A Model Assembly Approach of Planning Urban–Rural Transportation Network: A Case Study of Jiangxia District, Wuhan, China

Abstract

Planning transportation networks between urban and rural areas is of crucial importance for the integration of urban and rural development, for socio-economic connectivity, and for sustainable growth. The study offers a model assembly approach in order to logically plan an integrated urban–rural transportation network that may support the coordinated development of its living–production–ecological space. Within this approach, the ordinary least squares (OLS) linear regression analysis method is used to investigate the correlation between urban and rural areas of a transportation network and the influencing factors in the living–production–ecological space so as to objectively analyze their degree of influence. These factors are size of town, urban and rural settlements, life services, supporting transportation facilities, trunk layout, external transport links, cargo hubs, logistics and transportation, enterprise distribution, agricultural production, terrain, distribution of water systems, tourism resources, heritage preservation, and ecological protection. The analytic hierarchy method is used to assign weight to the urban and rural transportation network planning impact index system. As a result, a transportation network planning decision hierarchy model is implemented to identify suitable areas for urban and rural transportation network construction and to provide guidance and reference for planning. Jiangxia District, Wuhan, China is selected as the study area to verify the feasibility and effectiveness of the model. The findings indicate that the influencing factors of urban and rural industrial and ecological space have a significant impact on the transportation network in the research area. Planning should prioritize optimizing the central region’s transportation network structure and enhancing traffic flow between urban and rural communities, which is effectively in line with the current reality. The suggested approach is helpful in establishing case-study-specific planning and development strategies of urban and rural integrated transportation networks in the age of big data, as well as in balancing these influencing factors in living, production, and ecological spaces when planning an integrated urban and rural transportation network.

1. Introduction

People have frequently blurred the geographical and social barriers that generally divide urban from rural areas during the spread of urbanization, particularly in the early period of urban sprawl [1]. The first argument made by Howard’s pastoral city theory was that there was a need for urban planning to incorporate both urban and rural growth [2]. Recognition of the spatial interdependence of “rural” and “urban” areas is starting to happen. In some locations, cities start to build interactions between urban and rural development in a variety of social developments and economic industries [3]. Urban areas encourage rural areas’ efforts to develop, and vice versa, leading to integration of urban and rural development. Throughout the integrated development stage of countries around the world, urban–rural integrated development and transportation networks should be closely linked, and the two areas influence and restrict each other [4].

The German government attaches importance to the layout of roads in villages and the rational planning of external transportation, which promotes the development of villages with their own characteristics and development potential. Finland makes full use of urban and rural resources in regional spatial planning, focusing on integrating key aspects such as transport, housing, and food production [5]. The United States proposed increasing the efficiency of transportation systems in The Civil Development Act of 1964 [6], and transportation construction had a direct and significant effect on sustainable social development [7]. In Asia, Japan’s “Economic and Social Development Plan” focuses on the rational planning of transportation infrastructure, which has enabled many large cities to drive the development of large areas around them, balancing urban and rural development. The main methods of the “New Village Movement” in Korea involve improving rural roads and effectively promoting the integrated development of urban and rural areas. Although these countries have different national conditions, they attach importance to urban and rural transportation network planning in the common development of urban and rural areas. However, recent research tends to focus more on urban transportation networks and often ignore several aspects of the transportation networks in suburban and rural areas [8]. China has a large number of rural areas, and roads are the most important or even the only mode of traffic in some areas. Therefore, urban and rural transportation planning is an urgent issue for urban and rural integration.

In addition, in the context of China’s ongoing territorial spatial planning, the coordination between transportation and other elements should be fully considered in traffic planning, and the transportation network planning should closely work with the living–production–ecological life space to extend the traffic planning content. The living–production–ecological space is an important strategy for the development and optimization of land space in the new era of China. Living space is a space used by people for daily activities. Production space is a specific functional area where people engage in productive activities, such as industrial production, agricultural production, and other activities. Ecological space is a regional space with an ecological protection function that can provide ecological products and services. The problems faced in the planning and development of urban and rural transportation networks are different in China, so the selection of urban and rural transportation network planning methods is also different. The suburban and rural areas have a wide range of complex land use functions, including urban function transfer to the outer metropolitan area, and the ecological agricultural area mainly dominated by agriculture and tourism. Therefore, there are some differences in the demand for transportation services in different parts of the far city, which should be distinguished in the comprehensive transportation planning. To address this issue, this study introduces a model assembly approach for planning an urban–rural transportation network. This approach can consider the social, economic, and ecological influencing factors of urban and rural areas from the perspective of living–production–ecological space. The study area covering urban, suburban, and rural areas was chosen to demonstrate the feasibility of the proposed approach.

The contributions of this paper are summarized as follows: (1) A model assembly approach for planning an urban–rural transportation network is proposed. This model can comprehensively analyze the correlation between urban–rural transportation network planning and the influencing factors of living–production–ecological space, and can analyze the aspects that have made outstanding contributions to the construction of urban and rural regional transportation networks. (2) A transportation network planning decision hierarchy model is constructed by combining a qualitative and quantitative analytic hierarchy, which considers 15 influencing factors in living, production, and ecological spaces. It can reduce the subjective arbitrariness of qualitative and quantitative analysis of influencing factors in transportation network planning. (3) Some planning and development strategies of the urban and rural integrated transportation network is proposed according to the modeled results of urban and rural for transportation network planning in the study area.

The remainder of this paper is organized as follows. The existing literature is reviewed in Section 2. The study area, dataset used, and study process are described in Section 3 and Section 4. Section 5 traduces the regression analysis results, model assembly results, and some discussion. Finally, Section 6 summarizes the conclusions of the study and discusses future research.

