Enhancing the Green Construction Performance Resilience in Infrastructure Projects: A Complexity Perspective

Abstract

Green construction in infrastructure projects has emerged as a crucial approach for reducing environmental impacts, yet its implementation is fraught with numerous uncertainties. To assess the capacity to maintain and restore green construction performance in complex environments, this study proposes the concept of Green Construction Performance Resilience (GCPR) for infrastructure projects and develops methodologies for its management and optimization. This study constructs a project network based on the labeled property graph (LPG) technique and demonstrates its dynamic evolution throughout the entire project lifecycle. A series of indicators for quantifying GCPR are constructed and applied, enabling the quantification of green construction performance resilience in infrastructure projects. An optimization method for GCPR based on genetic algorithms is proposed. Finally, the applicability and effectiveness of the proposed methodologies are validated through the analysis of real-world infrastructure project cases. The results demonstrate that the project network model can comprehensively capture the complexity of large-scale infrastructure projects, and that the GCPR indicators effectively measure green construction performance resilience, providing valuable decision-making support for project managers. The optimization algorithm has been validated and shown to improve the GCPR level of the project. This study enriches interdisciplinary research on project resilience and project complexity theory and provides project managers with quantitative analysis and visualization tools to facilitate the attainment of green construction performance objectives in infrastructure projects and accelerate the transition towards low-carbon practices.

1. Introduction

Against the backdrop of accelerating global economic growth and urbanization, infrastructure projects have become a crucial pillar in driving socio-economic development [1,2,3]. However, the construction processes of such projects are often accompanied by high levels of resource consumption, carbon emissions, and environmental degradation [4,5,6]. Reducing the environmental impacts of construction activities has thus emerged as a major challenge in infrastructure development. Promoting green construction practices in infrastructure projects is of great significance for mitigating climate change [7,8].

Green construction performance has attracted increasing attention from academia and industry to assess the environmental performance of construction processes of infrastructure projects. Green construction practices in infrastructure projects can effectively reduce long-term maintenance costs and improve energy efficiency. However, green construction performance is significantly affected by the inherent complexity of infrastructure projects. Variations in technical requirements, construction tasks, contractor composition, and construction environments lead to increased complexity, thereby imposing greater challenges to green transformation efforts [9,10]. Green construction not only involves traditional construction management but also introduces multiple sustainability requirements, such as reducing carbon emissions, enhancing resource efficiency, and optimizing energy consumption [11]. Achieving these goals often depends on interdisciplinary collaboration and advanced technological support, which further amplifies construction complexity. From the perspective of dissipative structure theory, infrastructure projects maintain continuous exchanges of resources and information with their external environment during the green construction process [12]. Green construction performance is influenced by multiple factors, including policy and regulatory frameworks, market demand, technological innovation, and construction process control. Over time, the interactions among different construction sections, contractors, tasks, and technologies evolve from a chaotic to a more ordered state, shaping the performance of green construction [13]. Therefore, investigating green construction performance from the perspective of project complexity helps to reveal its structural, dynamic, and uncertain characteristics, and provides a more scientific basis for decision-making.

Although significant progress has been made in academic research on green construction performance of projects and in-depth discussions have been conducted from multiple perspectives, studies exploring the evolution of green construction performance from the perspective of infrastructure project complexity remain a gap in the existing literature. Furthermore, static research focusing solely on influencing factors and measurement indicators is insufficient to effectively explain how infrastructure projects can adapt to and recover from external uncertainties during green construction, thereby improving green construction performance.

Against this background, this study proposes the concept of Green Construction Performance Resilience (GCPR), aiming to assess the ability of infrastructure projects to maintain and recover green construction performance under complex conditions. Based on project complexity theory, the labeled property graph technique is used to model the project network, illustrating its dynamic evolution as the project progresses. On this project network foundation, a series of GCPR indicators is developed and quantified to evaluate the resilience of green construction performance in infrastructure projects. To further enhance GCPR, the genetic algorithm is employed for optimization. Finally, the proposed approach is empirically validated through a real-world case study.

2. Literature Review

2.1. Green Construction Performance Indicators for Infrastructure Projects

The construction industry is progressively advancing towards green transformation, with infrastructure projects placing greater emphasis on the sustainability of the construction process [14]. The Triple Bottom Line (TBL) concept, proposed by Elkington in 1994, provides a theoretical foundation for measuring sustainable development performance [4,15,16]. This concept explains sustainability from three dimensions: economic, environmental, and social, and offers a framework for measuring green construction performance in infrastructure projects. As research on green construction deepens, scholars have expanded the evaluation indicator system for green construction performance. Tupebaitewe developed a sustainability measurement tool and identified “energy and atmosphere” as the most critical performance indicator [17]. Shad developed a set of indicators for evaluating the green construction performance of infrastructure projects, comprising eight criteria and 61 sub-criteria, with water use efficiency identified as a core factor [18]. Yao incorporated resource sustainability into the green construction performance evaluation system and divided green construction performance into four dimensions: resources, environment, economy, and society [14]. Meanwhile, Yadegaridehkordi further expanded the indicators for green construction performance, considering energy efficiency, sustainable site planning and management, materials and resources, water efficiency, and innovation as core indicators for measuring sustainability [19]. However, while there have been numerous discussions on measuring the green construction performance of infrastructure projects, most of these have been post-hoc static assessments, which have limited effectiveness in optimizing the green construction performance of infrastructure projects.

2.2. Complexity of Green Construction in Infrastructure Projects

Infrastructure projects exhibit uncertain features due to their structural, dynamic nature, and the interactions among various components [9,20]. The design and adjustment of construction processes, coordination and governance among stakeholders, and the rational supply and allocation of resources, such as construction materials and machinery within their construction and management processes, render the entire green construction process highly complex [20]. This complexity may lead to minor execution deviations in the early stages that, due to cascading effects, ultimately result in unpredictable environmental impacts [21]. Green construction in infrastructure projects imposes numerous sustainability requirements, such as reducing carbon emissions, enhancing resource utilization efficiency, and optimizing energy consumption [19,22,23]. The achievement of these performance objectives relies on interdisciplinary collaboration and advanced technological support [13]. The complexity of green construction in infrastructure projects exerts an inhibitory effect on green performance [24]. Hu investigated project complexity across sectors and organizations, explored the relationships between project structure, workflow, and coordination mechanisms, and proposed a more inclusive collaborative governance network to enhance green construction performance [25]. Other studies have attempted to measure project complexity to improve green construction performance. Luo developed a systematic computational method based on Bayesian inference to explore project complexity in uncertain environments [26]. Kermanshachi employed a Project Complexity Index (CI) to construct a complexity management model, enabling practitioners to reasonably prioritize the allocation of limited resources during construction [27]. Many studies have focused on surveys combining expert interviews, case analyses, and statistical methods to enhance green construction performance by identifying and mitigating key influencing factors of project complexity [28]. However, decision-making solely based on experience and intuition is often criticized [9].

