Fuzzy Multi-Objective Optimization Model for Resilient Supply Chain Financing Based on Blockchain and IoT

Abstract

Managing finances in a supply chain today is not as straightforward as it once was. The world is constantly shifting—markets fluctuate, risks emerge unexpectedly—and companies are continually trying to stay one step ahead. In all this, financial resilience has become more than just a strategy. It is a survival skill. In our research, we examined how newer technologies (such as blockchain and the Internet of Things) can make a difference. The idea was not to reinvent the wheel but to see if these tools could actually make financing more transparent, reduce some of the friction, and maybe even help companies breathe a little easier when it comes to liquidity. We employed two optimization methods (Non-dominated Sorting Genetic Algorithm II (NSGA-II) and Multi-Objective Particle Swarm Optimization (MOPSO)) to achieve a balanced outcome. The goal was lower financing costs, better liquidity, and stronger resilience. Blockchain did not just record transactions—it seemed to build trust. Meanwhile, the Internet of Things (IoT) provided companies with a clearer picture of what is happening in real-time, making financial outcomes a bit less of a guessing game. However, it gives financial managers a better chance at planning and not getting caught off guard when the economy takes a turn.

1. Introduction

In today’s world, managing finance in supply chains has become one of the main challenges for companies. The increasing complexity of supply chains, market fluctuations, and economic uncertainties has compelled companies to seek new and efficient solutions for financing. In the meantime, financial resilience has been proposed as one of the key components that can make companies more resilient to financial crises and supply chain disruptions [1]. This research aims to present a multi-objective optimization model for resilient supply chain financing that utilizes new technologies, such as blockchain and the Internet of Things (IoT), to reduce financing costs, optimize liquidity, and enhance flexibility in the face of market uncertainties [2].

One of the main challenges in supply chain financing is the lack of transparency in financial transactions and the credit risk of suppliers and manufacturers. This study uses blockchain as a decentralized and transparent technology to solve this challenge. By providing an immutable record system, blockchain enables the real-time verification of suppliers’ credit histories, preventing risks associated with default or late payments. In addition, IoT, as a monitoring tool, provides real-time information on the operational status of suppliers, warehouses, and transportation, which can help to assess financial risks more accurately [3]. These two technologies will help reduce uncertainty in supply chain financing and improve stakeholder trust. The model presented in this study employs two multi-objective optimization algorithms, namely NSGA-II and MOPSO, which simultaneously minimize financing costs, enhance liquidity levels, and maximize supply chain resilience. This model enables financial decision-makers to select the optimal financing strategy based on varying market conditions and risk tolerance levels. This study examines the optimal solutions for supply chain financing, considering various demand scenarios, operational risk conditions, and varying levels of blockchain trust.

In many decision-making processes in the supply chain, especially in the field of financing, variables such as “trust between parties”, “financial risk”, “supplier credit”, or “perception of digital technology qualities” are evaluated in conditions of complete uncertainty and in a verbal or subjective manner. For this reason, the use of fuzzy approaches in modeling such systems is logical and mandatory, as it allows the transformation of the subjective knowledge of decision-makers into quantitative mathematical relationships. Unlike deterministic models, fuzzy logic explicitly models uncertainty and ambiguity using membership functions and helps optimize systems where precise data is not available. In this study, fuzzification of some key parameters, including the level of trust, the level of technology adoption, and credit risk, has allowed us to simulate the real and complex behavior of suppliers more accurately.

The main innovation of this study lies in combining advanced optimization methods with digital technologies to create a smart and resilient financing model. Unlike previous studies that focused solely on reducing financing costs, this study demonstrates that liquidity, operational risk, and the ability to adapt to market changes are also key factors in financing decisions. This research can help companies, financial institutions, and policymakers manage supply chain financing more effectively, leveraging new technologies to mitigate risks and enhance financial efficiency.

The value of this study lies not merely in the combination of digital technologies and optimization models, but in how it operationalizes abstract and uncertain financial indicators—such as trust and risk using fuzzy logic and quantifies their impact on real financing decisions within supply chains. The proposed framework also fills a methodological gap by adapting advanced metaheuristics (NSGA-II and MOPSO) to this multi-dimensional decision space, enabling practical, data-driven strategies for resilient and technology-enabled financial planning.

Based on the identified research gap and practical challenges in digitally enabled supply chain finance, this study addresses the following core research questions:

• How can blockchain and IoT technologies be integrated into supply chain financing models to reduce financial risks and increase transparency?
• To what extent do fuzzy multi-objective optimization techniques enhance decision-making under uncertainty in the context of supply chain finance?
• How do trust levels (based on blockchain) and operational risks (monitored via IoT) influence financing costs, liquidity access, and supply chain resilience?
• What are the comparative strengths of NSGA-II and MOPSO algorithms in optimizing complex, digitally augmented financing decisions?

The remainder of this research is divided into several main sections. In the second section, a literature review is presented, which examines previous studies in the field of supply chain financing, the impact of new technologies such as blockchain and the Internet of Things, and multi-objective optimization models. In the third section, the research method and data used are described, along with the data collection method and the conceptual framework of the research. The fourth section is dedicated to the problem-solving process, in which two multi-objective optimization algorithms, NSGA-II and MOPSO, are introduced, and their parameter settings are examined. In the fifth section, numerical analysis and results are presented, which include examining the impact of interest rates, blockchain trust scores, IoT risk indicators, and demand changes on supply chain financing. In this section, a performance comparison of the two optimization algorithms is presented, along with an examination of the decision-making criteria. Finally, the sixth section of the paper is dedicated to conclusions and future suggestions and summarizes the findings and suggests future research directions to improve the proposed model and expand its applications.

2. Literature Review

Supply chain financing is a key issue in modern supply chain management, and it has become a strategic challenge for organizations due to its increasing complexity and economic volatility [4]. In recent years, numerous studies have examined financing models, financial risks, and optimal strategies for managing financial resources within supply chains. Studies have shown that liquidity and financing conditions have a direct impact on supply chain performance and an organization’s ability to manage disruptions effectively. In the meantime, financial resilience has been proposed as a determining factor in maintaining supply chain sustainability [5].

Past studies have mainly focused on optimizing financing costs and reducing investment costs in supply chains. Classical financing models primarily emphasize minimizing financing costs and optimizing capital structures. However, operational risks and the effects of economic uncertainties on supply chain financing have received less attention in these models. On the other hand, new supply chain finance models, emphasizing resilience, have addressed cost reduction and financial and operational risk management [6].