2. Literature Review

Scholars have also used these information data and various analytical models to conduct research and planning practice at all phases of transportation network design in order to scientifically plan an urban and rural transportation network.

In the evaluation of the current situation of the transportation network, various mathematical analysis methods can be used to establish a comprehensive transportation network evaluation model [9,10], including the density, scale, structure, and other aspects of the transportation network, and judge the deficiencies of the transportation network layout according to the evaluation results. In traffic analysis, multi-source data such as license plate recognition (LPR) data, cellular signal (CS) data, AVI data, IC card data, taxi GPS data, and other multi-source data can be used to extract traffic origin and destination (OD) information [11,12], analyze travel demand, and provide a basis for urban transportation planning. In transportation network planning, the transportation network, travel demand and land use are considered comprehensively, and geoinformatics technology is used to depict and determine the new linear transportation layout [13,14]. The application of multi-source data for transportation network planning mainly concentrated on the evaluation of the current transportation network and traffic demand analysis [15,16]. These studies demonstrate the values of multi-source big data in finding and solving the issue of urban–rural transportation networks.

The construction of transportation not only considers the needs of people but also needs to combine other elements of space. Chaishushan et al. constructed a basic network according to the location of large passenger flow and main passenger flow corridors [17] and studied the matching degree of network with travel demand and land use. Surabuddin Mondal et al. used geoengineering parameters, such as slope, aspect, geology, land use, drainage, and soil, and geoinformatics techniques to delineate and determine new routes between two cities [18]. This was based on an understanding that ecological protection of land space and intensive land use need to pay attention to the synergy between transportation elements and other spatial elements. Therefore, this study provides a thorough investigation of transportation networks of suburban and rural areas, coordinates land use and transportation network planning, and protects ecological, agricultural, cultural, and other resources.

However, considering the social, economic, and ecological influencing factors, the degree of importance of these factors were calculated mainly by subjective judgment [19], which should result in strong subjective arbitrariness of qualitative and quantitative analysis, especially in the era of big data. The analytic hierarchy method is a decision-making method proposed by Saaty et al. [20] for subjective analysis and objective evaluation. It effectively converts some qualitative and partial quantitative problems into quantitative problems, hierarchizes various elements, compares various related elements layer by layer, and determines the weight coefficient of comprehensive evaluation. It is widely used in decision-making work. In many planning works, analytic hierarchy approaches are used to calculate the weights of various indicators in the index evaluation system to calculate the comprehensive score. For example, the main influencing factors of residents’ travel in transportation research are summarized and the weight of the travel impact factor in the travel prediction stage was established by the analytic hierarchy method [21]. Their works could be extended to modify the location influence coefficient in the population, land use, and urban travel models to characterize the dynamic state of residents’ travel demand in the travel stage. In addition, when determining the weights of various factors, most of the studies generally use the proportional scale of the relative importance between the two factors by experts to score, and the weight is calculated in the construction of the judgment matrix, and qualitative analysis still accounts for a large proportion in the weight judgment. Therefore, the qualitative analysis still accounts for a large proportion of the weight calculation in the construction of the judgment matrix for subjective analysis.

In order to reduce the proportion of subjective analysis, analytic hierarchy methods can be combined with other techniques to rank the importance of factors according to other methods [22], thereby reducing the subjective arbitrariness of analytic hierarchy and expanding the application scope of analytic hierarchy methods. The OLS model, the least squares linear regression model, was proposed by Giusep Piazzi and later invented by Legendre or Gauss [23,24,25], which is one of the commonly used spatial regression methods. It is widely used in the field of economics to study the correlation analysis of the influencing factors of various phenomena [26]. The OLS models are also combined with mathematical statistical methods. Bothe et al. combined descriptive statistics with OLS models to analyze the factors associated with intra-city job growth in workplaces after the metro opens [27]. Therefore, in order to improve the objectivity of analysis, this paper incorporates the OLS linear regression model for factor correlation analysis, which more closely follows the actual characteristics of the research object and provides a reference for the weighted ranking of the analytic hierarchy method.

In summary, rational planning of urban and rural transportation networks requires coordinating transportation construction with urban and rural social, economic, and ecological development, and ensuring that transportation networks follow the actual conditions of urban and rural development. Based on the research on the main influencing factors of transportation construction, OLS linear regression analysis could be used to study the relationship between an urban and rural transportation network and the influencing factors of living–production–ecological space. The correlation coefficient that affects features can further guide the importance ranking of factors. The analytic hierarchy method determines the weights of each element at each level and constructs the final urban and rural transportation network planning decision model. The method developed in this study may provide a valuable practice sample for the application and quantitative analysis of data in urban and rural transportation network planning.

3. Methodology

In the era of big data, the planning of urban and rural transportation network should balance the influencing factors in living, production, and ecological spaces through a scientific approach of assembling models. Each model can control its spatial statistics or importance ranking of factors, even to make a decision between the balancing of factors, but how to assemble them for easily finding several planning and development strategies of urban and rural integrated transportation networks is still a difficult task in urban and rural transportation network planning. This paper introduces a model assembly approach for planning an urban–rural transportation network to address this problem.