The complexity of green construction and management in infrastructure projects typically manifests as multi-dimensional attributes, with a variety of inducing factors [9]. The construction teams for infrastructure projects are composed of diverse stakeholders who may come from different regions [9] and possess distinct values and interests [29]. These differences in interests can lead to free-riding behavior among organizations and greenwashing within organizations, thereby reducing the effectiveness of coordination and negotiation [30]. The green construction process of infrastructure projects is constrained by relevant environmental protection and energy-saving standards [14] and involves multiple stakeholders, such as government regulatory agencies, owners, contractors, suppliers, and environmental protection organizations [9]. Moreover, due to interdependencies among work tasks, an non-environmentally friendly outcome in one task can often lead to cascading failures in green construction management for other tasks or even the entire project [31]. Naveed pointed out in his research that temporary changes resulting from construction environment uncertainty, delayed information acquisition, and inefficient communication among stakeholders in infrastructure projects are all reasons for the low green construction performance of such projects [32]. Meanwhile, infrastructure projects often involve multiple construction sites, each with distinct environmental characteristics, and within the same site, there may be multiple specialized contractors working simultaneously or consecutively, further increasing the complexity of green construction management [13]. Unlike traditional construction, green construction technologies demand high levels of expertise and possess certain particularities [33]. The uncertainties and adaptability arising from the diversity of green construction technologies increase management difficulties, reflecting technological complexity [33,34]. Furthermore, this technological complexity has a negative impact on the green construction performance of infrastructure projects [9]. Existing research has demonstrated that project complexity directly influences project management performance, including aspects such as duration, cost, and quality [35]. Green construction not only involves these traditional performance indicators but also encompasses environmental and social sustainability goals [19], making it more susceptible to the effects of complexity. Based on a literature review of the complexity of infrastructure projects and its impact on green construction performance, this study categorizes the complexity of infrastructure projects into task complexity, organizational complexity, technological complexity, and spatial complexity [3,36,37], as shown in

Table 1. Types of complexity in infrastructure projects.

Type	Concept
Task Complexity	Overlapping relationships, task changes, and construction processes among tasks in the green construction of infrastructure projects.
Organizational Complexity	Organizational structure, cooperation, and competition among contractors, and inter-organizational information exchange in the green construction of infrastructure projects.
Technological Complexity	Use of new green construction technologies and materials, as well as innovative green building structures and functions.
Spatial Complexity	Traffic logistics, environmental sensitivity of green construction, and geological conditions in the green construction of infrastructure projects.

. These four types of complexity are interrelated and continuously influence the green construction performance of infrastructure projects over time [38,39,40,41].

2.3. Vulnerability of Green Construction Performance in Infrastructure Projects

The green construction performance of infrastructure projects exhibits a certain degree of vulnerability when confronted with project complexity. Focusing on the green construction and management processes of infrastructure, researchers have attempted to promote cross-organizational interactions among stakeholders through positive group norms [42,43], aiming to achieve close collaboration among different stakeholders and enhance green construction performance [44,45]. They argue that the realization of green construction performance in infrastructure projects not only depends on the application of green materials and technologies but is also influenced by inter-organizational collaborative relationships, which exhibit vulnerability [46,47]. This vulnerability is primarily manifested in the project’s inability to achieve anticipated sustainable development goals under the influence of external shocks or internal management deficiencies [46], particularly due to the cascading effects of implementation deviations in green construction on subsequent construction activities. In infrastructure projects, contractors exhibit strong interdependencies, especially at critical time nodes and important construction sites, where excessive dependency may become a vulnerability point for the project’s green construction performance [48]. When multiple contractors collaborate simultaneously in the same temporal and spatial coordinates, it may trigger a risk resonance effect, exacerbating the likelihood of cascading performance disruptions (CPDs) [49,50] and ultimately reducing the green construction performance of infrastructure projects. CPD refers to a dynamic process in which initial disturbances—such as local task failures, specification changes, or schedule delays—occur within an interdependent and dynamically evolving project system, gradually spreading and amplifying throughout the project. This propagation ultimately leads to a decline in overall project performance, or even systemic failure [13,51,52]. Contractors serve as the core executing entities in the green construction of infrastructure projects, and their construction technologies, resource management, and implementation of environmental protection measures directly determine the stability and adaptability of green construction performance [37]. In the green construction of infrastructure projects, high-impact contractors often occupy key positions within the project network, and fluctuations in their green construction performance may lead to cascading performance disruption (CPD) effects, thereby reducing the overall green construction performance of the project [13]. High-impact contractors or construction sites may become vulnerability points for the project’s green construction performance [53]. The interaction mechanism between the complexity of infrastructure projects and their green construction performance is illustrated in Figure 1.

2.4. Resilience of Green Construction Performance in Infrastructure Projects

The enhancement of green construction performance in infrastructure projects not only relies on the analysis of measurement indicators and their influencing factors but also necessitates an exploration of how it can adapt to and recover from the impacts of uncertainties. In traditional research on project construction performance, the resilience perspective has been widely applied to performance management. Studies have found that if the task dependencies among different stakeholders are not effectively managed, it will increase project complexity and subsequently reduce project resilience [13,38]. Promoting unity and collaboration among team members, as well as appreciating their capabilities, contributes to improving project resilience [54]. Additionally, factors, such as working conditions, trust, team building, and commitment, can all drive team capabilities and ultimately enhance project resilience [55]. Project resilience can be defined as the ability of a project system to understand its surrounding environment and vulnerabilities, and to recover from disruptive practices and achieve its goals [38]. Therefore, this study proposes the concept of Green Construction Performance Resilience (GCPR) to measure the adaptability and recovery capacity of green construction performance in infrastructure projects under uncertain environments. Green Construction Performance Resilience can be defined as the ability of a project to understand the vulnerabilities of its surrounding environment, effectively adapt and adjust, and ultimately achieve its green construction performance goals.

Complex Dynamic Network Theory (CDNT) is considered an ideal approach for analyzing project complexity and uncertainty [13,37,56,57]. CDNT provides a robust set of tools for modeling, analyzing, and understanding dynamic interdependencies within complex systems [13,58]. Among these tools, the Labeled Property Graph (LPG), as an expression of complex networks, exhibits exceptional flexibility and adaptability. It is particularly well-suited for dynamic and scalable network analysis tasks [59]. LPG is renowned for its efficient storage capabilities, rapid traversal abilities, and effectiveness in modeling real-world scenarios [59,60].

The application of green construction in infrastructure projects has become a vital approach to promoting sustainable development and reducing environmental impacts. However, existing research predominantly focuses on static, ex-post performance evaluations, overlooking the dynamic influence of project complexity on green construction performance. The complexity of infrastructure projects, encompassing task, organizational, technological, and spatial dimensions, renders their green construction performance notably vulnerable, particularly when confronted with uncertainty shocks that may trigger cascading performance disturbances (CPDs), ultimately undermining the achievement of sustainability goals. Although prior studies have endeavored to enhance green construction performance by optimizing collaboration networks, quantifying complexity, or refining management strategies, they largely rely on empirical decision-making and lack systematic methodologies to bolster the adaptability and resilience of green construction performance in dynamic environments. Moreover, the enhancement of green construction performance in infrastructure projects necessitates not only the analysis of measurement indicators and their influencing factors but also an exploration of how it can adapt to and recover from the impacts of uncertainties. Therefore, this study introduces the concept of green construction performance resilience (GCPR), aiming to gauge the capacity of infrastructure projects to maintain and restore green construction performance under complex circumstances. Building upon project complexity theory and integrating the labeled property graph (LPG) modeling approach, a framework for managing, quantifying, and optimizing GCPR is constructed to address the gaps in existing research. This theoretical perspective not only broadens the scope of green construction performance management but also equips project managers with novel tools to tackle uncertainties, thereby facilitating the low-carbon transformation of infrastructure projects and the realization of sustainable development objectives.