One of the recent developments in supply chain finance is the utilization of digital technologies to enhance transparency and mitigate financial risks. Recent studies have demonstrated that blockchain can serve as an effective tool in improving trust among suppliers, manufacturers, and financial institutions. Since blockchain is a decentralized and immutable record-keeping system, it makes all financial transactions in the supply chain transparent, reducing the likelihood of fraud, payment delays, and liquidity problems [7].

Moreover, recent works have emphasized that blockchain-based financing mechanisms can strengthen supplier–buyer relationships by reducing information asymmetry and enhancing transactional security. Research in this field has shown that suppliers’ trust scores on blockchain can directly impact their financing conditions, such that suppliers with better credit histories receive lower interest rates and obtain more financing facilities [8].

Along with blockchain, the Internet of Things (IoT) is also recognized as a key tool in supply chain finance management. Utilizing real-time data in the supply chain through IoT sensors enables companies to assess their financial status, inventory levels, and operational risks more accurately. Recent research has shown that companies that utilize IoT to monitor supply chain risks exhibit better liquidity and are less vulnerable to sudden risks. Additionally, studies highlight that IoT-driven data analytics contribute to proactive risk mitigation strategies and enhance the financial resilience of supply chains in volatile markets [9].

Therefore, combining blockchain and IoT technologies can provide a powerful solution to increase supply chain financial resilience. Several studies have examined multi-objective models in supply chain finance optimization to balance financing costs, liquidity, and financial risk. Studies have shown that multi-objective optimization algorithms, such as NSGA-II and MOPSO, can assist decision-makers in selecting an optimal solution among various financing options. Recent research has demonstrated that NSGA-II can yield Pareto-optimal solutions, allowing decision-makers to choose the most suitable strategy for their specific circumstances. In contrast, MOPSO is a good choice for financial applications that require rapid decision-making due to its faster performance and ability to converge to optimal solutions [10].

A key challenge in applying optimization methods to supply chain financing is properly tuning model parameters and considering various scenarios of market uncertainty. Studies have shown that market uncertainties, demand fluctuations, and interest rate changes can have a direct impact on the performance of optimization models. For this reason, some studies have used fuzzy and robust models to manage uncertainties. These models enable decision-makers to make informed financing decisions, even when input data are uncertain [11].

In recent years, the application of fuzzy logic in decision-making modeling under uncertainty has attracted considerable attention in the fields of finance, management, and logistics. The real conditions of supply chains are always accompanied by ambiguity, incomplete data, demand fluctuations, and uncertainty in performance indicators; issues that traditional deterministic models are unable to accurately reflect. Meanwhile, multi-objective optimization models based on fuzzy logic, as a flexible approach, enable the simultaneous analysis of conflicting goals such as cost reduction, resilience enhancement, and liquidity management, even in situations where decision variables and model parameters are not fully accurate. Some scholars have also underlined the synergy between fuzzy logic and blockchain systems in enabling transparent yet flexible financial decision-making under uncertainty [12].

Various studies have shown that fuzzy logic can be effectively used to model variables such as supplier trust, operational risk, real-time performance indicators, or even market fluctuations, especially in environments where input information is not complete, accurate, or reliable [13]. Fuzzy modeling not only recognizes uncertainty but also numerically incorporates it into the optimization process, allowing the decision maker to evaluate different solutions across a wide range of possible scenarios [14].

In the context of supply chain finance, the use of fuzzy logic has led to the creation of models that are capable of more accurate credit risk analysis, more flexible liquidity management, and a more dynamic understanding of the interactions between technological and financial factors. One of the key advantages of fuzzy models is that they can incorporate the lack of data transparency in environments such as blockchain-based trust or the fluctuations caused by real-time indicators of the Internet of Things [15]. In such a structure, financial decisions are made not only based on numerical data but also on qualitative interpretations of risk and trust.

From a technological perspective, the combination of fuzzy modeling with new technologies such as blockchain and the Internet of Things has opened up new horizons in the intelligent automation of financial decisions. Data generated by real-time sensors are usually accompanied by some degree of volatility and noise, while fuzzy algorithms can analyze this data in a meaningful way and convert it into reliable operational recommendations [16]. This has made the use of fuzzy logic in today’s digital and fast-changing environments a practical and reliable approach.

In other areas such as energy, urban logistics, crisis management, and sustainable supply chains, multi-objective fuzzy models have also been able to act as an effective tool in creating a balance between cost, time, and risk [17]. These experiences show that fuzzy logic is not only limited to mathematical modeling but can also be used as a decision-making language in complex and multidimensional environments. However, what has been neglected in many previous studies is the design of models that simultaneously combine the three main dimensions of decision-making in the financial supply chain, namely cost efficiency, liquidity flexibility, and operational resilience, using digital technologies [18].

Considering the trend of previous studies and the position of multi-objective models in the field of supply chain finance, in recent years, new approaches have been introduced in the simultaneous analysis of cost, risk and liquidity that use meta-heuristic algorithms, especially in data-driven environments. Some studies have shown that the use of genetic algorithms, particle swarm optimization and honeybee colony optimization along with fuzzy methods can be effective in the simultaneous optimization of several conflicting objectives in the supply chain [19]. These studies have focused more on how to manage financial and operational risks, reduce financing costs and improve liquidity levels in complex supply chains [20]. Also, in some studies, multi-objective modeling has been introduced as a smart solution in financial decision-making by considering market uncertainty and interest rate changes [21].

In addition to the development of optimization algorithms, studies have examined the role of new technologies in improving supply chain finance processes. In particular, researchers have paid special attention to the importance of real-time data received from IoT sensors and their ability to reduce operational risks [22]. This data, which allows real-time monitoring of product flow, inventory, and production conditions, plays a vital role in improving financial forecasting and decision-making [23]. On the other hand, the use of blockchain to increase transparency and reduce risks arising from lack of trust between partners in the supply chain has been the focus of several studies [24]. By providing a secure and immutable infrastructure for recording transactions, this technology has paved the way for the development of smart financing models [25].

