Coupling Coordination Research on Disaster-Adapted Resilience of Modern Infrastructure System in the Middle and Lower Section of the Three Gorges Reservoir Area

Abstract

Rapid incremental urbanization in China has resulted in an incomplete modern infrastructure system and multiple point-like flaws. This is due to a lack of funding and poor scientific construction concepts and procedures. This also contributes to the infrastructure system’s low disaster-adapted resilience and insufficient coupling coordination of production-oriented and service-oriented infrastructure subsystems. Based on the “Robustness-Rapidity-Redundancy-Resourcefulness-Durability” (4R-D) frameworks, this study screens 53 indicators across three tiers of “production-oriented, service-oriented, intelligent” infrastructure subsystems to establish a modern infrastructure resilience evaluation system. We examined the overall infrastructure resilience and coupling coordination development among subsystems in the Three Gorges Reservoir Area (TGRA) from 2009 to 2020 using a coupling coordination degree model (CCDM). Grey relational analysis (GRA) was used to analyze the significant control aspects of infrastructure resilience and coupling coordination degree based on grey system theory. The findings show the following: (1) at the macro level the overall resilience, resilience of each subsystem, and coupling coordination among subsystems in the research region show an upward trend from 2009 to 2020, with the rise from 2018 to 2020 being the most significant; (2) at the micro level, from 2010 to 2013, there was no obvious spatial divergence and from 2014 to 2020, driven by the radiation of the two major urban agglomerations, the resilience and coupling coordination of Yiling and Wanzhou both show a trend of more substantial increase, while the rest of the counties have a small increase; and (3) at the meso level, seven factors have a more significant impact on the coupled and coordinated development of urban infrastructure than other indicators, including urbanization rate, average annual rainfall, the number of health technicians per 10,000 people, and the percentage of GDP in the tertiary industrial sector.

1. Introduction

In the context of intensifying extreme weather disasters, rising urban security incidents, and global economic integration, China’s urbanization model of high-speed incremental development over the last 40 years has resulted in inadequate system networks, structural and functional deficiencies, and limited infrastructure operational capacity of modern infrastructure systems due to insufficient construction science and technology. Low catastrophe resilience is widespread in China’s cities, particularly in less developed central and western regions. Natural and man-made disasters have put modern infrastructure systems under strain in recent years. For example, the COVID-19 outbreak in 2019 exposed the weaknesses of many forms of infrastructure, with a more significant impact on service-oriented infrastructure. China’s total logistics freight volume was 4.5 billion tons from January to February 2020, a 19.8% year-on-year decline [1,2]. In 2021, heavy rains in Zhengzhou, Henan province, damaged 2639 highways and 6554 familiar main stem roads, resulting in direct losses of 10.9 billion yuan [3,4]. Extreme heat blanketed China in 2022, with high-temperature warnings issued for an entire month and a maximum temperature of 45 °C, resulting in power outages in multiple regions and significantly impacting the power infrastructure system [5,6]. The figures shown above indicate that the current infrastructure system’s risk resilience in the face of calamities is considerably lacking. Furthermore, the structure and function of urban hardware facilities are vulnerable in terms of disaster-adapted operation, the network system is insufficient, the connection between related facilities is hampered, and software facilities’ operation and management capability are limited. Moreover, the post-disaster recovery has revealed the inadequacy of software facilities’ operational capacity. The National 14th Five-Year Plan emphasizes the importance of coordinating and promoting the construction of both traditional infrastructure and new infrastructure, as well as developing a modern infrastructure system that is comprehensive, efficient, practical, intelligent, environmentally friendly, safe, and dependable.

Traditional production-oriented infrastructure, service-oriented infrastructure, and innovative, intelligent infrastructure comprise the modern infrastructure system. The traditional production-oriented infrastructure includes the transportation systems (water, land, and air), energy systems (water, electricity, and gas), water supply and drainage systems, communication systems (radio, television, and networks), and emergency disaster protection systems. The traditional service-oriented infrastructure includes, among other things, the administrative management system, the medical and health system, the cultural and education system, and the social welfare system. The emerging intelligent infrastructure primarily consists of the latest information infrastructure, such as 5G, the internet of things (IoT), and the industrial internet, as well as new convergence infrastructures, such as big data centers and artificial intelligence. Production-oriented infrastructure functions as the skeleton of cities and towns, ensuring the safe and efficient flow of “ecology-production-life” factors, whereas service-oriented infrastructure functions as the meridian of cities and villages, operating and controlling the safety and efficiency of the “ecology-production-life” state. Furthermore, as the town’s brain, the emerging intelligent infrastructure couples and coordinates with production-oriented and service-oriented infrastructure systems to jointly support the town system’s safe, healthy, and effectual functioning. In the face of disasters the structural and functional resilience of production-oriented infrastructure, and the “coupled-linked” and “coordinated-interacted” ability of service-oriented infrastructure, express the resilience of modern infrastructure systems. Additionally, it refers to the emerging intelligent infrastructure’s ability to analyze massive volumes of data, the system’s adaptability to post-disaster recovery, and the provision of scientific data analysis to support decision-making (Figure 1).