3. Methodology

The technical roadmap of this study is illustrated in Figure 2. The research methodology can be broadly divided into four parts. The first part involves the construction and dynamic evolution of infrastructure project networks. Based on the infrastructure project schedule, a network model is constructed using Labeled Property Graph (LPG) technology, grounded in project complexity theory, and subsequently subjected to dynamic evolution. The second part focuses on the construction and application of quantitative indicators for quantifying the green construction performance resilience in infrastructure projects. Utilizing an infrastructure project network model, a series of critical indicators are provided for the quantification of green construction performance resilience. These indicators not only furnish critical data for quantifying green construction performance resilience but also assist project managers in identifying potential vulnerabilities in the green construction performance of projects. The third part involves the quantification of green construction performance resilience of infrastructure projects. The final part, the fourth, addresses the optimization of green construction performance resilience of infrastructure projects. The following sections will elaborate on these steps in detail.

3.1. Construction and Dynamic Evolution of Infrastructure Project Network Models

(1)

Initial Project Network Construction

Figure 3 illustrates the specific steps for constructing the initial project network based on an infrastructure project schedule. The data for initial network modeling originates from the project schedule, which contains rich multi-dimensional project data. Furthermore, as a core document in infrastructure project management, the project schedule is widely recognized by stakeholders in the project. This modeling process primarily encompasses two phases: dimension construction and relationship establishment. In the dimension construction phase of the initial project network, we partition the project network into time dimension, task dimension, contractor dimension, and location dimension based on the project schedule, corresponding, respectively, to task complexity, organizational complexity, and spatial complexity. In contrast, technical complexity is more implicit and systemic, making it difficult to model directly through a single dimension. Instead, it acts more as a systemic complexity permeating various elements of the project, indirectly reflected across multiple dimensions [37,40]. For instance, higher technical requirements may lead to prolonged task durations, strengthened logical relationships, and consequently increased task complexity and scheduling pressure. Meanwhile, they may also necessitate more specialized contractor resources, enhancing the complexity of organizational structures, or impose construction constraints under specific spatial conditions, increasing the difficulty of spatial coordination. Therefore, technical complexity is regarded as a cross-dimensional intrinsic driving factor in the project network. Subsequently, elements within each dimension are mapped as node sets in the initial project network. This approach effectively preserves the multi-dimensional structural characteristics of the project schedule, enabling different types of information to be uniformly integrated into the network model for modeling and analysis. Specifically, the task dimension corresponds to various construction tasks in the schedule, with each task mapped as an independent node; the location dimension corresponds to construction location, with each construction location mapped as a location node; the contractor dimension represents contractors in the project, with each contractor modeled as an independent node; and the time dimension is discretized based on task start and end times, transformed into multiple time-window nodes. Additionally, each node is assigned a “dimension property” to specify its dimension type, such as setting the dimension property of task node Task7 as “Task”.

In the phase of constructing relationships between dimensions, edges are established between nodes from different dimensions based on the correspondences between dimensions in the schedule (such as the associations between tasks and their construction locations, contractors, and execution times), thereby constructing a multi-dimensionally interconnected project network structure. The setup of these edges reflects key semantic relationships in the actual operation of the project, such as “who executes the task, where it is executed, and when it is executed”. In this modeling, it is explicitly stipulated that edges are not established between nodes within the same dimension. This is based on the following considerations: firstly, the schedule itself does not define direct relationships between entities within the same dimension, lacking clear semantic support; secondly, forcibly establishing edges within the same dimension may introduce redundant connections or semantic confusion, reducing the clarity and interpretability of the network structure. Therefore, this modeling method ensures a logically clear and semantically accurate network structure by restricting the establishment of edges to occur only between different dimensions, facilitating subsequent network analysis and visualization processing. In specific implementation, each task node is fully connected to its corresponding location node, contractor node, and time-window nodes. Since edges are established among the four types of nodes: task, location, contractor, and time, local areas exhibit typical small-world network structure characteristics. Furthermore, as the task data in the schedule does not reflect a fixed sequence or directionality among the four dimensions, the initial project network is modeled as an undirected graph.

(2)

Dynamic Evolution of Project Networks

As the infrastructure project evolves over time and space, the network structure formed by the project also undergoes changes. Consequently, as depicted in Figure 4, the initial project network continually adjusts with the progression of time, primarily manifested as the disappearance of corresponding time-window nodes and associated changes in related nodes. Taking the project schedule shown in Figure 4a as an example (where task start and end times have been converted into time-windows), the initial project network, as outlined in Step 1 for constructing the initial project network, can be derived as illustrated in Figure 4b. This initial project network can also be regarded as the network structure at the project’s inception. Due to the advancement of construction progress, the time-window node D1 vanishes from the network. All relationships associated with D1 (task-contractor-construction site) consequently disappear. With the disappearance of node D1, task M1 occurring within time-window D1 and contractor C1 solely engaged in construction during time-window D1 will also exit the project upon the completion of their tasks, represented in the network by the disappearance of task node M1 and contractor node C1. Figure 4c illustrates the network structure formed after the disappearance of D1. As project progress continues, time-window D2 disappears, and the project network has been further downsized; (as shown in Figure 4d). With the disappearance of time-window D3, all nodes in the network are no longer associated with any time-window. All task, contractor, and construction site nodes are removed. The project network completely vanishes, marking the official conclusion of the project.

3.2. Construction and Application of Quantitative Indicators for Quantifying the Green Construction Performance Resilience in Infrastructure Projects

(1)

Identification of the importance of contractors and construction locations

In the green construction process of infrastructure projects, contractors and construction locations with high importance are usually situated at critical positions within the project network. Fluctuations in their green construction performance may trigger “single point of failure” risks, thereby undermining the green construction performance of the entire project [15]. These high importance nodes are often the vulnerable points of the project’s green construction performance [52]. Contractors with high importance typically undertake multiple critical tasks, and a decline in their green construction performance may initiate a chain reaction, ultimately affecting the green construction performance of the entire project. Conversely, construction locations with high importance frequently concentrate a substantial number of construction tasks. If management becomes disorganized, it may impact multiple tasks and contractors, further threatening the project’s green construction performance objectives.

Contractors and construction locations with a high degree of importance exhibit strong network centrality. Currently, various methods within Complex Dynamic Network Theory (CDNT) can be employed to quantify the importance and centrality of nodes in a network, such as weighted degree centrality (WDC), network density (ND), and eigenvector centrality (EC). However, methods, like ND and WDC, only measure node importance by examining local network structures. To address this limitation, the PageRank algorithm was introduced, which calculates node importance by considering the entire network structure, thereby providing a more comprehensive perspective on the importance and centrality of nodes in a network. This makes it particularly effective in ranking influential nodes [61,62]. The ability of the PageRank algorithm to consider the overall structure of a network aligns precisely with the needs of this study, given the multifaceted complexity of infrastructure projects.

The importance of contractors and construction locations calculations are performed based on the dynamic evolution of project networks. The PageRank algorithm is applied to the initial project network to compute node importance. PageRank calculations are conducted using the NetworkX library. After computation, each node’s PageRank value is obtained. However, these current PageRank values represent node importance relative to the entire network and lack practical significance in this study, as construction sites and contractors cannot be directly compared in terms of importance. Therefore, it is necessary to transform node importance from the network-wide scale to the dimension-specific scale. This is achieved by normalizing node PageRank values according to dimensions. Ultimately, the importance levels of nodes within different dimensions at the initial state are determined. As construction progresses and the project network structure evolves, the aforementioned calculation steps are repeated to obtain node importance under different dimensions at the current time-window. The same procedure is followed for node importance in subsequent time-windows. By sequencing node importance according to temporal windows, we can ultimately derive the evolution of node importance across different dimensions over time.