Recent studies have emphasized the importance of integrating the three pillars of digital technology, fuzzy modeling, and multi-objective optimization. These integrated approaches have not only improved optimization results but also helped to increase the financial sustainability and resilience of supply chains in high-risk environments [20]. For example, models have been designed that use hybrid algorithms to simultaneously optimally manage financing costs, liquidity levels, and operational risks [26]. Also, the development of hybrid frameworks that leverage the benefits of blockchain, IoT, and optimization algorithms has been identified as an emerging direction in recent studies [27]. These studies provide a strong foundation for the development of the proposed model in this study and show that the integration of these tools can lead to significant improvements in supply chain financial efficiency. Despite these advancements, an integrated framework that systematically combines blockchain, IoT, and fuzzy-based multi-objective optimization for resilient financial decision-making remains underexplored [28].

According to this literature review, it is clear that previous studies have emphasized the importance of digital technologies, optimization models, and uncertainty management in supply chain financing. However, integrating all these factors into a comprehensive model that can simultaneously reduce financing costs, optimize liquidity, and increase resilience has received less attention. By combining blockchain and IoT technologies with multi-objective optimization models, such as NSGA-II and MOPSO, this study proposes a novel and comprehensive approach to optimizing supply chain financing, thereby enhancing financial efficiency and increasing supply chain resilience.

3. Mathematical Modeling

This section presents a fuzzy multi-objective optimization model for resilient supply chain financing that enhances transparency, security, and efficiency in the financing process by utilizing blockchain and the Internet of Things (IoT). As a distributed technology, blockchain enables the immutable recording and validation of financial transactions, while IoT reduces operational risks and optimizes financial decisions through real-time data collection.

The proposed model minimizes financing costs, improves liquidity, and increases supply chain resilience to potential shocks. Fuzzy variables are employed to model financial and operational uncertainties, and several constraints are implemented to manage credit risks, ensure economic stability, and facilitate flexibility in resource allocation.

The following will define the model’s sets, parameters, and decision variables and introduce the objective functions and constraints in detail.

Sets and Indices

$I$	Set of suppliers, indexed by i
$J$	Set of manufacturers, indexed by j
$K$	Set of financial institutions (banks, fintechs), indexed by k
$L$	Set of customers (buyers), indexed by l
$T$	Set of time periods, indexed by t
$S$	Set of possible disruption scenarios, indexed by s

Parameters

Financial Parameters

$C_{ijk}$	Interest rate charged by financial institution k for financing supplier i and manufacturer j
$R_{il}$	Revenue obtained by supplier i from customer l
$P_j$	Production cost of manufacturer j
$\overline{C_j}$	Credit limit available for manufacturer j

Blockchain and IoT-Based Risk Factors

$U_{ik}^b$	Trust level of financial institution k towards supplier i using blockchain (fuzzy variable)
$V_{jt}^I$	Real-time risk indicator of manufacturer j using IoT data (fuzzy variable)
$S_i^{fuzzy}$	Supply chain disruption risk for supplier i (fuzzy variable)
$D_j^{fuzzy}$	Demand uncertainty at manufacturer j (fuzzy variable)
$F_{ijk}$	Transaction fees incurred when financing from institution k

Cash Flow Constraints

$L_i^{min}$	Minimum liquidity requirement for supplier i
$L_j^{min}$	Minimum liquidity requirement for manufacturer j

Decision Variables

$X_{ijk}$	Amount of financing received by supplier i from financial institu-tion k
$Y_{jk}$	Amount of financing received by manufacturer j from financial institution k
$Z_j$	Production level at manufacturer j
$W_i$	Supply level at supplier i
Qil	Quantity delivered by supplier i to customer l
δik	Binary variable, 1 if supplier i receives financing from institution k, otherwise 0
θjk	Binary variable, 1 if manufacturer j receives financing from institu-tion k, otherwise 0

$$Min Z_1=\sum _{i\in I}\sum _{j\in J}\sum _{k\in K}{C}_{ijk}{X}_{ijk}+\sum _{j\in J}\sum _{k\in K}{C}_{jk}{Y}_{jk}+\sum _{i\in I}\sum _{j\in J}\sum _{k\in K}{F}_{ijk}$$

(1)

$$Max Z_2=\sum _{i\in I}\sum _{k\in K}{U}_{ik}^b{\delta }_{ik}+\sum _{j\in J}\sum _{t\in T}\sum _{k\in K}{V}_{jt}^I{\theta }_{jk}-\sum _{i\in I}{S}_i^{fuzzy}-\sum _{j\in J}{D}_j^{fuzzy}$$

(2)

$$Max Z_3=\sum _{i\in I}(R_{il}-X_{ijk})+\sum _{j\in J}(Z_j-P_j-Y_{jk})$$

(3)

S.t

$$\begin{gathered} X_{ijk}\le {\delta }_{ik}.\overline{C_j}, \forall i,j,k \\ {Y}_{jk}\le {\theta }_{jk}.\overline{C_j}, \forall j,k \end{gathered}$$

(4)

$$\begin{gathered} \sum _{l\in L}{Q}_{il}\le W_i, \forall i \\ \sum _{i\in I}{W}_i\le \sum _{j\in J}{Z}_j, \forall j \end{gathered}$$

(5)

$$z_j\le P_j+\sum _{k\in K}{Y}_{jk} \forall j$$

(6)

$$X_{ijk}\le U_{ik}^b.{\delta }_{ik}.\overline{C_j} ,\forall j,k$$

(7)

$$Y_{jk}\le V_{jt}^I.{\theta }_{jk}.\overline{C_j} \forall j,k$$

(8)

$$\sum _{i\in I}{R}_{il}\ge \sum _{i\in I}\sum _{j\in J}{X}_{ijk}+\sum _{j\in J}{Y}_{jk}, \forall k,l$$

(9)

$$X_{ijk}\ge \left(1-S_i^{fuzy}\right).X_{ijk}^{normal} ,\forall i,j,k$$

(10)

$$\alpha {\mathrm{Z}}_1+\beta {\mathrm{Z}}_2+\gamma {\mathrm{Z}}_2=1, \alpha ,\beta ,\gamma are \mathrm{W}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{C}\mathrm{o}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{s}$$

(11)