3.1. Workflow of Model Assembly Approach of Urban and Rural Transportation Network Planning

The workflow of the model assembly approach for urban and rural transportation network planning can be processed through the following steps:

• Collect multi-source data on network map data and POI data in the research area. All data are imported into the unified coordinate system and projected onto the administrative subdistrict map of Jiangxia. The China Geodetic Coordinate System 2000 (CGCS2000) is used for geographical coordinates in this study.
• Calculate and unify data form. The point data and line data are calculated by density analysis, and raster data are obtained. The area data are directly converted to raster data.
• Calculate the density of the transportation network in each subdistrict individually.
• Analyze the correlation between traffic density and its influencing factors by OLS linear regression model. The correlation results of factors provide a reference for the weight ranking of influencing elements of transportation network planning.
• Combined with the analytic hierarchy method, the weight values of various transportation network planning considerations are calculated, and the urban and rural transportation network planning decision-making model is constructed.
• Calculate the suitable area for traffic construction by raster math.
• Put forward corresponding optimization suggestions to solve the limit in urban and rural transportation network.

In steps i–vii, the methods are drawn and divided into three phases: (a) multi-source data processing phase, (b) model assembly, and (c) output results and discussion phase (Figure 1). The first phase and second phase are described in Section 3.2 and Section 3.3, respectively.

In order to link the attribute information obtained by the multi-source data processing in the first step through the weights calculated by the model assembly in the second step, this study developed a special calculation formula, that is, Equation (1):

$$S_i={\sum }_{i=1}^n({{\beta }_{ij}}_{(x_i,y_i)}\ast {\omega }_j)$$

(1)

where i is raster cell, j is the transportation network planning influencing factor, $S_i$ is the i-th raster cell calculation result, ${\omega }_j$ is the weight value of the j-th transportation network planning influencing factor, $(x_i,y_i)$ represents the geographic center coordinates of the i-th raster cell, ${\beta }_{ij}$ is the data attribute information of the j-th influencing factor in the i-th raster cell.

3.2. Transportation Network Planning Influencing Factor Analyses and Data Fusion

Transportation network density is an important indicator reflecting the level of urban transportation infrastructure construction [28,29,30]. In this study, traffic density was used as the independent variable to reflect the construction of urban and rural transportation networks. Common living spaces such as residential areas and living support service facilities are classified as living spaces. Enterprises, industrial zones, logistics areas, and agricultural production service areas are classified as production spaces. Natural elements such as parks, green spaces, and rivers are classified as ecological spaces. Then, an evaluation element system was constructed by considering these three types of spaces.

In the living spaces, the construction of the transportation network is oriented by the behavior of urban and rural residents, and the scale development level and land use nature of the town are closely related to the traffic [31,32,33]. At the same time, trunk transportation networks in urban and rural spaces run between urban and rural areas, and their layout scale directly affects the layout and scale of rural transportation networks [34]. Related supporting transportation facilities such as bus stations, parking lots, and long-distance bus stations also constitute the structural space of urban and rural transportation networks [35,36,37,38]. In the production spaces, it is necessary to coordinate the relationship between transportation and industrial development, ensure interregional interaction, transportation of goods, industrial and agricultural production, etc. and configure transportation facilities around industrial production, so as to improve the transportation capacity of urban and rural industries [39,40,41,42]. In the ecological spaces, natural factors such as geological topography and water system distribution affect the distribution of transportation networks. When planning the transportation network, scenic areas, nature reserves, cultural relics and monuments, and areas where mineral resources are distributed should also be considered as key objects [31,43,44,45], it is necessary to consider tourism and mineral resources and protect historical and cultural resources and ecological function areas.

By combining them, the study addressed 15 influencing factors, namely, size of town, urban and rural settlements, life services, supporting transportation facilities, trunk layout, external transport links, cargo hubs, logistics and transportation, enterprise distribution, agricultural production, terrain, distribution of water systems, tourism resources, heritage preservation, and ecological protection. These influencing factors were defined as decision-making factors for transportation network planning (Figure 2) in this study.

This study preprocessed the data separately according to different attributes and analyzed each element corresponding to each research unit (Figure 3).

3.3. The Model Assembly Approach

Based on the constructed urban and rural transportation network impact index system, the vector data of various influencing elements are processed to obtain the research data of the research unit. After normalizing the values, the OLS regression analysis model is used to calculate the correlation coefficient between the influencing features and the density of the transportation network. The priority is judged according to the size of the coefficient, and the final weight is obtained by combining the analytic hierarchy method. Then, each feature attribute is spatially associated with the corresponding weight. The assembly of the two models can effectively improve the objectivity of the analysis in transportation planning, which links transportation with social, economic, and ecological development, reducing problems such as inconsistencies between transportation networks and actual travel needs, insufficient links between roads and enterprise production, and damage to agricultural, ecological, and cultural protection. This model assembly approach can identify suitable areas for urban and rural transportation network construction and provides guidance and reference for urban and rural transportation network planning, which consists of four parts, namely, regression analysis model, transportation network planning decision hierarchy model, calculating the weight vector and testing the consistency of the judgment matrix, and multiplying the first-level and second-level weights to obtain the comprehensive weight.

In this aspect of the regression analysis model, each element corresponds to a research unit for analysis and calculation (

Table 1. Influence factor calculation.