(2)

Identification of Contractor Dependencies and Calculation of Dependency Levels

Identifying contractors with high dependency relationships is crucial for enhancing the green construction performance of infrastructure projects, as they may represent vulnerabilities in the project’s green construction performance [49,50]. The level of dependency between two contractors depends on several factors: whether they work jointly at critical times; whether they operate at the same or adjacent critical construction locations; and their importance within the project. Based on this rationale, the dependency level between contractors can be calculated using the following formula:

$$\begin{array}{c}D = T \times \left({\displaystyle \sum _{k=1}^n}{L}_k\right)\times C_i\times C_j\end{array}$$

(1)

where, D is the dependency level between contractors i and j; T is the importance of the time-window node; L_k is the importance of the k-th common work location at that time-window point, with multiple common work locations for different contractors increasing dependency risks; C_i and C_j is the importance of contractors i and j, respectively, such that lower contractor importance correlates with reduced dependency levels between them; and n is the number of common work locations.

(3)

Quantitative Analysis of Contractor Influence Scope and Influence Degree

To quantify the potential impact of the deterioration of a contractor’s green construction performance on the green construction performance of infrastructure projects, this study introduces two concepts: the Green Construction Performance Influence Scope (GCPI-Scope) and the Green Construction Performance Influence Degree (GCPI-Degree). These concepts are proposed to identify and evaluate the role of contractors in the overall green construction performance of a project. The GCPI-Scope aims to measure the equivalent duration of the impact of a single contractor’s green construction performance issues on the overall project timeline. If a contractor has a large GCPI-Scope value, it implies that any potential decline in their performance will affect a broader construction period, thereby causing a significant impact on the green construction performance of the infrastructure project. Such contractors often become vulnerabilities in the project’s green construction performance, and any decline in their construction capabilities may lead to a deterioration in the overall green construction performance of the project. The calculation of GCPI-Scope can be found in Equation (2).

$${GCPI-Scope}_i={\displaystyle \frac{{\sum }_t{w}_i\left(t\right)}{{\sum }_t{w}_{total}\left(t\right)}}\times T$$

(2)

where, GCPI-Scope_i is the GCPI-Scope value for contractor i; w_i(t) is the importance of contractor i on day t; w_total(t) is the sum of the importance of all contractors on day t, and T is the total duration of the infrastructure project.

The GCPI-Degree is utilized to quantify the intensity of the impact caused by a decline in a contractor’s green construction performance on the overall green construction performance of an infrastructure project. The decline in green construction performance among different contractors does not uniformly affect the project, this disparity depends on the importance of the contractor at a specific time, the magnitude of the deterioration in construction performance, and the buffering effect of their historical green construction performance. The calculations for GCPI-Degree can be found in Equations (3) and (4).

$${GCPI-Degree}_i=\left({\sum }_t{w}_i\left(t\right)\times {(1-p}_i\left(t\right))\right)\times \left[1-\sigma \left(H_i\right)\right]$$

(3)

$$\sigma \left(H_i\right)={\displaystyle \frac{1}{1+e^{-k\left(H_i-H_0\right)}}}$$

(4)

where, GCPI-Degree_i is the GCPI-Degree value for contractor i; w_i(t) is the importance of contractor i on day t; p_i(t) is the performance deterioration value of contractor i on day t, which project managers can set p_i(t) to proactively assess the impact of reduced contractor performance on the project’s green construction performance. [1 − σ(H_i)] is the historical performance adjustment factor, quantifying the buffering effect of a contractor’s historical performance. If a contractor’s historical green construction performance is more excellent, it demonstrates stronger buffering against potential performance shortfalls, thereby significantly mitigating the overall impact of potential performance deterioration. Here, H_i is the historical performance score of contractor i, ranging from [0, 1]. The better the past performance of contractor i, the closer its historical performance score approaches 1, consequently reducing the overall impact of the contractor’s current performance decline on the project. H₀ is the baseline value, which can be set to 0.5 by default, representing an average level of historical green construction performance; k is the adjustment coefficient used to control the buffering effect of historical performance, and it is recommended to be set within the range of [6, 10].

(4)

Prediction of Green Construction Performance in Infrastructure Projects

In this study, we focus on six green construction performance indicators for infrastructure projects, which are widely applied in green construction management to assess the sustainability of the construction process [19,63,64]. These indicators are Material Utilization Rate (MU), Waste Recycling Rate (WR), Water Reuse Rate in Construction (WUR), Renewable Energy Utilization Rate (REU), Wastewater Compliance Rate (WCR), and Green Construction Technology Application Rate (GCTA). The calculations for each indicator are presented in

Table 2. Green construction performance indicators for infrastructure projects.

Material Utilization Rate (MU) = Actual Material Usage (t/m3)/Total Purchased Material Quantity (t/m3)	This indicator reflects the efficiency of material procurement and actual usage, effectively measuring resource wastage during the construction process. It is widely adopted in green building assessment systems such as LEED and serves as a crucial indicator in the category of resource conservation.
Waste Recycling Rate (WR) = Recyclable Construction Waste Volume (t)/Total Construction Waste Volume (t)	This metric assesses the resource recovery capacity of waste generated during construction. It is a core indicator for evaluating the environmental friendliness of construction sites and reflects the management ability to mitigate negative impacts on ecosystems.
Water Reuse Rate in Construction (WUR) = Reused Water Volume (m3)/Total Construction Water Consumption (m3)	Construction water, though often overlooked, is a vital resource in green construction. The extent of its reuse is a key measure of water resource management efficiency in construction, particularly in areas facing water scarcity.
Renewable Energy Utilization Rate (REU) = Renewable Energy Consumption (kWh)/Total Energy Consumption (kWh)	This indicator measures the effective integration of renewable energy sources (such as solar and wind energy) during the construction process. It is a critical metric for achieving low-carbon construction and energy transition, reflecting the greenness of the project’s energy structure.
Wastewater Compliance Rate (WCR) = Compliant Wastewater Discharge Volume (m3)/Total Construction Wastewater Volume (m3)	This reflects the construction site’s capacity to treat and meet wastewater discharge standards. It is a core aspect of environmental protection and government regulatory assessment, particularly crucial in urban environments.
Green Construction Technology Application Rate (GCTA) = Total Construction Processes Employing Green Construction Technologies/Total Construction Processes	This represents the extent of adoption of new green construction techniques and technologies (such as prefabricated construction and intelligent energy-saving systems). It reflects the project’s level of green innovation and serves as a representative indicator for assessing the advancement of green construction practices.

. These indicators encompass the four core objectives of green construction: resource conservation, low-carbon energy, pollution control, and technological innovation, and they are representative, quantifiable, and accessible, enabling a systematic reflection of the green construction performance level of infrastructure projects. Additionally, they possess feasibility and monitorability in practical engineering applications [65,66,67].

The green construction performance of infrastructure projects is not only influenced by the behavior of individual contractors but also by the contractor’s importance within the project network. Therefore, we assess the potential green construction performance level of projects in future time periods by correlating each contractor’s performance indicators with their importance within the project network. The calculation process can be referred to in Equation (5).

$$\begin{array}{c}KPIP={\displaystyle \sum _{k=1}^n}\left({KPIC}_k\times {IM}_k\right)\end{array}$$

(5)

where, KPIP is any green construction performance considered for the project, such as Material Utilization Rate (MU); n is the number of contractors associated with the green construction performance of the current project; KPIC_k is the green construction performance of contractor k; and IM_k indicates the degree of importance of contractor k.