The proposed model presents a multi-objective optimization framework for resilient supply chain finance, leveraging blockchain and Internet of Things (IoT) technologies to enhance transparency, reduce financial costs, and improve resilience. The model is designed to achieve three fundamental objectives: minimize financing costs, maximize supply chain resilience, and optimize liquidity and cash flow management. To achieve these goals, a set of constraints has been implemented to ensure financial stability, mitigate risks, and enhance flexibility in resource allocation. The following section outlines and explains the objective functions and constraints embedded in the model. The first objective function (1) aims to minimize financing costs, including interest payments to financial institutions, transaction fees, and penalties incurred from payment delays. By optimizing financing strategies, this objective ensures efficient resource utilization while minimizing the financial burden on suppliers and manufacturers. Beyond cost reduction, one of the most critical aspects of the model is enhancing the supply chain’s resilience against potential disruptions. Objective function (2) seeks to maximize the supply chain’s resilience by integrating blockchain technology to foster trust among suppliers, manufacturers, and financial institutions. Through its decentralized structure, blockchain enables the tracking and verification of financial transactions, thereby reducing fraud risks and preventing payment delays. Additionally, real-time data from IoT devices plays a crucial role in assessing operational risks within the supply chain by monitoring inventory levels, production processes, and logistics status. This function ensures that the supply chain remains robust against market fluctuations, unforeseen changes, and external disruptions. Another essential component of the model is efficient cash flow and liquidity management. Objective function (3) optimizes cash flow distribution to prevent liquidity shortages or excessive capital allocation. Poor liquidity management can lead to delayed supplier payments, reduced production capacity, and overall inefficiencies in the supply chain. This function ensures that each entity in the supply chain receives the necessary liquidity to maintain operational stability without accumulating excess funds that could result in unnecessary costs. To achieve these objectives, the model incorporates a series of constraints that regulate financial, operational, and risk management aspects, ensuring the feasibility and resilience of financing decisions. Constraint (4) establishes credit limits for suppliers and manufacturers, preventing them from exceeding the maximum allowable financing from financial institutions. Constraint (5) maintains the balance between supply and demand, ensuring that the quantity of raw materials supplied aligns with production needs and customer orders, thereby preventing shortages or excessive inventory accumulation. Moreover, Constraint (6) accounts for production capacity constraints, ensuring that manufacturers do not exceed their operational capabilities or financial limits. Additionally, Constraint (7) introduces blockchain-based trust levels, allowing financing to be allocated only to suppliers with a proven track record of financial reliability. This mechanism reduces credit risks and prevents funding from being extended to high-risk entities. Furthermore, IoT-based risk assessment is integrated into Constraint (8), ensuring that only suppliers and manufacturers meeting acceptable performance criteria are eligible to receive financing. This measure mitigates risks associated with unstable or unreliable suppliers while enhancing overall supply chain resilience. Constraint (9) focuses on liquidity requirements, ensuring that all supply chain participants maintain a minimum level of cash reserves to handle financial uncertainties and prevent insolvency. Additionally, Constraint (10) is designed to address supply chain disruptions, ensuring that financial and operational strategies remain adaptable to market fluctuations and economic uncertainties. Lastly, Constraint (11) enforces a balanced approach to multi-objective optimization, guaranteeing that financing decisions simultaneously address cost minimization, resilience enhancement, and liquidity optimization. This ensures that no single objective dominates at the expense of others, thereby maintaining a well-rounded and effective financial strategy.

In this model, parameters such as blockchain-based trust level, IoT real-time risk index, and demand uncertainty are modeled as fuzzy variables. For this purpose, triangular or trapezoidal membership functions were used that grade risk or trust based on low, medium, and high values. For example, a trust level with a value of 0.2 is considered “low”, and its membership in the corresponding function gradually decreases or increases. The use of these fuzzy variables allows decision-making to be made in a continuous space of probabilities and uncertainties instead of relying on definite values.

In order to capture the inherent ambiguity and linguistic nature of certain key factors in supply chain financing, especially those related to trust, technology adoption, financial risk, and creditworthiness, this study incorporates fuzzy logic into the mathematical model. Each fuzzy parameter is modeled using a well-established membership function that reflects its qualitative interpretation in the real-world context.

Table 1. Definition of fuzzy variables and membership functions used in the model.

Fuzzy Concept	Membership Function Type	Variable Domain	Sample Linguistic Values
Supplier Trust Level	Triangular	[0, 1]	Low, Medium, High
Technology Adoption Level	Trapezoidal	[0, 10]	Poor, Acceptable, Good, Excellent
Financial Risk	Triangular	[0, 100]	Low, Medium, High
Creditworthiness Level	Trapezoidal	[0, 1]	Unsuitable, Average, Suitable, Ideal

provides an overview of the fuzzy variables used in this model, along with their membership function types, defined domains, and representative linguistic terms.

These fuzzy variables are subsequently integrated into the optimization model using fuzzy logic-based evaluation rules, allowing the model to represent subjective assessments and real-world ambiguity with greater precision and realism.

4. Solution Methods

This study employs a quantitative approach to examine the resilience of supply chain financing in the face of uncertainty. It seeks to balance financing costs, resilience, and liquidity management through the use of multi-objective optimization techniques. The research framework is designed based on realistic financial scenarios, incorporating interest rates, liquidity requirements, blockchain trust scores, IoT risk indicators, and demand variability. The primary objective of this study is to optimize financial decisions by striking a balance between cost efficiency and economic stability.

To ensure the accuracy and validity of the research results, the dataset is constructed as a combination of synthetic financial data and industry-aligned economic assumptions. The synthetic data is generated based on realistic financial constraints and observed behavioral patterns in supply chain financing, representing fundamental financial dynamics. All variables used in the model are derived from empirical studies in supply chain risk management, financial optimization, and multi-objective decision-making.

The present study examines various scenarios of demand changes, ranging from low-demand conditions to severe market fluctuations, to investigate the impact of demand uncertainty on financial stability. Additionally, blockchain trust scores and IoT risk indicators are also incorporated to evaluate the effect of digital transformation on contemporary financial decision-making. These elements provide a comprehensive view of the financial decision-making process under uncertainty, making the model reliable for various economic applications.

This study’s optimization framework uses two multi-objective optimization algorithms, NSGA-II and MOPSO, which are selected to examine the trade-off between financing costs, liquidity access, and supply chain resilience. These methods enable a systematic and goal-oriented evaluation of financial strategies, rendering the research results suitable for scientific analysis and informed financial planning.

The choice of these two algorithms is based on their distinct optimization philosophies: one rooted in evolutionary mechanics (NSGA-II) and the other in swarm-based heuristics (MOPSO). Comparing their performance allows for a better understanding of how different optimization strategies perform under uncertainty-driven financial decision-making scenarios in supply chain finance.