Space	Influencing Factor	Analysis Description
Living Space	Size of town	The ratio of the area enclosed by the permitted construction area of the town to the area of the administrative subdistrict jurisdiction.
	Urban and rural settlements	Point kernel density analysis is used to obtain the raster cell values of the kernel density of urban and rural settlements in the study area. Then, the average value of the kernel density per unit area of the cells in each research unit is calculated, that is, the average point density of urban and rural settlements (the average point density is calculated as below).
	Life services	The total number of POIs of regional government, catering, medical, education, hotel, sports, shopping, and other life service facilities is calculated, and the average point density of life service functions is determined.
	Supporting transportation facilities	The mean point density of POI at bus stops and bus terminals.
	Trunk layout	The ratio of trunk line transportation network mileage to rural transportation network mileage.
Production space	External transport links	The total number of high-speed and expressway entrances and exits is counted, and the average point density of the external entrance and exit is determined.
	Cargo hubs	Cargo hub facility footprint statistics.
	Logistics and transportation	The average point density of the distribution of logistics and transportation-related service facilities.
	Enterprise distribution	Mean enterprise POI point density is determined.
	Agricultural production	The ratio of the area of cultivated land concentration in the study unit to the area of the study unit.
Ecological space	Terrain	The mean value of the slope of the study unit is determined based on the DEM data.
	Distribution of water systems	The ratio of the area of rivers and lakes in the study unit to the area of the study unit.
	Tourism resources	Average point density of distribution of tourism resources.
	Heritage preservation	Statistics on the number of cultural relics protected.
	Ecological protection	The ratio of the area of ecological protection redlines in the study unit to the area of the study unit.

). In order to facilitate the study comparison and analysis, it is necessary to standardize the data of each impact element. This study uses the range method for normalization, compressing the value to the range of 0 to 1 [46]. The calculation formula is as follows (Equation (2)):

$$c_n=\frac{C-C_{min}}{C_{max}-C_{min}}$$

(2)

where c is the data corresponding to each street administrative district of the influence element, $c_n$ is the normalized value, $c_{max}$ and $c_{min}$ are the maximum and minimum values of the indicator data, respectively.

In this study, a regression model was used to explore the influence of various factors on transportation network planning, and the relationship between different influencing factors and transportation network density. One of the most commonly used regression models in the linear regression model chamber, ordinary OLS is usually used to construct suitable intercept coefficients and slope coefficients for regression models [47,48,49,50]. In this study, the regression coefficients are calculated based on the least squares regression method, and the calculation model is as follows (Equation (3)):

$$y_i={\beta }_0+{\sum }_{j=1}^n{\beta }_j{X}_{ij}+{\epsilon }_i$$

(3)

where $y_i$ is the i-th variable, which represents the density value of the transportation network of the i-th study unit; $X_{ij}$ is the independent variable, which represents the analytical value of the j-related influencing element in the i-th street; ${\beta }_0$ is the intercept; ${\beta }_j$ represents the i-th variable regression coefficient, estimated by the least squares method; and ${\epsilon }_i$ is random error.

Based on the analysis of the influencing factors related to the transportation network planning, this study constructed the decision-making model of transportation network planning by using the three aspects of living space, production space, and ecological space as the criterion layer of the system, and 15 influencing factors as the index layer. As a decision-making analysis method combining qualitative and quantitative factors, analytic hierarchy quantifies the experience judgment of decision-makers, hierarchizes the relevant elements, and tests the rationality of the comparison results layer by layer [51].

In this study, analytic hierarchy was used to determine the weights of each influencing factor, and the basic steps and calculation Formulas (4)–(7) are as follows:

(i)

Establish the hierarchical structure model of influencing factors of road network planning;

(ii)

Based on the correlation degree analysis of the OLS linear regression model, the importance of indirect measurement of the weight of decision-making criteria is judged by pairwise comparison.

$$\mathrm{A}=\left\{\begin{array}{c}{\mathrm{a}}_{11}, {\mathrm{a}}_{12},\dots ,{\mathrm{a}}_{1\mathrm{j}}\\ ⋮, ⋮,\dots ,⋮\\ {\mathrm{a}}_{\mathrm{i}1}, {\mathrm{a}}_{\mathrm{i}2},\dots ,{\mathrm{a}}_{\mathrm{i}\mathrm{j}}\end{array}\right.$$

(4)

where A is the judgment matrix, ${\mathrm{a}}_{\mathrm{i}\mathrm{j}}$ represents the comparison result of the importance of element i to element j, and a quantitative scale of 1–9 can be used to reflect its importance [52].

(iii)

In terms of calculating the weight vector and testing the consistency of the judgment matrix, this study defines the follows equations:

$$\mathsf{\lambda }\mathrm{m}\mathrm{a}\mathrm{x}=\frac{1}{\mathrm{n}}{\sum }_i\frac{{\left(A\omega \right)}_i}{{\omega }_i}$$

(5)

$$\sigma \left(CI\right)=\frac{{\lambda }_{max}-n}{n-1}$$

(6)

$$\sigma \left(CR\right)=\frac{\sigma \left(CI\right)}{\sigma \left(RI\right)}$$

(7)

where n is the number of rows in the decision matrix and λ_max is the largest eigenvalue of the comparison matrix, A is the matrix operating on the ratio of $\omega$_i and $\omega$_j, $\omega$_i is the local weights, CI is the Consistency Index, and RI is the Random Consistency Index. When σ (CR) < 0.1, it is considered that the consistency is passed and logically reasonable; when σ (CR) > 0.1, it means that the consistency of the judgment matrix does not pass, and the judgment matrix needs to be re-corrected.

In terms of multiplying the first-level and second-level weights to obtain the comprehensive weight, this study implemented density analysis of various influencing factors or factor conversion to raster to obtain raster data. To avoid calculation errors caused by different raster data units, the data of each layer of raster cells were reclassified and assigned. According to the literature and the correlation analysis above, the influencing factors are assigned as values of 1 to 5 at different levels. The processing and grading of the influencing factors are described in

Table 2. Statistical table of data processing and grading assignment description.