3.3. Quantification of Green Construction Performance Resilience in Infrastructure Projects

The green construction performance resilience of infrastructure projects comprises three components: resistance capacity, adaptability capacity, and recovery capacity. Each component reflects the project system’s distinct capabilities when confronted with disturbances. The resistance capacity aims to measure the project’s ability to withstand external disturbances and maintain green construction performance. The adaptability capacity aims to evaluate the project’s capability to adjust its strategies to adapt to disturbances. The recovery capacity aims to assess the project’s ability to revert to the targeted level of green construction performance after being impacted. The calculation of green construction performance resilience can be referred to in Equation (6).

$$R=\alpha \times R_{resist}+\beta \times R_{adapt}+\gamma \times R_{recover}$$

(6)

where R is the green construction performance resilience of the infrastructure project; R_resist is the resistance capacity; R_adapt is the adaptability capacity; R_recover is the recovery capacity; α, β, and γ are weights with the condition that α + β + γ = 1. Project managers can adjust the weight distribution according to different construction scenarios to adapt to varying construction conditions and management requirements.

The calculations of resistance capacity, adaptability capacity, and recovery capacity can be referred to in Equations (7) and (9).

$$R_{resist}=1-{\displaystyle \frac{{\sum }_i\left(p_i\times {Scope}_i\right)}{T\times P_{threshold}}}$$

(7)

$$R_{adapt}=1-Gini\left(\left\{w_i\right\}\right)$$

(8)

$$R_{recover}={\displaystyle \frac{\sum _{i=1}^{n}min\left(\frac{A{P}_i}{T{P}_i},1\right)}{n}}$$

(9)

where pi is the GCPI-Degree value of contractor i; Scope_i is the GCPI-Scope value of contractor i; P_threshd is the maximum interference threshold that the system can withstand; T is the total project duration; Gini({w_i}) is the Gini coefficient of contractor importance in the infrastructure project, which is used to measure the imbalance in the distribution of contractor importance within the project. A higher Gini coefficient indicates a more uneven distribution of contractor importance in the project and, consequently, lower adaptability. Here, w_i is the average importance of contractor i; AP_i is the actual performance of the project’s green construction performance indicator i, and TP_i is the target green construction performance.

3.4. Optimization of Green Construction Performance Resilience in Infrastructure Projects

To enhance the Green Construction Performance Resilience (GCPR) of infrastructure projects, this study employs a Genetic Algorithm (GA) as a metaheuristic optimization strategy to systematically optimize the project schedule and contractor allocation. The genetic algorithm is a powerful technique for finding approximate optimal solutions to such problems and has been proven highly effective in solving many challenging multi-objective optimization problems [13,37,68]. This study focuses on improving green construction performance resilience by adjusting the project schedule. Therefore, the genetic algorithm is adopted, as it can efficiently rank non-dominated solutions and generate a set of Pareto-optimal solutions in a single run.

The Genetic Algorithm approach uses GCPR as the objective function for optimization and iteratively searches for the optimal scheduling plan for maximizing GCPR through processes of population initialization, selection, cross-over, mutation, and iterative updating. The optimization process is illustrated in Figure 5. Initially, parameters need to be set. The GA optimization strategy involves two types of parameters: task parameters and algorithm parameters. Task parameters are based on the structural characteristics of the original project schedule, including start/end times, free float, task dependencies, and contractor assignments for each task. Algorithm parameters include population size, maximum number of iterations, cross-over probability, and mutation probability, which control the search efficiency and global optimization capability of the genetic algorithm. Higher iteration counts and cross-over/mutation probabilities facilitate the generation of better solutions but may also incur higher computational costs. Subsequently, cyclic optimization is conducted: In each iteration, the GA first evaluates the fitness of the current population based on the GCPR of the infrastructure project. New generations of individuals are then generated through selection, cross-over, and mutation operations to construct new project schedules. Each individual represents a complete and feasible project schedule, including the sequencing of construction tasks and contractor assignments. These schedules are transformed into project networks, and the GCPR of each individual is calculated. If the GCPR of the current best individual exceeds the historical best value, the optimal solution is updated, and the search continues. This process persists until the number of iterations reaches the predefined limit or the optimal solution converges. Ultimately, this study outputs the optimized project schedule and its corresponding GCPR. The detailed implementation process is outlined in Algorithm 1.

Algorithm 1 Genetic Algorithm for Optimizing Green Construction Performance Resilience
Input:   Original project schedule O_SD   Maximum number of generations max_gen   Mutation rate MR Output:   Optimized project schedule P_SD   Optimized green construction performance resilience R
1:	Initialize population P = {S_1, S_2, …, S_n} by generating pop_size variations of O_SD  # Each individual S_i is a feasible project schedule (chromosome)
2:	For each individual S_i in P do
3:	R_i = Resilience_analyse(S_i)  # resilience evaluation
4:	For generation = 1 to max_gen do
5:	Select parents based on R_i
6:	Apply crossover on selected parents to generate offspring  # Crossover mixes task sequences or contractor task
7:	Apply mutation on offspring with rate MR  # Slightly adjust task timing or contractor allocation
8:	Evaluate fitness R_j = Resilience_analyse(S_j) for each offspring S_j
9:	Combine parents and offspring to form new population
10:	Select top pop_size individuals for next generation
11:	End for
12:	Return best schedule P_SD with maximum R in final population

4. Case Study

The method developed in this study will be applied to a real infrastructure project to validate the effectiveness of its methodology. Due to confidentiality concerns related to relevant contractual documents, the project name and specific details of the contractors are omitted in this paper. The infrastructure project comprises two main buildings and their ancillary facilities. Given its substantial scale, the project is divided into five construction zones, involving a total of ten contractors, with a planned total duration of 14 months (375 days for the construction phase). Details of the participating contractors and their descriptions are provided in

Table 3. Contractor information.

Contractor ID	Type
C0	Civil works contractor
C1	Waterproofing works contractor
C2	Electrical contractor
C3	Drywall and interior finishing contractor
C4	Doors and windows contractor
C5	Plumbing and drainage contractor
C6	HVAC (Heating, Ventilation, and Air Conditioning) contractor
C7	Structural steel erection contractor
C8	Road construction contractor
C9	Landscaping contractor

. Given the large scale of this infrastructure project, its construction activities may have significant impacts on the surrounding environment. Therefore, effective management of the project is crucial to ensure it meets the predetermined green construction performance objectives. This characteristic makes the project an ideal application scenario for the method proposed in this study, facilitating a comprehensive demonstration of the method’s applicability and effectiveness in practical engineering. However, constrained by data availability and the absence of contractors’ historical performance data, this study makes reasonable assumptions regarding the contractors’ green construction performance indicators in the case analysis. Historical performance scores (Hi) are assigned to each contractor, and the project’s target performance values are clearly defined. Specific assumptions are presented in

Table 4. Green construction performance of contractor.