Disaster-adapted resilience and sustainable system development are linked to the safety and health of the nation and the region as a cornerstone of the regular operation of cities and towns, as vital support for regional economic and social development, as a material guarantee for people’s healthy lives, and as a solid barrier for national and regional security and stability. The government and relevant scholars have been focusing on assessing the disaster-adapted resilience of modern infrastructure systems, analyzing the coupling and coordination among subsystems, identifying the central control elements influencing the coupling coordination degree, and exploring effective technical strategies to increase resilience.

The concept of infrastructure resilience evaluation began in the realm of earthquake engineering. Bruneau et al. [7,8] established the first widely used framework for seismic resilience evaluation, which treats disaster-adapted resilience performance and recovery duration as resilience indicators. Qualitative infrastructure resilience assessments primarily focus on developing a conceptual system framework for resilience [9,10,11] and assessing resilience using the Likert scale or Delphi method. On the other hand, quantitative assessments are mainly divided into static quantity measures and dynamic evolution of resilience. The former primarily uses fixed Bayesian networks [12,13,14], whereas the latter primarily uses dynamic network flow models [15,16] based on scenario tree generation algorithms, cascading failure models [17,18] based on complex networks, and Hornberger–Spear–Young (HSY) algorithms [19,20] based on Monte Carlo simulations to conduct resilience process simulation modeling for different perturbation scenarios, resolve infrastructure system structures and investigate resilience mechanisms. Most earlier studies have focused on the resilience assessment of a single infrastructure under a specific disturbance event scenario [21,22,23]. As research has progressed, scholars have obtained a more thorough grasp of infrastructure resilience. They began concentrating on the correlation between two or more infrastructures [24,25,26,27] and used interdependence models to examine them. With the continuous development of intelligent supporting technologies such as big data and IoT technology in recent years, some progress has been made in researching new intelligent infrastructure resilience assessments [28]. Still, less attention has been paid to the degree of coupling coordination between traditional infrastructure and new infrastructure. The multi-dimensional analysis of systemic thinking has to be expanded.

The study examined the coupling mechanism between three systems based on the architecture of a modern infrastructure system: production-oriented infrastructure, service-oriented infrastructure, and emerging intelligent infrastructure. The resilience features of urban infrastructure in the TGRA were discovered using the 4R and 4R-D frameworks. In addition, a framework for analyzing the resilience of urban infrastructure was built using a mathematical method, the amount of coupling and coordination among the three subsystems was quantitatively analyzed, and the critical control components influencing the level of coupling were evaluated.

2. Study Area and Data Sources

2.1. Study Area

The TGRA is a 58,000-square-kilometer area flooded by the Three Gorges Project development, lying in the transition zone between China’s second and third terraces. It is a critical hurdle in ensuring the Yangtze River Basin’s ecological security. It is also an essential environmental function area [29,30] on the Yangtze River Economic Belt, which is strategically vital. Because the TGRA is a geographical area formed by the Three Gorges Project, its distinct formation method, development pattern, geographical environment, and climatic conditions have resulted in a slew of urban diseases during the urbanization process, including unbalanced urban–rural development, insufficient infrastructure construction, and severe soil erosion [31]. With the continued promotion of strategic policies such as the development of the Yangtze River Economic Belt and the middle reaches of the Yangtze River urban agglomeration, the study of the multi-dimensional coupling resilience between the “productive-service-oriented-intelligent” infrastructure subsystems has emerged as a central driving force for the region’s sustainable development of the “ecology-production-life” space [32,33] (Figure 2).