Influencing Factor	Reclassification of Hierarchical Assignments	Illustration
Size of town	The inside and outside of the construction area are allowed to be assigned 5 and 3 at two levels, respectively.	There is a large demand for transportation network construction within the permitted scope of urban construction, and the construction conditions are relatively good, while the transportation network construction in other regions will be limited to a certain extent.
Urban and rural settlements	According to the natural discontinuity method, it is divided into five levels, and the value is assigned 5 to 1 according to the density value from high to low, respectively.	In areas with dense urban and rural residential areas, there is more traffic demand, so the transportation network construction should be more intensive and more complete.
Life services	Hierarchical assignment is the same as urban and rural settlements.	In areas with dense living service facilities, residents have more travel needs.
Supporting transportation facilities		If the transportation supporting service facilities are concentrated, the construction of the transportation network should be coordinated with the construction of supporting transportation facilities.
Trunk layout		In areas with intensive rural traffic construction, local rural traffic can be upgraded and incorporated into urban trunk roads according to actual needs.
External transport links		In the distribution area of the entrances and exits of expressways, the construction of the transportation network should be improved to facilitate urban and rural external transportation.
Cargo hubs	The inside and outside of the cargo hub area are allowed to be assigned 5 and 3 at two levels, respectively.	Transportation network construction should be improved in urban and rural cargo hub areas to facilitate cargo storage and transportation.
Logistics and transportation	Hierarchical assignment is the same as urban and rural settlements.	Transportation network construction should be improved in areas where urban and rural logistics and transportation service facilities are concentrated to facilitate logistics and transportation.
Enterprise distribution		The construction of transportation networks should be strengthened in areas where urban and rural enterprises are concentrated, so as to quickly transport out the production products.
Agricultural production	The inside and outside of the cultivated land concentrated distribution area are allowed to be assigned 1 and 3 at two levels, respectively.	In order to protect the distribution area of concentrated cultivated land, the construction of the transportation network should reduce or avoid planning in the area.
Terrain	Assign values according to the double evaluation grading, slope from low to high, assign a value of 5 to 1 respectively.	Areas with gentle slopes are more suitable for transportation network construction to reduce its cost and ensure road safety.
Distribution of water systems	The inside and outside of the distribution of water systems area are allowed to be assigned 1 and 3 at two levels, respectively.	Transportation network construction should be avoided in water system distribution areas to reduce the damage and pollution to water bodies.
Tourism resources	Hierarchical assignment is the same as urban and rural settlements.	In areas rich in tourism resources, transportation network construction should be strengthened to promote the development of urban and rural tourism industry.
Heritage preservation	The inside and outside of the heritage conservation buffer zone are allowed to be assigned 1 and 3 at two levels, respectively.	Cultural relics protection units shall reduce the construction of transportation networks in the control construction areas to reduce the damage caused by traffic congestion and pollution to cultural protection units.
Ecological protection	The inside and outside of the ecological protection redline areas are allowed to be assigned 1 and 3 at two levels, respectively.	Transportation network construction should reduce the number of areas crossing the ecological protection red line to protect ecological functional areas, environmental quality, and natural resources.

. Then, a raster of the 15 affected features was superimposed and determined in Equation (8) to obtain a raster layer suitable for planning a transportation network [53]. The result is divided into five levels according to the natural point discontinuity method. Areas with higher scores are more suitable for road construction, thereby providing a spatial reference for urban and rural transportation network planning.

$$V={ V}_1\ast {\omega }_1+{ V}_2\ast {\omega }_2+{ V}_3\ast {\omega }_3+\dots \dots +V_{15}\ast {\omega }_{15}$$

(8)

where V is the final evaluation result score, $V_n$ is the raster cell values calculated from Table 2 of the i-th influencing element, ${\omega }_n$ is the weight value of the i-th influencing element, which is calculated by step 2.

4. Study Area and Data

4.1. Study Area

Jiangxia District is located in the south of Wuhan, China, with a total area of 2009 km². The district has a low urbanization rate among Wuhan districts. There are numerous rural settlements, so the issue of transportation between urban and rural areas must be addressed. This paper uses subdistricts as research units to explore the correlation between urban and rural transportation network construction and the influencing factors in three types of spaces: living, production, and ecology, and is divided into 12 research units according to the administrative districts of streets in 2017 (Figure 4).

4.2. Dataset Description

The research data include the administrative subdistrict map of Jiangxia District in 2017, the transportation network data of Jiangxia District in 2022, the POI data of Jiangxia District in 2022, the administrative map and main functional area planning map of Jiangxia District in 2022, the DEM data of Jiangxia District in 2022, and the water system vector data of Jiangxia District in 2022 (

Table 3. Research data source statistics.

Data Name	Data Source	Data Format
Subdistrict map	Map sharing website, Self-drawn	Vector
Land use status	Wuhan Land Use Master Plan (2006~2020) and satellite imagery, Self-drawn	Vector
Transportation network	Open Street Map open-source map download platform	Vector
Service facility POI	Map of Gaode	Excel
Digital elevation model	Geospatial data cloud	Raster
Satellite imagery		Tiff
Water system	Satellite imagery, Self-drawn	Vector
Distribution of urban and rural settlements	Map of Jiangxia District, Self-drawn	Vector
Distribution of natural resources	Boyaa tourism sharing website and Wuhan mineral resources development and protection and utilization plan, Self-drawn	Vector
Logistics park	Planning layout of different functional units in Jiangxia District, Self-drawn	Vector
Agricultural production functional area		Vector
Ecological main functional area		Vector
Heritage sites	Boyaa Travel Sharing Website, Self-drawn	Vector

). Among them, the transportation network data come from the OpenStreetMap website, combined with the satellite image downloaded from the geospatial data cloud platform to collate the transportation network system including highways, national highways, provincial highways, county roads, township roads, and other roads (Figure 5).