Contractor	Hi	Green Construction Performance
		MU	WR	WUR	REU	WSR	GTA
C0	0.85	0.85	0.65	0.45	0.20	0.90	0.65
C1	0.79	0.90	0.70	0.55	0.15	0.95	0.70
C2	0.71	0.80	0.55	0.60	0.30	0.85	0.60
C3	0.74	0.78	0.60	0.50	0.10	0.92	0.75
C4	0.79	0.92	0.80	0.70	0.18	0.88	0.68
C5	0.76	0.82	0.68	0.35	0.25	0.96	0.72
C6	0.71	0.79	0.58	0.65	0.35	0.82	0.62
C7	0.75	0.87	0.75	0.40	0.12	0.93	0.67
C8	0.77	0.84	0.62	0.80	0.22	0.87	0.71
C9	0.85	0.95	0.85	0.52	0.40	0.98	0.80
Target		0.90	0.75	0.65	0.25	1.00	0.60

. Through this approach, this study can evaluate the practical feasibility of the proposed method under data-limited conditions and verify its role in enhancing the green construction performance resilience of infrastructure projects. The modeling of green construction complexity in infrastructure projects, the calculation of related indicators, resilience assessment, and optimization steps are illustrated in Figure 6.

4.1. Construction of the Infrastructure Project Network Model

The total duration of the construction phase for this infrastructure project is 375 days. Using months as the time-window may overlook subtle changes within the project, potentially leading to analytical deviations. Therefore, this study sets the time-window to days to ensure the accuracy of the analysis. Initially, the initial network model of the infrastructure project is constructed. The project schedule of the infrastructure project is processed by removing hierarchical structures used solely for project task arrangement to enhance the convenience of data processing. The start and end times of tasks are converted into corresponding time-windows to ensure the accuracy of the temporal dimension. Nodes are established based on multiple dimensions such as construction tasks, contractors, and construction location, and associations between different dimensions are constructed. Ultimately, after the aforementioned processing, the initial project network model is formed, which comprises 515 nodes and 3658 edges. Subsequently, this study conducts a dynamic evolution of the project network based on the initial network model to simulate the dynamic changes in time and space during the construction process.

4.2. Calculation and Application of Resilience Indicators for Green Construction Performance in Case Infrastructure Projects

(1)

Identification of the importance of contractors and construction locations

Following the methodology outlined in Section 3.2, this study employs the PageRank algorithm to calculate the importance of each node within the project network and normalizes the node importance scores across different dimensions. Ultimately, the temporal variations in node importance for the infrastructure project in the dimensions of contractors and construction location are depicted in Figure 7 and Figure 8.

In Figure 7, the darker the color, the higher the importance of the contractor during that time-window. The C0 civil engineering contractor is involved throughout the entire construction period and holds a relatively high level of importance across multiple time intervals. This characteristic indicates that the C0 contractor has a significant impact on the green construction performance of the infrastructure project, and fluctuations in its green construction performance may lead to deterioration in the overall green construction performance of the project. From Days 43 to 115 (the single-contractor phase), C0 becomes the sole construction unit, and the stability of its green construction performance directly determines the overall green construction performance of the project. As there are no other contractors to share the risks, the project manager needs to closely monitor the construction status of C0 to ensure the achievement of its green construction performance goals. From Days 140 to 305 (the high-contractor-density phase): A total of eight contractors enter the site for construction during this period, and the importance of C0, C2, C3, and C5 significantly increases, indicating that they undertake more critical tasks and play a role in multiple important time-windows. Due to the dense overlap of construction activities among multiple contractors, a decline in the green construction performance of some contractors may lead to the occurrence of CPD, thereby deteriorating the overall green construction performance of the infrastructure project. From Day 305 until the project’s completion, the tasks of contractors C0, C5, C8, and C9 are highly overlapping, and their interdependencies increase. When they work separately, their importance correspondingly rises, necessitating enhanced monitoring of their green construction performance.

As can be observed from Figure 8, the importance of different construction locations within the infrastructure project varies over time, providing the project manager with the basis for dynamic management of the construction layout. During Days 1 to 10, L0 is the sole construction location and holds the highest importance, necessitating the assurance that the construction environment complies with green construction standards. From Days 10 to 230, L0 and L1 undergo simultaneous construction with similar levels of importance. As L0 and L1 are the construction locations for the two main buildings in the infrastructure project, the project manager must pay particular attention to the green construction performance in these two locations. From Days 230 to 385, the project is under construction at multiple sites, including L0 to L3, with increased interactions between locations and a rise in the complexity of construction location management. Poor management may lead to construction chaos, thereby affecting the achievement of green construction performance goals. Therefore, during this phase, efficient construction scheduling strategies should be employed to reasonably arrange the sequence of tasks at different construction locations and mitigate the risks associated with cross-construction.

(2)

Identification of Contractor Dependencies and Calculation of Dependency Levels

The complexity of infrastructure projects largely stems from the interdependencies among contractors [13,37]. Due to the strong need for collaborative operations among contractors, certain high-level dependencies often emerge as vulnerability points for the green construction performance of infrastructure projects [48]. Furthermore, such interdependencies may trigger CPD, leading to the spread of risks throughout the project and ultimately affecting its green construction performance. Additionally, the higher the level of dependency among contractors, the greater the risk of inducing CPD. Therefore, project managers should prioritize examining the dependencies among contractors during the management process and pay extra attention to contractors with higher dependency levels to prevent the loss of control over the overall green construction performance goals due to fluctuations in individual green construction performance. Based on calculations using the relevant formulas outlined in Section 3.2, the changes in dependencies among contractors in the case infrastructure project over time are depicted in Figure 9.

In Figure 9, each vertical bar indicates the existence of a dependency between the corresponding contractors. The darker the color of the bar, the stronger the dependency. During the early stage of the case infrastructure project, the dependencies among contractors were relatively straightforward, with each contractor’s tasks being relatively independent. The dependencies among contractors primarily concentrated in the mid-stage of project implementation. As an increasing number of contractors entered the construction site successively, the interdependencies among them became more intricate, leading to gradually escalating the risk of CPD. Among these, the dependency relationship between Contractor C0 and Contractor C3 persisted for a relatively long duration, spanning from Day 142 to Day 225, totaling 83 days. During this period, the dependency level between the two exhibited a fluctuating trend, initially weakening, then strengthening, and subsequently declining gradually. This fluctuation indicates that during certain construction phases, the work progress of C3 may have been directly influenced by C0, and any issues in green construction practices by C0 could potentially lead to a deterioration in the green construction performance of C3, thereby impacting the overall green construction performance. Although the duration of dependency between C3 and C5, as well as between C2 and C5, was shorter than that between C0 and C3, the dependency levels were higher and continued to intensify over time. This situation suggests that the work of C5 is highly susceptible to the influence of C3 and C2, and fluctuations in the green construction performance of C5 could similarly reverberate to affect C3 and C2, forming a bidirectional dependency. If the green construction performance of either party declines, it may trigger a broader CPD, exacerbating the deterioration of the project’s green construction performance. Therefore, project managers need to focus on supervising the construction coordination between C3-C5 and C2-C5 to ensure that all parties maintain stable green construction performance during the high-dependency phase. For instance, establishing a real-time monitoring system to dynamically assess the green construction performance of highly dependent contractors. Despite the extremely short duration of the dependency relationship between Contractor C0 and C8, the dependency level was exceedingly high and consistently maintained at an extremely high level throughout the existence of the dependency. This high-dependency relationship emerged during the project closeout stage, and if left unmanaged, it could potentially exacerbate CPD risks and impact the final green construction performance of the entire project.