The study region includes the Hubei section, which consists of four counties and districts in Hubei province (Yiling district, Zigui county, Xingshan county, Badong county), and the Chongqing section, which consists of four counties and districts in Chongqing city (Wanzhou district, Wushan county, Fengjie county, Yunyang county). The Hubei section is the radiation of the urban agglomeration in the middle reaches of the Yangtze River. In contrast, the Chongqing section belongs to the Da-Wan city-and-town concentrated area and is the radiation area of the Chengyu urban agglomeration. The Three Gorges Project was completed in 2009, and follow-up work for immigration began in 2010, the first year of the post-Three Gorges period. With the progressive improvement of town construction and the acceleration of urbanization, it has amassed more than 5724 km of various roads in the TGRA’s immigrant resettlement region. It will enhance the environment of the urban immigrant resettlement area by more than 168 square kilometers by 2021. Furthermore, the construction efficiency of modern infrastructure has gradually improved. As a result, the research period for this study is from 2009 to 2021, based on the post-Three Gorges period.

2.2. Data Sources

The primary data for this study include statistical data from the municipal, county, and district level statistical yearbooks of Yichang, Enshi, and Chongqing from 2009 to 2021. Some data are also acquired from statistical information and statistical bulletins of relevant government departments. Some counties and districts’ missing data are estimated using municipal data, while some years’ missing data are interpolated from adjacent years. Some intelligent infrastructures were not generally created in the county and district until the post-Three Gorges period, therefore, the data for some missing years are represented by 0 to reflect that new infrastructures were not regularly made in the county and district. The specific sources of the relevant data are shown in

Table 1. Sources of representative data.

Guideline Layer	Type of Data	Representative Indicators	Data Sources
Production-oriented Infrastructure	Extreme weather data	Frequency of extreme weather.	1. Meteorological administration; 2. Historical Weather Website (https://lishi.tianqi.com/, accessed on 3 August 2022)
	Geological hazard data	Geological disaster potential sites.	Bureau of Natural Resources and Planning
	Structure resilience data	Urbanization rate; highway road network density; annual water supply per 10,000 people	1. Yichang Statistical Yearbook; Chongqing Statistical Yearbook (section of urban construction; government finance resources and environment); 2. Statistical Yearbook of Chongqing Construction System
	Regulation resilience data	Percentage of transportation expenditure; water coverage rate
Service-oriented Infrastructure	Disturbance data	Average annual rainfall; average regional ambient noise.	1. Meteorological administration; 2. Statistical bulletins
	Structure resilience data	Number of tertiary hospitals per 10,000 people; number of library books per 10,000 people; greening coverage rate of built-up areas.	1. Yichang Statistical Yearbook; Chongqing Statistical Yearbook (section of public health, education, science, technology and culture); 2. Baidu Map open data platform (https://lbsyun.baidu.com/, accessed on 3 August 2022)
	Regulation resilience data	Percentage of expenditure for general public services; comprehensive radio and TV coverage rate.	Yichang Statistical Yearbook; Chongqing Statistical Yearbook (section of culture and government finance);
Emerging Intelligent Infrastructure	Disturbance data	Frequency of major safety accidents per 10,000 people.	1. Emergency Management Bureau; 2. Statistical bulletins
	Structure resilience data	Number of 5G base stations per 10,000 people; degree of comprehensive construction of emerging intelligent infrastructure.	1. Big Data Application and Development Administration; 2. Science and Technology Bureau
	Regulation resilience data	Ratio of R & D internal expenditure to GDP.	Yichang Statistical Yearbook; Chongqing Statistical Yearbook (section of science);

.

3. Resilience Measurement and Research Method

3.1. Construction of Infrastructure Resilience Index System

3.1.1. 4R Framework and 4R-D Framework

Bruneau et al. established the 4R framework [7,34], which consists of four dimensions: robustness, rapidity, redundancy, and resourcefulness, by evaluating the features of resilience processes. For example, robustness denotes the ability to preserve the performance of the infrastructure system when it is disrupted, limiting the damage caused by the disaster. Rapidity is the system’s ability to promptly rebound to a specified level following a disturbance. Redundancy means that the vital infrastructure has a specific reserve capacity to perform at the appropriate level during a crisis without being paralyzed. The ability of the system to intelligently regulate resources and optimize decisions is represented by resourcefulness. Following research, the durability dimension was introduced to the 4R framework and expanded into the 4R-D resilience assessment framework [35]. Durability measures the average time a system remains above a critical condition after being perturbed.