5. Research Process and Results

5.1. Regression Analysis Results

Based on the collected research data and the constructed research methods, all indicators related to the first statistics and research were obtained, data processing was carried out, descriptive analysis (

Table 4. Descriptive statistics tables for all variables.

Influencing Factor	Minimum	Maximum	Average	Variance
Size of town	0	84.992	23.484	904.242
Urban and rural settlements	0.419	0.776	0.586	0.011
Life services	440	44,701	7880.25	204,348,132.932
Supporting transportation facilities	0	1.303	0.327	0.179
Trunk layout	0	3	0.92	0.674
External transport links	0	0.249	0.042	0.005
Cargo hubs	0	17.762	3.202	29.777
Logistics and transportation	0	0.038	0.010	0.000
Enterprise distribution	0.096	7.206	1.942	6.555
Agricultural production	0.077	0.773	0.413	0.049
Terrain	18.974	36.374	28.034	28.327
Distribution of water systems	0.008	0.328	0.170	0.008
Tourism resources	0.007	0.283	0.086	0.009
Heritage preservation	0	3	1.17	1.061
Ecological protection	0	0.447	0.108	0.018

) was carried out to clarify the overall situation of the study area, and the completeness and accuracy of the verification data were checked.

According to the dependent and independent variables selected in previous research, we constructed an OSL model of the influencing features of each space, and obtained the regression coefficient values with the least squares calculation method. According to the output regression results, the closer the value is to 1, the better the model fits. If $R^2$ (Living space) = 0.995035, $R^2$ (Production space) = 0.995035, $R^2$ (Ecological space) = 0.995035, it shows that the sample fits very well. The range of SRV (standardized residual values) for each space research unit were [−2.5, 2.5] (Figure 6), so all streets in the study area passed the residual test.

The final results were obtained by OSL regression analysis (

Table 5. OSL regression analysis statistical table.

Space	Living Space
Influencing factor	Size of town	Urban and rural settlements	Life services	Supporting transportation facilities	Trunk layout
Regression coefficient	0.145	−0.007	0.433	0.035	0.361
Space	Production Space
Influencing factor	External transport links	Cargo hubs	Logistics and transportation	Enterprise distribution	Agricultural production
Regression coefficient	0.072	−0.034	−0.131	0.779	−0.021
Space	Ecological Space
Influencing factor	Terrain	Distribution of water systems	Tourism resources	Heritage preservation	Ecological protection
Regression coefficient	−0.094	−0.107	0.808	0.017	−0.077

), and it can be seen from the regression model of each influencing element and transportation network density that most of the influencing factors are positively correlated with the transportation network density. However, the regression coefficient of urban and rural settlements was negative, and the correlation coefficient was low, which was not in line with the expected situation, indicating that some of the study areas should improve traffic accessibility by building roads in areas where urban and rural settlements are concentrated. Agricultural production, cultural relic protection, and ecological protection are negatively related to the transportation network density, which is consistent with the reality. This indicates that Jiangxia District has considered the protection of farmland and ecological environment in the transportation network planning, but the correlation intensity was not high, indicating that the management and control of ecological protection and farmland construction in Jiangxia District need to be strengthened. In addition, the correlation between terrain, the distribution of water system, and the density of transportation network is not high, which is related to the location of Jiangxia District in Jianghan Plain and the uniform distribution of lakes and rivers in the interior. As can be seen from the correlation analysis results of influencing factors, the transportation network planning in urban and rural areas was mainly affected by production and ecology. The common development of urban and rural industries is an important way to coordinate the development of urban and rural areas, with a large number of urban industrial zones distributed in the suburbs and a large number of agricultural production areas distributed in the countryside, so the transportation network in Jiangxia District is mainly arranged around promoting industrial production. At the same time, a large number of suburban and rural areas are also the main areas of spatial control planning. The distribution of ecological function areas in Jiangxia District also has an important impact on urban and rural transportation network layout.

5.2. Model Assembly Results

This paper conforms to the actual urban and rural evolution and development of Jiangxia District, and according to the coefficients obtained from the OSL model analysis, we take the sum of the absolute values of the coefficients of each influencing factor to judge the importance ranking of the criterion layer. The greater the influence of influencing factors on transportation network planning, the closer the absolute value of the regression coefficient is to 1. That is, production space > ecological space > living space. The matrix scale method was applied, combined with the coefficient size of each influencing factor and the expert score, and the criterion layer and the index layer were scored in pairs to obtain the judgment matrix. Using the normalize table by sum–product method, we obtained the weights of each index under its corresponding criterion layer and finally obtained the relative weight, absolute weight, and consistency test results of each index of the urban and rural transportation spatial comprehensive evaluation system (

Table 6. Standard layer judgment matrix.

Index	Living Space	Production Space	Ecological Space
Living space	1	1/2	1/3
Production space	3	2	1
Ecological space	2	1	1/2

,

Table 7. Standard layer judgment matrix of living space.

Index	Size of Town	Urban and Rural Settlements	Life Services	Supporting Transportation Facilities	Trunk Layout
Size of town	1	1/5	1/9	2	1/3
Urban and rural settlements	5	1	1/2	7	2
Life services	9	2	1	9	6
Supporting transportation facilities	1/2	1/7	1/9	1	1/5
Trunk layout	3	1/2	1/6	5	1

,

Table 8. Standard layer judgment matrix of production space.