(3)

Quantitative Analysis of the Influence Scope and Degree of Contractors’ Green Construction Performance

In infrastructure projects, fluctuations in contractors’ green construction performance may significantly impact the overall green construction performance of the project. Therefore, this study quantifies the boundary of influence and the specific impact of a decline in green construction performance on infrastructure projects by calculating the influence scope (GCPI-Scope) and influence degree (GCPI-Degree) of contractors’ green construction performance. If a contractor’s GCPI-Scope and GCPI-Degree are relatively high, it indicates that the contractor represents a vulnerability in the project’s green construction performance. Once this contractor’s green construction performance deteriorates, it may pose a serious threat to the overall green construction goals of the project, thereby reducing its green construction performance. To quantify this impact, this study assumes a 10% deterioration in contractors’ green construction performance and analyzes each contractor in the case infrastructure project. The results are shown in Figure 10.

In Figure 10, from the perspective of GCPI-Scope, Contractor C0’s influence scope of green construction performance is significantly higher than that of other contractors. This implies that deterioration in C0’s green construction performance will affect a broader range of tasks and contractors, potentially impacting multiple critical nodes in terms of construction quality, resource consumption, and environmental effects. From the perspective of GCPI-Degree, Contractor C0 still occupies the top position, followed by Contractors C2, C3, and C5. This indicates that a decline in C0’s green construction performance not only affects a wide range but also exerts a relatively severe direct impact on the overall green construction performance of the infrastructure project. However, despite C0’s large GCPI-Scope, its historically high-performance score buffers, to some extent, the impact of C0’s performance deterioration on the project. Given C0’s high influence scope and degree, project managers should strengthen construction management for C0, including real-time monitoring of its construction performance, optimizing resource allocation, and ensuring strict adherence to green construction standards to prevent widespread impacts resulting from deterioration in green construction performance.

To further explore the effect of the setting of contractors’ historical performance scores on their GCPI-Degree, this study conducted additional calculations and analyses to help project managers clarify the impact of contractor selection on the green construction performance of infrastructure projects. In Formulas (3) and (4), k is the adjustment coefficient used to control the buffering effect of historical performance. The results in Figure 11a indicate that when the value of k is small, the buffering effect of historical performance is weak, and the contractors’ current green construction performance has a greater impact on their GCPI-Degree. Research shows that in project environments characterized by high complexity, high uncertainty, significant risks in critical tasks, and high long-term collaboration value, project managers are more inclined to trust contractors with excellent historical performance [69]. Therefore, in such circumstances, the value of k can be appropriately increased to make GCPI-Degree more sensitive to historical performance. When k = 8, the impact of contractors’ historical performance scores on their GCPI-Degree in the case infrastructure project is shown in Figure 11b. Contractor C0’s GCPI-Degree is most sensitive to its historical performance score. When C0’s historical performance score decreases, its GCPI-Degree increases most significantly, indicating that this contractor’s historical performance is crucial to the stability of the project’s green construction performance. This further suggests that during the contractor selection process for the case infrastructure project, project managers should pay special attention to C0 to ensure it has sufficient capacity to cope with the risk of deterioration in green construction performance.

4.3. Prediction of Green Construction Performance in Infrastructure Projects

The predicted values of green construction performance over time in the case infrastructure project are shown in Figure 12.

In Figure 12, the color intensity represents the level of green construction performance in the case infrastructure project. The darker the color, the better the performance during that time window. The green construction performance of the case infrastructure project in the later stages is notably superior to that in the early stages. This trend is primarily influenced by the varying levels of green construction performance among contractors across different construction phases. From Day 43 to Day 115, Contractor C0 was the sole contractor operating in this phase, rendering the project’s green construction performance entirely contingent upon C0’s construction performance. In the later stages of the project, there was a marked improvement in green construction performance. With the entry of Contractors C5, C8, and C9, who demonstrated favorable green construction performance, the overall green construction performance of the project saw an upward trend. Project managers can prioritize the management of Contractor C0 in the early and mid-stages of the project, strictly monitoring C0’s green construction performance to ensure adherence to green construction standards such as energy conservation, emission reduction, and resource utilization optimization. Simultaneously, optimizing C0’s construction plan can mitigate unnecessary resource waste and enhance green construction efficiency. In the later stages of the project, collaborative work among high-performance contractors can be promoted. Strengthening collaborative management of C5, C8, and C9 ensures that their high green construction performance can be fully leveraged. Additionally, optimizing construction organization and resource allocation can maximize the advantages of high-performance contractors and elevate the overall green construction standard of the project.

4.4. Quantification of Green Construction Performance Resilience in Case Infrastructure Projects

This study further conducted a quantitative assessment of the green construction performance resilience of the case infrastructure project and calculated its resistance capacity, adaptability capacity, and recovery capacity. According to Equation (6), when α = β = γ = 1/3, the final score of the project’s green construction performance resilience is R = 0.443, with the assessment results presented in Figure 13.

In Figure 13, the resistance, adaptability, and recovery capacities of the case infrastructure project are 0.21, 0.31, and 0.81, respectively. The recovery capacity of the case infrastructure project is notably higher than its resistance capacity and adaptability capacity, indicating that after being exposed to risks that deteriorate green construction performance, the project still possesses strong recovery capabilities and can gradually return to the expected level of green construction performance. However, the project’s resistance capacity and adaptability capacity are relatively weak, suggesting its poor defensive and adaptive abilities against external disturbances, making it susceptible to sudden incidents that could lead to a decline in green construction performance. The primary reason for the inadequate resilience of green construction performance in the case infrastructure project lies in the existence of multiple internal vulnerabilities within the system. For instance, Contractor C0 exerting far greater influence than other contractors, and this structure may lead to potential “single point of failure” risks. To enhance the green construction performance resilience of the project, project managers should adopt systematic management strategies to bolster the project’s resistance to external disturbances and improve adaptability. For example, they could introduce backup contractors and further optimize task allocation, having multiple contractors jointly undertake critical construction tasks to enhance system flexibility and fault tolerance, ensuring that when one contractor’s green construction performance declines, others can quickly take over, thereby avoiding the impact of single-point failures on the entire project. The project should also further establish a robust emergency management mechanism for green construction, addressing strategies for responding to emergencies, such as extreme weather, supply chain fluctuations, and construction safety accidents, and introduce intelligent monitoring systems to track contractors’ green construction performance in real time, enabling timely adjustments to construction plans before issues arise and minimizing the impact on the project’s green construction performance.

4.5. Optimization of Green Construction Performance Resilience in Case Infrastructure Projects

To further enhance the green construction performance resilience of the case infrastructure project and assist project managers in better managing its green construction performance, optimization of the project’s green construction performance resilience was conducted following the optimization steps outlined in Section 3.4. Prior to formal optimization, relevant parameters for the genetic algorithm were set. Task parameters were configured based on the original project schedule, including the start and end times of each task, available free float, logical relationships between tasks, and initial contractor allocations. Regarding algorithm parameters, considering the limited adjustability range among tasks in the case infrastructure project, and to avoid premature convergence while maintaining solution diversity, this study set the population size to 50 and the maximum number of iterations to 300 to achieve stable and representative optimization results. Meanwhile, the cross-over rate (CR) was set to 0.8 and the mutation rate (MR) to 0.1. A higher cross-over rate aids in enhancing the population’s global search capability, while a moderate mutation rate prevents the algorithm from getting trapped in local optima. Based on the aforementioned parameter settings, the genetic algorithm optimization of the green construction performance resilience of the case infrastructure project was conducted, ultimately yielding an optimized construction scheduling plan along with its corresponding resistance capacity, adaptability capacity, and recovery capacity, as illustrated in Figure 14.