The infrastructure resilience features of the TGRA are studied as “disturbance type, structure resilience, and regulatory resilience” based on the 4R-D framework (Figure 3), taking the characteristics of the infrastructure construction and development in the TGRA into account. The types of disturbances experienced by each infrastructure subsystem are primarily divided into natural and artificially made disorders, which reflect the system’s disaster riskiness and resilience; into structural resilience, which demonstrates the robustness and redundancy of the infrastructure network structure and is used to measure its self-adaptive capacity in the face of various disturbances; and into regulatory resilience, which is critical to whether the system can return to normal operation.

3.1.2. Selection of Indicator Set

Under the principles of science, comprehensiveness, objectivity and feasibility, a preliminary set of indicators for assessing infrastructure resilience was established based on the “production-oriented, service-oriented, intelligent” infrastructure system, combining the above three dimensions: disturbance type, structural resilience, and regulatory resilience, with a total of 53 indicators, as shown in

Table 2. The preliminary set of indicators for assessing the resilience of urban infrastructure in the TGRA.

Guideline Layer	Criteria Layer	Index Layer	Index Direction
Production-oriented Infrastructure	Disturbance Type	R1 Frequency of extreme weather (%)	Negative
		R2 Industrial wastewater emission per unit GDP (million tons/billion yuan)	Negative
		R3 Industrial solid waste generation per unit GDP (tones/billion yuan)	Negative
		R4 Geological disaster potential sites (unit)	Negative
	Structure Resilience	R5 Urbanization rate (%)	Positive
		R6 Number of industrial enterprise units above the scale per unit GDP (unit/billion yuan)	Positive
		R7 Highway road network density (km/km2)	Positive
		R8 Waterway freight volume per unit GDP (million tons/billion yuan)	Positive
		R9 Effective service area coverage rate of civil airports (%)	Positive
		R10 Annual water supply per 10,000 people (10,000 tons/10,000 people)	Positive
		R11 Density of water supply pipeline network in built-up areas (km/km2)	Positive
		R12 Electricity Consumption in Rural Areas per 10,000 people (million kWh/million people)	Positive
		R13 Natural gas supply per 10,000 people (million m3/10,000 people)	Positive
		R14 Number of emergency shelters per 10,000 people (unit/10,000 people)	Positive
	Regulation Resilience	R15 Percentage of GDP in the secondary industrial sector (%)	Positive
		R16 Percentage of transportation expenditure (%)	Positive
		R17 Percentage of expenditure for urban and rural community affairs (%)	Positive
		R18 Percentage of expenditure for disaster prevention and emergency management (%)	Positive
		R19 Water coverage rate (%)	Positive
		R20 Gas coverage rate (%)	Positive
		R21 Integrated utilization rate of industrial solid waste (%)	Positive
		R22 Urban sewage treatment rate (%)	Positive
Service-oriented Infrastructure	Disturbance Type	R23 Average annual rainfall (mm)	Negative
		R24 Amount of domestic waste removal per 10,000 people (10,000 tons per 10,000 people)	Negative
		R25 Average regional ambient noise dB (A)	Negative
		R26 Number of persons receiving minimum living allowance per 10,000 population (person/million people)	Negative
	Structure Resilience	R27 Number of township governments per 10,000 people (unit/10,000 people)	Positive
		R28 Number of village committees per 10,000 people (unit/10,000 people)	Positive
		R29 Number of tertiary hospitals per 10,000 people (unit/10,000 people)	Positive
		R30 Number of CDCs per 10,000 people (unit/10,000 people)	Positive
		R31 Number of medical institution beds per 10,000 people (bed/10,000 people)	Positive
		R32 Health technicians per 10,000 population (person/10,000 people)	Positive
		R33 Number of library books per 10,000 people (10,000 books/10,000 people)	Positive
		R34 Number of students in general secondary schools per 10,000 people (person/10,000 people)	Positive
		R35 Number of general secondary school staff per 10,000 people (person/10,000 people)	Positive
		R36 Greening coverage rate of built-up areas (%)	Positive
		R37 Number of beds in adoption-type institutions per 10,000 people (beds/10,000 people)	Positive
	Regulation Resilience	R38 Percentage of GDP in the tertiary industrial sector (%)	Positive
		R39 Percentage of expenditure for general public services (%)	Positive
		R40 Percentage of health expenditure (%)	Positive
		R41 Percentage of education expenditure (%)	Positive
		R42 Percentage of social security expenditure (%)	Positive
		R43 Comprehensive radio and TV coverage rate (%)	Positive
Emerging Intelligent Infrastructure	Disturbance Type	R44 Frequency of major safety accidents per 10,000 people (unit/10,000 people)	Negative
		R45 Number of criminal cases filed per 10,000 people (unit/10,000 people)	Negative
	Structure Resilience	R46 Number of big data centers per 10,000 people (unit/10,000 people)	Positive
		R47 Number of high-tech enterprises per 10,000 people (unit/10,000 people)	Positive
		R48 Number of 5G base stations per 10,000 people (unit/10,000 people)	Positive
		R49 Number of new energy vehicle charging piles per 10,000 people (unit/10,000 people)	Positive
		R50 Degree of comprehensive construction of emerging intelligent infrastructure (%)	Positive
	Regulation Resilience	R51 Number of R & D personnel per 10,000 people (person/million people)	Positive
		R52 Ratio of R & D internal expenditure to GDP (%)	Positive
		R53 Percentage of science and technology expenditure (%)	Positive