Index	External Transport Links	Cargo Hubs	Logistics and Transportation	Enterprise Distribution	Agricultural Production
External transport links	1	8	4	1/2	1/2
Cargo hubs	1/8	1	3	1/8	1/6
Logistics and transportation	1/4	1/3	1	1/9	1/6
Enterprise distribution	2	8	9	1	2
Agricultural production	2	6	6	1/2	1

,

Table 9. Standard layer judgment matrix of ecological space.

Index	Terrain	Distribution of Water Systems	Tourism Resources	Heritage Preservation	Ecological Protection
Terrain	1	1/3	3/5	2	1/2
Distribution of water systems	3	1	3/2	3	5/4
Tourism resources	5/3	2/3	1	4	6/7
Heritage preservation	1/2	1/3	1/4	1	2/7
Ecological protection	2	4/5	7/6	7/2	1

and

Table 10. Weight and consistency test of each decision index.

Criterion layer	Weight	Index Layer	Weight	CI Value	Absolute Weight
B1 Living space	0.163	C1 Size of town	0.056	0.032 < 0.1	0.009
		C2 Urban and rural settlements	0.262		0.043
		C3 Life services	0.511		0.083
		C4 Supporting transportation facilities	0.035		0.006
		C5 Trunk layout	0.135		0.022
B2 Production space	0.297	C6 External transport links	0.207	0.081 < 0.1	0.061
		C7 Cargo hubs	0.052		0.015
		C8 Logistics and transportation	0.037		0.011
		C9 Enterprise distribution	0.424		0.126
		C10 Agricultural production	0.280		0.083
B3 Ecological space	0.540	C11 Terrain	0.128	0.013 < 0.1	0.069
		C12 Distribution of water systems	0.311		0.168
		C13 Tourism resources	0.231		0.125
		C14 Heritage preservation	0.073		0.039
		C15 Ecological protection	0.257		0.139

).

The remap value tool was used to assign values of 1 to 5 at different levels for the influencing factors (Figure 7). The single factor score was multiplied by the weight and all the factors were added to obtain the sum, to achieve the comprehensive evaluation of the suitability for planning transportation network.

5.3. Urban and Rural Transportation Network Planning Decision Model Results

The analysis results of transportation network planning decision model are shown in Figure 8. The range of the scores is between 1.6 and 3.88. Five layers of results are created based on the natural discontinuity point. The results of the regional calculations are divided primarily between 2.36 and 2.61. Higher scores are found in the red and orange regions, which is in line with the actual transportation network in Jiangxia District. The dark-blue and light-blue values are lower, which is consistent with the natural water system, ecological protection area, and concentrated distribution of cultivated land in the Jiangxia District. Therefore, the evaluation results of the decision model have a certain reference value. At the same time, according to the density analysis of the existing urban and rural transportation network in the Jiangxia District using the same evaluation method above, the density value from high to low were assigned values of 5 to 1, respectively. We compare the current transportation network construction situation with the subtraction of the analysis results of the decision model, and the results were divided into positive and negative values for visualization processing. From the analysis of the difference between the current transportation network and the results (Figure 9), it can be seen that the positive distribution area is consistent with the concentrated construction of the transportation network in the urban area of the Jiangxia District, which further verifies the validity of the analysis results of the transportation network planning decision model. The yellow area with the evaluation result of the decision-making model is also the area with a negative result of different analyses, so the yellow area is the key area to be considered for the future urban and rural transportation development in the Jiangxia District.

5.4. Discussion

5.4.1. Planning Strategies

The urban areas in Jiangxia District are concentrated in the northern half of the district, and the rural areas are distributed across the southern portion of the territory. The transition zone between urban and rural areas in Jiangxia District is the area with yellow findings from the planning decision-making model, indicating that Jiangxia District needs to strengthen the transportation connection between urban and rural areas. Combined with the previous OSL correlation analysis, the focus should be on the transportation network, urban and rural residential areas, cargo hubs, and agricultural production coordination planning. According to the evaluation results, Jiangxia District can be divided into planning units. As a result, different planning schemes and transportation policies can be formulated for the problems of different units. Areas with higher scores can be listed as improvement areas, most of which are urban construction concentration areas, which can develop new types of transportation, etc. The areas in the middle of the score can be considered as perfect areas, which are mainly urban and rural areas, and rationally build transportation networks, strengthen connections to rural settlements, and improve overall traffic accessibility. The remainder with low scores are listed as restricted areas in traffic planning, which are mainly ecologically concentrated areas that need to be protected and controlled to reduce ecological damage. After the basic division, and their refinement according to the actual situation, the spatial pattern of urban and rural transportation network can be effectively optimized.

Furthermore, based on the current situation and development needs of the urban and rural transportation network spatial layout in Jiangxia District, this paper proposes the following planning and development strategies of an urban and rural integrated transportation network:

(1)

Planning urban and rural transportation network system with urban and rural integration as the goal. In the central area of Jiangxia District, the supply level of transportation service facilities should match the living space of urban and rural residents. Rural roads in areas with concentrated distribution of urban and rural residential areas and living service facilities should be upgraded and connected to urban arterial roads. Thus, the urban and rural transportation axis is formed, which can gradually penetrate urban public transportation and slow-moving traffic in rural areas and establish a multi-mode public transport system in urban and rural areas.

(2)

Layout urban and rural transportation network based on industrial development. In the Jiangxia District, traffic development policies can be formulated separately according to concentrated industrial production and agricultural production areas. Industrial production zones, according to the production distribution of enterprises in Jiangxia District, give full play to the advantages of agglomeration and industrial corridor cooperation, and allow for planning of production and transportation corridors. Network nodes are planned according to cargo storage and transportation, logistics transportation hubs, so as to establish a comprehensive urban and rural transportation network. For agricultural production areas, transportation construction should reduce the damage and pollution to cultivated land and plan the transportation of agricultural products to the outside world on the basis of ensuring the integrity of agriculturally concentrated contiguous cultivated land.