Solution Method and Optimization Algorithms

This study uses two multi-objective metaheuristic algorithms, NSGA-II and MOPSO (Multi-Objective Particle Swarm Optimization), to achieve an optimal balance between cost minimization, financial resilience enhancement, and liquidity optimization. Each method is selected to investigate different financing approaches and risk mitigation strategies [29].

NSGA-II is a genetic evolution-based algorithm that efficiently identifies Pareto-optimal solutions. This method employs non-dominated sorting to categorize solutions based on dominance relations, ensuring that the most effective financial strategies are preserved. The primary advantage of this method is its ability to handle complex financial constraints and nonlinear decision variables, which provides decision-makers with various financial options [30].

MOPSO, a multi-objective version of the particle swarm optimization (PSO) algorithm, adapts the principles of crowd intelligence to solve multi-objective problems. Unlike NSGA-II, which employs genetic operators to evolve solutions, MOPSO utilizes a population of particles that move through the solution space, improving their position based on previous best solutions and swarm dynamics. The main advantage of this method is its high computational efficiency, as it typically converges faster than NSGA-II while maintaining a high level of solution quality. This feature makes MOPSO ideal for real-time financial decision-making and rapid optimization of financing strategies. To ensure the efficiency and accuracy of the optimization, the parameters of NSGA-II and MOPSO are carefully tuned. The population size, mutation rate, crossover probability, and iteration limit are adjusted to balance exploration (finding diverse solutions) and exploitation (improving quality solutions).

To better illustrate the optimization mechanisms used in this study, the main procedural steps of both the NSGA-II and MOPSO algorithms are summarized in Figure 1.

As shown in Figure 1, although both algorithms share evolutionary principles, they differ in aspects such as selection strategy and diversity maintenance, which influence their performance in multi-objective supply chain optimization.

NSGA-II has a higher ability to maintain the diversity of solutions due to its more precise sorting structure, while MOPSO performs favorably in faster convergence of local optima. These two algorithms were chosen to compare the performance of genetic-based approaches and particle swarm algorithms in the fuzzy and blockchain-based modeling space, to show how differences in search logic can affect the quality of decision-making in supply chain financing. The exact parameter settings for this study are presented in

Table 2. NSGA-II and MOPSO parameter settings.

Parameter	NSGA-II Setting	MOPSO Setting
Population Size	100	100
Number of Generations	500	500
Crossover Probability	0.9	N/A
Mutation Rate	0.1	N/A
Swarm Inertia Weight	N/A	0.7
Personal Best Factor	N/A	1.5
Global Best Factor	N/A	2

, which ensures that the optimization results are reproducible and valid. All algorithms and fuzzy modeling procedures were implemented using MATLAB R2023a. Custom scripts were developed to ensure flexible manipulation of fuzzy rules, optimization parameters, and evaluation of blockchain-based decision metrics.

The parameters of NSGA-II and MOPSO algorithms are selected according to previous studies in the field of supply chain and multi-objective optimization. For example, a population size of 100 and an iteration number of 500 have been used in the parametric optimization literature and provide a good balance between accuracy and convergence speed. The mutation rate of 0.1 for NSGA-II and the particle weight coefficients for MOPSO are also adjusted based on experimental results.

In order to increase the methodological accuracy and examine the stability of the model to changes in algorithmic parameters, a sensitivity analysis was conducted on key NSGA-II and MOPSO settings. In this analysis, the population size was varied between 50 and 200, the mutation rate was varied between 0.05 and 0.2, and the number of iterations was varied between 300 and 700. The effects of these changes on the algorithm execution time, Pareto front diversity, and resilience score stability were then examined. The results showed that although increasing the population size improved the diversity of solutions, it significantly increased the algorithm execution time. A high mutation rate made the model convergence unstable and caused fluctuations in resilience scores, while a mutation rate lower than 0.05 also led to a decrease in solution diversity. The initial settings used in the paper (population 100, mutation rate 0.1, iterations 500) provided the best balance between solution quality, diversity, and computational speed. Therefore, these settings were chosen as the empirical optimum for the final model.

5. Analysis of Results

To obtain the final results, the following methodological steps were applied. First, synthetic input data were prepared, and fuzzy logic was used to convert linguistic assessments (e.g., trust, risk, and cost evaluations) into triangular fuzzy numbers. Next, these fuzzy values were processed using the proposed optimization framework. Two evolutionary algorithms (NSGA-II and MOPSO) were separately applied to solve the fuzzy multi-objective model, with population size, crossover, and mutation rates predefined based on a sensitivity analysis. During each iteration, supplier performance was evaluated using blockchain-integrated metrics including transaction trust and financing eligibility. Pareto-optimal solutions were then extracted and compared based on objective trade-offs. The final decisions were made by defuzzifying the optimal solutions and selecting the most balanced alternatives according to the supply chain’s priorities.

To illustrate the effectiveness of our fuzzy robust optimization model in supply chain finance, we construct an example dataset inspired by real-world scenarios. This dataset includes key decision variables, cost parameters, demand variations, and resilience factors influencing supply chain financing decisions. Analyzing

Table 3. Example dataset for fuzzy robust supply chain finance model.

Supplier	Manufacturer	Financial Institution	Interest Rate (%)	Trust Score (Blockchain)	IoT Risk Indicator	Demand Variation (%)	Liquidity Requirement (USD)
S1	M1	FI1	4.5	High	Low	10	50,000
S2	M2	FI2	5.2	Medium	Medium	15	40,000
S3	M1	FI3	3.9	High	Low	8	60,000
S4	M3	FI2	6.1	Low	High	20	35,000
S5	M2	FI1	5.7	Medium	Medium	12	45,000

, we can assess how different factors impact financial strategies and operational resilience.

Table 2 is a foundation for analyzing financial risk and supply chain resilience. It presents a structured view of financial interactions among suppliers, manufacturers, and financial institutions, incorporating blockchain-based trust scores and IoT-driven risk assessments. These elements allow for a data-driven approach to optimizing financial strategies and ensuring liquidity stability. For example, a supplier with a high blockchain trust score may receive more favorable financing terms. In contrast, a supplier flagged with a high IoT risk indicator may face higher interest rates or stricter liquidity requirements. By incorporating these variables into our fuzzy robust optimization model, we can simulate real-world supply chain dynamics and assess the financial sustainability of various stakeholders. This example dataset helps us better understand how different financial conditions influence the stability of supply chains under uncertain market conditions.