.

3.2. Indicator Data Pre-Processing

3.2.1. Standardization of Indicator Data

In order to eliminate dimensional differences in the data of different indicators and to facilitate comparison between indicators, this paper applies the extreme value method [36,37] to standardize the specific data of indicators with the following formula.

$$X_{ij}^{\prime }=\{\begin{array}{c}\frac{x_{ij} - \underset{1\le j\le m}{\min }(x_{ij})}{\underset{1\le j\le m}{\max }(x_{ij}) - \underset{1\le j\le m}{\min }(x_{ij})} j\in PI \\ \frac{\underset{1\le j\le m}{\max }\left(x_{ij}\right) - x_{ij}}{\underset{1\le j\le m}{\max }(x_{ij}) - \underset{1\le j\le m}{\min }(x_{ij})} j\in NI\end{array}$$

(1)

where PI stands for positive indicator, i.e., the larger the data, the stronger the resilience; NI stands for negative indicator, the larger the data, the weaker the resilience; $x_{ij}$ stands for the jth indicator in year i of the county; $X_{ij}^{\prime }$ stands for the standardized value.

3.2.2. Correlation Screening of Data

In order to eliminate the influence of redundant information on the evaluation results, this paper adopted Pearson’s correlation analysis [38,39], aiming to verify the independence of indicators by calculating the correlation coefficient between two evaluation indicators. Firstly, according to the relevant studies, the critical value of the correlation coefficient E was defined as 0.9 [40]. Secondly, the Pearson correlation coefficient between each indicator was calculated separately, and the maximum value of the correlation coefficient in each indicator was set as ${\left|r\right|}_{max}$. When ${\left|r\right|}_{max}>E$, it means that there was a high overlap of indicator information, the corresponding indicator should be deleted, and the correlation coefficient should be recalculated for the remaining indicators until the ${\left|r\right|}_{max}$ of all indicators was less than E. The Pearson correlation coefficient was calculated as follows.

$$r=\frac{{{\displaystyle \sum }}_{i=1}^n\left(X_i-\overline{X}\right)\left(X_i^{\prime }-{\overline{X}}^{\prime }\right)}{\sqrt{{{\displaystyle \sum }}_{i=1}^n{\left(X_i-\overline{X}\right)}^2}\times \sqrt{{{\displaystyle \sum }}_{i=1}^n{\left(X_i^{\prime }-{\overline{X}}^{\prime }\right)}^2}}$$

(2)

where r represents the Pearson correlation coefficient; $X_i$ and $X_i^{\prime }$ are individual samples; n represents the sample size; $\overline{X}$ and ${\overline{X}}^{\prime }$ are the mean values of the sample.

3.2.3. Calculation of Indicator Data Weighting

Weights were assigned to the indicators in

Table 3. Indicator correlation coefficients and weighting results.