The changes in resistance capacity, adaptability capacity, and recovery capacity of the case infrastructure project before and after optimization are depicted in Figure 14. Compared to the pre-optimization state, all capacities exhibited improvement post-optimization, with the overall green construction performance resilience of the project increasing from 0.443 to 0.54, indicating a significant optimization effect. By readjusting the sequence of project tasks within the free float, without compromising the overall project duration and construction logic, the project’s adaptability to fluctuations in green construction performance was effectively enhanced, along with its resistance to external risks. The optimized plan is deemed reasonable and practically valuable, providing an important reference for the management of green construction performance in similar infrastructure projects.

5. Discussion

From the perspective of project complexity, this study systematically analyzed the green construction performance of infrastructure projects and proposed the concept of green construction performance resilience (GCPR). This concept aims to measure the adaptability and recovery capacity of green construction performance in infrastructure projects under uncertain environments. Furthermore, this study constructed a management and resilience optimization approach for green construction performance in infrastructure projects to assist stakeholders in more efficiently managing and controlling green construction performance, thereby promoting the green transformation of infrastructure project construction.

This study quantified the green construction performance resilience of infrastructure projects from three dimensions: resistance capacity, adaptability capacity, and recovery capacity. Unlike traditional methods for assessing the green construction level of projects, GCPR not only focuses on the current performance during the construction process but also emphasizes the ability to achieve long-term stability and sustainability through adjustment and optimization in complex and changing environments. Additionally, this concept highlights the proactive adjustment capability of infrastructure projects during the green construction process, i.e., whether the project can rapidly recover and continuously improve green construction performance under external shocks. This theory can provide a scientific basis for project managers to make informed decisions, helping them enhance the robustness of green construction performance in complex environments and thereby promoting the green transformation and high-quality development of infrastructure projects.

This study employed the PageRank algorithm to analyze the project network of the constructed infrastructure project, aiming to comprehensively measure the importance of various dimensional elements within the network. In previous studies, the importance of contractors was typically calculated using centrality metrics, such as weighted degree centrality (WDC), methods that have been widely applied and validated as effective [13,37]. However, traditional centrality approaches primarily rely on network relationships between contractors and are suitable for single-dimensional network analysis. In contrast, the project network constructed in this study encompasses multiple dimensions, with more complex associations between nodes. Consequently, relying solely on methods like WDC makes it difficult to accurately assess the importance of contractors and construction location. Moreover, methods such as WDC primarily focus on local network structures, whereas the PageRank algorithm can identify key elements from a global perspective. The results show that the PageRank algorithm is more accurate in measuring the importance of various dimensional elements within the network; the conclusion was further supported by the analysis results presented in Figure 7 and Figure 8. By analyzing the project network from the perspectives of contractor importance and construction location importance, project managers can more clearly identify contractors and construction location that are critical to the green construction performance of the project.

Furthermore, this study enhances the precision of measuring contractor dependency in infrastructure projects by calculating the degree of dependency between contractors through multi-dimensional indicators, thereby assisting project managers in identifying potential vulnerabilities in green construction performance. Previous studies typically quantified contractor dependency based on the number of days contractors spent concurrently performing interrelated tasks at the same construction location [13]. However, this approach solely relies on the temporal dimension and fails to comprehensively reflect the intricate nature of contractor dependencies. Further research has demonstrated that network analysis methods can more adequately depict dependency relationships, such as by considering the impact of task importance and criticality on the degree of contractor dependency [37,70]. Contractor dependencies are influenced by various factors, among which the superposition of critical time-window nodes and significant construction location may generate “risk potential energy” leading to a “risk resonance effect” that increases the likelihood of CPD and further intensifies the degree of dependency between contractors. Meanwhile, the network importance of contractors themselves also directly affects the strength of dependency relationships. Therefore, this study employs multi-dimensional elements to characterize the degree of contractor dependency, aiming to obtain more accurate measurement results of dependency relationships. The analysis results presented in Figure 8 further validate the effectiveness of this approach.

This study also correlates the project network of infrastructure projects with contractors’ green construction performance to forecast the overall project performance level. The results depicted in Figure 12 indicate that the method developed in this study can effectively demonstrate the dynamic changes in green construction performance of the project over time. Furthermore, to enhance green construction performance resilience in infrastructure projects, this study employs a genetic algorithm as a metaheuristic optimization strategy to adjust the project schedule, aiming to improve green construction performance resilience under given constraints. Metaheuristic algorithms have been proven to exhibit high effectiveness in project network optimization [13,37] and perform well in addressing complex optimization problems that are challenging for traditional mathematical optimization methods [68,71].

This study applies the proposed methodology to a real-world infrastructure project case to validate its effectiveness. Through the case study, the construction method of the project network is demonstrated, and the indicators of green construction performance resilience (GCPR) are calculated. These indicators also provide additional insights for green construction management. Figure 7 and Figure 8 clearly illustrate the importance of contractors at different stages of the project. Furthermore, Figure 10 shows the potential impact of contractor performance deterioration on the project, which can help project managers identify contractors that are critical to the project’s green construction performance. By identifying and quantifying contractor dependencies, project managers can better prevent the potential effects of cascading performance disruptions (CPD). This study proposes a specific calculation method for GCPR, assessing resilience across three dimensions. Finally, based on a genetic algorithm, the green construction performance resilience of the case project is optimized. Figure 14 presents the final optimization results, demonstrating the effectiveness of the proposed optimization approach.

6. Conclusions

To enhance green construction performance in infrastructure project development and reduce its adverse environmental impacts, this study analyzes the green construction performance of infrastructure projects from the perspective of project complexity and introduces the concept of green construction performance resilience (GCPR). This concept is used to measure the adaptability and recovery capacity of green construction performance within the complex environment of infrastructure projects, thereby providing scientific management insights for project managers. On this basis, this study constructs a series of green construction performance resilience management methods for infrastructure projects to assist project managers in formulating effective strategies, quantifying, analyzing, and proactively managing potential vulnerabilities in green construction performance as well as latent risks of performance deterioration. Based on project complexity theory, this study employs the labeled property graph technology to construct a project network and presents its dynamic evolution process as the project progresses to reveal the complexity structure of the project across different dimensions. Additionally, this study proposes a specific calculation method for green construction performance resilience in infrastructure projects and applies a genetic algorithm to adjust the project schedule, aiming to optimize the project’s green construction performance resilience and enhance its risk response capabilities. To verify the practicality of this method, it has been applied to real-world infrastructure projects. The research results indicate that this method can effectively assist project managers in identifying vulnerabilities in green construction performance within the project, quantifying the impact of contractors’ green construction performance deterioration risks on the project, accurately assessing, and improving the project’s green construction performance resilience.

From a theoretical standpoint, this study enriches research in the field of project resilience, expands the application scope of project complexity theory, and offers a novel academic perspective for research on green construction performance in infrastructure projects. From a practical perspective, this study provides project managers with a suite of systematic analysis and visualization tools to better address challenges posed by infrastructure project complexity and enhance the projects’ rapid adaptability and proactive adjustment capabilities in emergency situations. The outcomes of this study contribute to improving green construction performance resilience in infrastructure projects and provide scientific support for the green transformation and sustainable development of infrastructure projects. Future research could incorporate additional dimensions into the project network to more precisely quantify the interrelationships among different elements within projects. Meanwhile, labeled property graph could also be utilized in the future to expand the element attributes of project networks, such as the scale of construction location and resources required for tasks, to enhance the precision and applicability of analyses. These improvements will further refine the study of green construction performance resilience and provide stronger theoretical support and practical guidance for green construction management in infrastructure projects.