Due to the lack of access to a complete industrial dataset with real-time blockchain and IoT inputs, we constructed a synthetic dataset based on realistic financial patterns and prior studies. While this limits direct generalizability, it enables the development and benchmarking of the optimization framework under controlled yet representative conditions. Future research will incorporate real-world data from industry partners to validate and expand the current model.

The sensitivity analysis on interest rates and liquidity needs examines how variations in interest rates affect the liquidity requirements of supply chain participants. In supply chain financing, higher interest rates typically result in increased borrowing costs, thereby reducing the availability of liquidity. Conversely, lower interest rates improve liquidity conditions, enabling businesses to manage their cash flow more efficiently.

We applied two multi-objective optimization algorithms (NSGA-II and MOPSO) to analyze this relationship to generate optimal financing solutions under different interest rate conditions. The results are summarized in

Table 4. NSGA-II sensitivity analysis on interest rates and liquidity needs.

Interest Rate (%)	Liquidity Requirement (USD)	Financing Cost	Resilience Score
3.5	30,000	50,000	0.4
3.7	32,105.26	52,105.26	0.432
3.8	34,210.53	54,210.53	0.463
4.0	36,315.79	56,315.79	0.495
4.1	38,421.05	58,421.05	0.526
4.3	40,526.32	60,526.32	0.558
4.4	42,631.58	62,631.58	0.589
4.6	44,736.84	64,736.84	0.621
4.8	46,842.11	66,842.11	0.653
4.9	48,947.37	68,947.37	0.684
5.1	51,052.63	71,052.63	0.716
5.2	53,157.89	73,157.89	0.747
5.4	55,263.16	75,263.16	0.779
5.6	57,368.42	77,368.42	0.811
5.7	59,473.68	79,473.68	0.842
5.9	61,578.95	81,578.95	0.874
6.0	63,684.21	83,684.21	0.905
6.2	65,789.47	85,789.47	0.937
6.3	67,894.74	87,894.74	0.968
6.5	70,000	90,000	1

(NSGA-II) and

Table 5. MOPSO sensitivity analysis on interest rates and liquidity needs.

Interest Rate (%)	Liquidity Requirement (USD)	Financing Cost	Resilience Score
3.5	32,000	52,000	0.45
3.6	33,894.74	53,894.74	0.476
3.8	35,789.47	55,789.47	0.503
3.9	37,684.21	57,684.21	0.529
4.1	39,578.95	59,578.95	0.555
4.2	41,473.68	61,473.68	0.582
4.4	43,368.42	63,368.42	0.608
4.6	45,263.16	65,263.16	0.634
4.7	47,157.89	67,157.89	0.661
4.9	49,052.63	69,052.63	0.687
5.0	50,947.37	70,947.37	0.713
5.2	52,842.11	72,842.11	0.739
5.3	54,736.84	74,736.84	0.766
5.5	56,631.58	76,631.58	0.792
5.7	58,526.32	78,526.32	0.818
5.8	60,421.05	80,421.05	0.845
6.0	62,315.79	82,315.79	0.871
6.1	64,210.53	84,210.53	0.897
6.3	66,105.26	86,105.26	0.924
6.5	68,000	88,000	0.95

(MOPSO), which present the variations in interest rates, liquidity requirements, financing costs, and resilience scores.

A comparison of these two tables reveals key differences in how each algorithm prioritizes liquidity and resilience. NSGA-II produces solutions that cover a wider range of liquidity conditions, making it more suitable for scenarios requiring greater financial flexibility. On the other hand, MOPSO tends to stabilize liquidity requirements, ensuring that liquidity remains within a narrower, more controlled range even as interest rates fluctuate. This suggests that MOPSO may be a better choice for risk-averse organizations prioritizing financial stability over cost savings.

The impact of these differences is further illustrated in Figure 2, which presents a line chart comparing the liquidity requirements under NSGA-II and MOPSO. As interest rates increase, the liquidity requirements rise for both models, but the trend under MOPSO remains more stable compared to NSGA-II. This aligns with the expectation that MOPSO solutions prioritize financial robustness, whereas NSGA-II enables more dynamic adjustments to financing strategies.

This analysis confirms that interest rate fluctuations have a direct impact on liquidity management in supply chain finance. While lower interest rates enhance financial flexibility, decision-makers must carefully evaluate their financing strategy in light of their risk tolerance and liquidity needs. Firms that operate in volatile markets may benefit from NSGA-II’s flexibility, while those seeking stable long-term liquidity management may prefer the risk-controlled approach of MOPSO.

The blockchain-based trust score and financing conditions analysis explores how trust levels in blockchain-based transactions impact financial terms in supply chain financing. A higher blockchain trust score indicates a more reliable financial history, resulting in lower financing costs and higher loan approval rates. This relationship is crucial for supply chain participants as financial institutions use blockchain data to assess risk and determine lending conditions.

The results of this analysis are presented in

Table 6. Blockchain trust score and financing conditions.

Supplier	Blockchain Trust Score	Financing Cost (USD)	Loan Approval Rate (%)
S1	0.2	100,000	50
S2	0.29	94,444.44	55.56
S3	0.38	88,888.89	61.11
S4	0.47	83,333.33	66.67
S5	0.56	77,777.78	72.22
S6	0.64	72,222.22	77.78
S7	0.73	66,666.67	83.33
S8	0.82	61,111.11	88.89
S9	0.91	55,555.56	94.44
S10	1	50,000	100

, which summarizes blockchain trust scores, financing costs, and loan approval rates for a set of suppliers. The data clearly shows that financing costs decrease as the blockchain trust score increases, while loan approval rates rise. This trend highlights the importance of financial transparency and trust in securing more favorable economic terms.

Figure 3 provides a dual-axis line chart to visualize these relationships, illustrating how blockchain trust scores influence financing costs (blue line) and loan approval rates (red line). The downward slope of the blue line confirms that suppliers with higher trust scores receive financing at a lower cost, while the upward trend of the red line highlights increased loan approval rates for trusted suppliers. These findings underscore the strategic importance of leveraging blockchain technology to enhance financial credibility and secure improved access to capital.

While working on this part of the study, we started noticing how real-time data from IoT systems could actually shape financial thinking, especially in supply chain environments. These systems, by constantly tracking various performance signals, offer insights that can help lenders or financial teams get a better grasp of how risky or stable a supplier might be.