(3)

Formulate time sequence of transportation network construction based on demand intensity. According to the evaluation results of transportation network planning decision analysis, for the region with a higher score, the transportation network construction is relatively complete, and for the regions with low scores, the needs of cultivated land and ecological protection, the construction time of transportation network construction projects in these regions can be implemented later. For the central region, the urgency of transportation network construction is higher than that of other regions, so priority should be given to the implementation of transportation network construction projects in this region. From this, construction timing zoning can be delineated.

This paper systematically summarized and proposed 15 secondary sub-factors that can be used in urban and rural transportation network planning and layout decision-making against the backdrop of the national land and spatial planning system, which starts from the three major spaces of life, production, and ecology. Based on this, this paper built a hierarchical model of urban and rural transportation network planning and decision-making. By introducing regression analysis and analytic hierarchy for the weight ranking of decision-making factors and the weight values of sub-factors, the methods and steps of planning and line-selection decision-making were proposed, and the results were tested to compensate for the shortcomings of urban and rural transportation networks in big data application and other interdisciplinary research approaches.

5.4.2. Model Comparison

Construction of transportation is mostly centered in urban areas in accordance with China’s territorial spatial planning system. The limits of the living–production–ecological space are drawn using the dual-evaluation technique. Location analysis is the basic foundation for the national dual-evaluation analysis technique’s traffic evaluation. Additionally, the location analysis is evaluated by two indicators, transportation line location conditions and traffic network density. By comparing the transportation network analysis model in this study and in the dual evaluation (

Table 11. Model comparison.

	This Study	The Dual Evaluation
Analyze factors	Size of town, urban and rural settlements, life services, supporting transportation facilities, trunk layout, external transport links, cargo hubs, logistics and transportation, enterprise distribution, agricultural production, terrain, distribution of water systems, tourism resources, heritage preservation, and ecological protection.	Traffic artery, urban accessibility, transportation hub, accessibility of surrounding cities, road density.
Analysis tool	Density analysis, feature to raster, raster reclassification.	Buffer analysis, raster reclassification.
Weight assignment	Combining regression analysis and analytic hierarchy to give weight to each factor.	The weights are assigned according to the size of the road level, highway = 1.5, first-class highway = 1, secondary highway = 0.5, and third-level highway = 0.3.
Overlay calculations	The sum of the product of weights and factor raster attribute values, which are calculated by raster math.	The sum of 10 [Location traffic conditions.tif] and [Transportation network density] which are calculated by raster math.
Results output	The results are divided into five levels according to the natural point discontinuity method.	The results show the scores of the two evaluations, ten digits represent the evaluation scores of regional traffic conditions, and the number of each number is the evaluation of traffic network density. Location condition evaluation takes precedence over transportation network density evaluation. With the same score, the former will be higher than the latter. Therefore, the results are divided into six levels.

), it can be seen that the processing flow of the two models are similar, all data are analyzed into raster data, and the final result is obtained through raster math. However, in the dual evaluation, the analyzed factors only include traffic, and the weight assignment is also subjective in distinguishing between different hierarchical roads.

The purpose of the transportation network analysis in the dual evaluation is to obtain an area with convenient transportation. Therefore, under the new territorial spatial planning system, planning focuses on the overall development of ecology, agriculture, and towns, and the paradigm of transportation planning will require adjustment. However, the dual-evaluation technology is not adequate for transportation planning due to different purposes. A comprehensive analysis method between traffic and multiple elements of space is lacking. The model assembly approach for planning an urban–rural transportation network in this research can comprehensively analyze the correlation between urban–rural transportation network planning and the influencing factors of living–production–ecological space.

6. Conclusions

The main contribution of this study is to provide a viable approach for identifying suitable areas for the construction of an urban and rural transportation network. This approach effectively considers the development of key industrial areas in urban and rural areas, the protection of ecological agricultural areas, and residents’ living and travel needs. By using this proposed approach in the study area, this study develops several planning and development strategies for an urban and rural integrated transportation network. Concerning the case studied, the following conclusions are drawn:

• The construction of a transportation system is not simply designed to meet the needs of spatial flow, but it also needs to consider land for various uses, urban construction, regional development, ecological protection, and social progress. Transportation planning must pay attention to the synergy between transportation elements and other land and spatial elements, and focus on the spatial analysis of the supply level of transportation facilities. Based on this case, the methods and steps for the model assembly approach of planning urban–rural transportation network decision-making were proposed, and the results were tested.
• The 15 important transportation planning decision-making factors in the living–production–ecological space were systematically proposed. The hierarchical model of urban and rural transportation planning decision-making was established, and the data acquisition methods of 15 decision-making factors were proposed.
• Combined with OSL regression analysis and AHP analysis, the weight ranking of transportation planning decision-making factors was determined. Regression analysis is added to the weight ranking step of the analytic hierarchy method, which effectively considers the actual situation of the development of the research area and reduces the subjective judgment component.

In the future, we will extend this research in the following respects: (i) Improving the system of influencing factors in urban–rural transportation planning. (ii) Studying the acquisition, storage, and analysis methods of factor data. This will broaden the capability of data acquisition, such as from government yearbooks, geological literature, maps, transportation network big data, remote sensing technology and other advanced big data technologies to achieve rapid update of decision-making data. (iii) Integrating intelligent technology for big data processing and smart decision-making in the urban–rural transportation network planning.