As we looked closer, one pattern began to emerge. When the data showed higher IoT-related risks—things like system disruptions or inconsistent reporting—those suppliers tended to struggle more with liquidity. It was not just about operations; it had real financial consequences.

We gathered these observations in

Table 7. IoT Risk indicators and supply chain vulnerability.

Supplier	IoT Risk Indicator	Liquidity Impact (USD)
S1	0.1	80,000
S2	0.189	74,444.44
S3	0.278	68,888.89
S4	0.367	63,333.33
S5	0.456	57,777.78
S6	0.544	52,222.22
S7	0.633	46,666.67
S8	0.722	41,111.11
S9	0.811	35,555.56
S10	0.9	30,000

, where different suppliers were assessed based on the level of IoT-related risk they carried and how that seemed to affect their access to liquidity. Generally, the higher the risk score, the tighter the financial conditions became. It appears that lenders tend to see these suppliers as riskier and respond by limiting their access to funds. It is a reminder of how operational stability, or the lack of it, can influence financing decisions in very practical ways.

Figure 4 presents a line chart mapping IoT risk indicators against liquidity impact to further illustrate this relationship. The downward slope confirms that liquidity availability decreases as IoT risk levels increase. This visualization emphasizes the importance of operational stability in securing financial resources. Firms with lower IoT risk indicators tend to maintain more substantial liquidity positions, enabling them to negotiate more favorable financing terms.

This analysis highlights the importance of adopting smart strategies to manage risk, especially using Internet of Things (IoT) technologies. With tools like real-time monitoring, predictive analytics, and automated assessments, companies can respond more quickly to potential disruptions, reduce their exposure to operational risks, and improve their overall financial standing within the supply chain.

In another part of the study, we examined how changes in customer demand can affect financial planning. When market demand rises or falls unexpectedly, it puts pressure on liquidity and financing, making it harder for businesses to maintain stability. That is why having a solid, flexible financial plan in place is so important—it helps organizations deal with uncertainty and stay resilient even when the market is unpredictable.

The findings are summarized in

Table 8. Demand variability and resilient financial planning.

Scenario	Liquidity Requirement (USD)	Financing Cost (USD)	Resilience Score
Low Demand	40,000	55,000	0.8
Moderate Demand	50,000	65,000	0.7
High Demand	60,000	80,000	0.6
Extreme Demand	75,000	95,000	0.5

, which outlines four demand scenarios: Low, Moderate, High, and Extreme. As expected, when demand grows, so do liquidity needs because more working capital is required to keep up with production and delivery. Financing costs also go up since more capital is needed to sustain operations. Interestingly, under extreme demand, resilience tends to drop. This suggests that when companies are pushed to their limits, they become more financially vulnerable and face significant operational risks.

Our findings point to the importance of having flexible financial strategies when it comes to managing supply chains. Companies that can read the signs of changing demand and adjust their financial planning along the way are usually in a better position to remain steady, even when things get unpredictable. Tools like adjustable credit lines, financing plans that take risk into account, and investment choices based on real demand trends can all help reduce financial pressure and make better use of available resources.

As part of this research, we also explored how financial decisions in supply chains become more complex when several priorities need to be addressed at once. It is not just about saving money—staying financially secure and keeping enough liquidity for everyday operations are just as important. To better understand how companies might manage these competing goals, we developed a series of optimized financial plans, each designed to match different risk preferences and business needs.

Table 9. Multi-criteria decision-making (MCDM) optimization results.

Solution	Financing Cost (USD)	Resilience Score	Liquidity Level (USD)
Solution 1	68,727.01	0.5103	60,592.64
Solution 2	97,535.72	0.9850	36,974.69
Solution 3	86,599.70	0.9162	44,607.23
Solution 4	79,932.92	0.6062	48,318.09
Solution 5	57,800.93	0.5909	52,803.50
Solution 6	57,799.73	0.5917	69,258.80
Solution 7	52,904.18	0.6521	39,983.69
Solution 8	93,308.81	0.7624	55,711.72
Solution 9	80,055.75	0.7160	59,620.73
Solution 10	85,403.63	0.6456	32,322.52

gives a snapshot of what we found. Each row in the table shows a different financing option, aimed at balancing three main concerns: cost, resilience, and liquidity. What is interesting is the diversity in the results—some strategies lean toward minimizing expenses, while others give more weight to financial strength or liquidity reserves. These differences show that there is no universal answer. Instead, businesses need to build financial strategies that fit their unique situation and risk profile.

A crucial aspect of multi-objective optimization is the computational efficiency of various algorithms. Figure 5 compares the computational time for NSGA-II and MOPSO across multiple iterations.

The results indicate that MOPSO generally achieves slightly faster computation times, suggesting that it may be more efficient when rapid financial decision-making is crucial. However, NSGA-II tends to explore a more expansive solution space, which can be beneficial when decision-makers require diverse financing strategies.

While the current model treats trust scores and IoT risk indicators as static inputs, it is possible to conceptualize a dynamic feedback mechanism in which suppliers can improve their profiles over time. For example, a mid-ranked supplier such as S5 (currently categorized with a medium trust score and moderate IoT reliability) could progressively invest in IoT infrastructure to improve operational transparency. This, coupled with consistent financial behavior (e.g., on-time delivery, no payment disputes), would gradually increase their fuzzy trust level. As a result, their eligibility for blockchain-based financing would improve in subsequent decision cycles, potentially reducing their transaction fees and interest rates. Incorporating such an adaptive mechanism in future model iterations would enable a more realistic depiction of the strategic evolution of supplier behavior under digital finance ecosystems.

It is important to note that in the current model, trust scores are treated as scenario-driven inputs reflecting predefined reputational states of suppliers. These scores are not dynamically derived from actual blockchain interactions or longitudinal transaction behavior. As such, while the model shows that higher trust leads to more favorable financial terms, it does not capture the endogenous trust-building process facilitated by blockchain protocols. Future work should consider implementing a dynamic trust scoring mechanism based on transaction frequency, delay penalties, contract compliance, or smart contract execution history to more accurately reflect how trust evolves within blockchain environments.

6. Discussion

The results of the fuzzy multi-objective optimization model in this study provide significant evidence on the impact of new digital technologies, especially blockchain and trust assessment systems, on supply chain finance management. The analyses conducted show that adopting fuzzy approaches in modeling uncertainties provides a more accurate understanding of the decision-making space, especially in situations where indicators such as trust level, credit risk, and technology utilization are assessed qualitatively and imprecisely. This is especially evident in matching supplier credit with the conditions for using blockchain-based services, where the fuzzy mechanism has enabled more realistic categorization and scoring.

Also, comparing the performance of NSGA-II and MOPSO algorithms [28,29] in this study shows that metaheuristic optimization approaches still have a high power in discovering quality Pareto fronts, and the choice of algorithm depends on the type of problem and the priorities of the decision makers. In problems where the diversity of solutions and uniform coverage of the Pareto front are more important, NSGA-II shows its superiority, while the MOPSO algorithm has more advantages in convergence time and simplicity of implementation. These findings are in relative agreement with previous studies, but what distinguishes this research is the structured combination of these algorithms with fuzzy variables and blockchain-based decision-making mechanisms.

Although the dataset used in this study is synthetic, it was carefully designed to reflect realistic financial parameters, risk levels, and behavioral patterns documented in previous research on supply chain finance and risk management. The generation of artificial data was based on typical ranges of interest rates, trust scores, IoT risk indicators, and liquidity requirements commonly reported in the literature. The artificial dataset used in this study was designed to reflect commonly observed patterns in supply chain finance, including typical ranges of interest rates, supplier trust levels, IoT risk indicators, and demand variations. These ranges were selected based on standard financial practices and frequently reported industry benchmarks, aiming to simulate realistic financial decision-making environments for model evaluation purposes. However, we acknowledge that the absence of actual industrial data is a significant limitation that may affect the external validity and generalizability of our findings.

The analysis of the results related to different suppliers, especially S5, shows that the use of an adaptive approach in trust scoring allows for the creation of a path to improve the conditions of suppliers over time. Although the presented model is static in nature, its results pave the way for the formulation of policies that, in a prospective time frame, lead to the improvement of the competence level of medium-sized suppliers through investment in technology and performance improvement.

Finally, this study has taken a practical step towards integrating fuzzy modeling, multi-objective optimization, and blockchain technology into an integrated framework for sustainable financing decision-making in supply chains. Despite certain limitations such as the use of synthetic and simulated data, the designed structure has significant potential for implementation in real industrial environments and provides a suitable platform for the development of adaptive learning models in the future.

7. Conclusions

In this study, a comprehensive framework for multi-objective fuzzy modeling and optimization in the field of resilient supply chain financing was presented, which, by simultaneously using the capabilities of blockchain technology and the Internet of Things, attempted to respond to the challenges related to uncertainty, financial inefficiency, and operational risks. By utilizing fuzzy variables for indicators such as supplier trust level, instantaneous risk index, and demand fluctuations, the proposed model was able to provide a flexible and adaptable space for financial decision-making in unstable conditions.

Two meta-heuristic algorithms, NSGA-II and MOPSO, were used to solve the model in multi-objective conditions, and their performance was evaluated in the face of various financial and operational scenarios. The modeling findings showed that higher levels of blockchain-based trust directly reduce financing costs and improve loan approval rates. Also, the risk indicators identified by IoT had a negative impact on the amount of allocated liquidity, so that suppliers with higher operational risk had limited access to financial resources. This shows that digital smart technologies play an important role in shaping financing conditions.

The results of the analysis of different demand scenarios showed that a sudden increase in demand increases the need for liquidity and reduces the resilience index. This doubles the importance of adaptive financial planning for organizations operating in volatile environments. In this regard, the use of multi-objective algorithms made it possible to strike a proper balance between the three key objectives of reducing costs, increasing liquidity, and maintaining resilience.

From the perspective of algorithm comparison, the results showed that the NSGA-II algorithm performed better in generating a more diverse Pareto front and provided wider decision-making options. In contrast, MOPSO, with its shorter execution time and faster convergence, is considered a suitable option for real-time and time-constrained decision-making environments. This difference in the nature of the algorithms makes the choice between them dependent on the level of risk and the organization’s needs at the moment of decision-making. These findings suggest that the choice between NSGA-II and MOPSO depends largely on the specific requirements of the decision context whether prioritizing solution diversity and strategic flexibility with NSGA-II or seeking faster convergence and computational efficiency with MOPSO.

The model proposed in this study was able to provide an innovative model for intelligent financial decision-making in supply chains by integrating fuzzy structure, meta-heuristic algorithms, and decentralized digital technologies. This model is not only capable of being implemented in simulated environments but will also pave the way for use in real operational systems with real-time data.

It is suggested that in the continuation of this research, real industrial data in industries with complex supply chains be used to increase the empirical validity of the model. Also, developing the model using machine learning algorithms and reinforcement approaches can help improve the estimation of fuzzy variables and the selection of optimal strategies in real time. Furthermore, in the future, the economic value of the model’s proposed decisions can be enhanced by adding bioeconomic modules and cost-benefit analysis throughout the supply chain life cycle.

The observed outcomes, such as the inverse relationship between blockchain trust scores and financing costs, are aligned with findings in prior studies on decentralized trust-based finance. Similarly, the trend showing that higher IoT-based risk indicators lead to tighter liquidity access is consistent with risk-sensitive lending models in digital environments. While the optimization outputs are not statistically tested, their consistency with theoretical expectations and existing research strengthens the reliability of the findings.

This study presents a static multi-objective optimization model that evaluates financing decisions based on predefined fuzzy trust scores and IoT-based risk indicators. The current formulation does not simulate the temporal evolution of these indicators, nor does it capture the dynamic feedback loops typically associated with blockchain-based reputation systems. As such, while the model demonstrates the impact of varying trust levels on financial outcomes, it does not explicitly show how blockchain mechanisms generate or erode trust over time. Consequently, statements implying that “blockchain builds trust” should be interpreted in a limited sense, i.e., the model assumes trust levels as inputs rather than deriving them through behavioral transaction histories.

Additionally, the model assumes complete information regarding supplier profiles and does not account for learning, adaptation, or behavioral responses from agents. Future research should consider integrating longitudinal transaction data, smart contract dynamics, and adaptive trust-scoring mechanisms to more accurately reflect the evolving nature of trust in digitally enabled supply chain finance.

It is important to emphasize that the results presented in this study are based on a synthetic dataset calibrated to align with realistic industry patterns. Therefore, while the model demonstrates theoretical validity, its empirical validation requires testing with real-world industrial data. This remains a key limitation of our study and highlights a critical direction for future research to enhance the robustness and applicability of the proposed model.