Indicator	${\left|\mathit{r}\right|}_{\mathit{m}\mathit{a}\mathit{x}}$	W1j	W2j	W3j	$\overline{{\mathit{w}}_{\mathit{j}}}$
R1	0.348 ***	0.0100	0.0139	0.0443	0.0228
R2	0.609 ***	0.0069	0.0136	0.0888	0.0364
R3	0.609 ***	0.0088	0.0005	0.0801	0.0298
R4	0.756 ***	0.0169	0.6123	0.0417	0.2236
R5	0.860 ***	0.0320	0.0139	0.0175	0.0211
R6	0.557 ***	0.0592	0.0007	0.0536	0.0378
R7	0.749 ***	0.0264	0.0008	0.0228	0.0167
R8	0.756 ***	0.0763	0.0054	0.0725	0.0514
R9	0.759 ***	0.0486	0.0399	0.0269	0.0384
R10	0.795 ***	0.0556	0.0229	0.0491	0.0425
R11	0.657 ***	0.0470	0.0075	0.0266	0.0270
R12	0.656 ***	0.0494	0.1291	0.0361	0.0715
R13	0.860 ***	0.0577	0.0469	0.0632	0.0559
R14	0.784 ***	0.0622	0.0001	0.0337	0.0320
R15	0.545 ***	0.0282	0.0171	0.0188	0.0213
R16	0.581 ***	0.0261	0.0075	0.0436	0.0257
R17	0.509 ***	0.0577	0.0074	0.0634	0.0428
R18	0.768 ***	0.2956	0.0006	0.1815	0.1592
R19	0.815 ***	0.0038	0.0078	0.0045	0.0054
R20	0.815 ***	0.0081	0.0186	0.0116	0.0128
R21	0.622 ***	0.0109	0.0247	0.0152	0.0170
R22	0.706 ***	0.0125	0.0089	0.0045	0.0087
R23	0.378 ***	0.0129	0.3095	0.0261	0.1161
R24	0.697 ***	0.0258	0.0001	0.0422	0.0227
R25	0.472 ***	0.0082	0.0038	0.0067	0.0062
R26	0.854 ***	0.0403	0.3781	0.0824	0.1669
R27	0.746 ***	0.0727	0.0001	0.0324	0.0351
R28	0.760 ***	0.0783	0.0024	0.0501	0.0436
R29	0.811 ***	0.2518	0.0000	0.2079	0.1532
R30	0.877 ***	0.1092	0.0000	0.0895	0.0662
R31	0.762 ***	0.0217	0.0178	0.0453	0.0282
R32	0.771 ***	0.0220	0.0161	0.0444	0.0275
R33	0.877 ***	0.0664	0.0002	0.0925	0.0530
R34	0.792 ***	0.0466	0.2041	0.0330	0.0945
R35	0.784 ***	0.0395	0.0057	0.0158	0.0203
R36	0.744 ***	0.0098	0.0082	0.0175	0.0118
R37	0.547 ***	0.0429	0.0209	0.0610	0.0416
R38	0.529 ***	0.0217	0.0132	0.0283	0.0211
R39	0.527 ***	0.0390	0.0048	0.0418	0.0285
R40	0.564 ***	0.0297	0.0038	0.0314	0.0216
R41	0.642 ***	0.0333	0.0054	0.0231	0.0206
R42	0.502 ***	0.0196	0.0039	0.0267	0.0167
R43	0.706 ***	0.0087	0.0021	0.0022	0.0043
R44	0.827 ***	0.0036	0.0120	0.0554	0.0237
R45	0.380 ***	0.0044	0.1933	0.0187	0.0721
R46	0.789 ***	0.2000	0.0002	0.1463	0.1155
R47	0.703 ***	0.0523	0.0088	0.0601	0.0404
R48	0.620 ***	0.2452	0.0223	0.2005	0.1560
R49	0.349 ***	0.3240	0.0595	0.3363	0.2399
R50	0.768 ***	0.0400	0.4026	0.0380	0.1602
R51	0.553 ***	0.0485	0.2769	0.0531	0.1262
R52	0.682 ***	0.0550	0.0154	0.0568	0.0424
R53	0.725 ***	0.0270	0.0090	0.0349	0.0237

Note: *** represents 1% level of significance.

using the entropy-CRITIC-coefficient of variation (COV) combination weighting approach [41,42,43]. The COV method [44,45], among others, assigns weights to each indicator based on the degree of variation between the current and target values of each evaluation indicator. The entropy weight method [29,46] calculates the weight coefficients based on the quantity of information carried by each assessment component, overcoming information overlap between indicators and minimizing subjective factor interference. Additionally, Diakoulaki et al. [47] presented the CRITIC technique, which uses correlation analysis to identify comparisons between criteria, decreasing the influence of partially correlated indicators and further reducing indicator information overlap. The calculation is also carried out using Matlab software [48,49]. Assuming that the three methods have the same relevance [50,51] and setting α = β = γ = 1/3, the cumulative weights are as follows:

$$\overline{W_j}=\alpha W_{1j}+\beta W_{2j}+\gamma W_{3j}\left(\alpha =\beta =\gamma =\frac{1}{3}\right)$$

(3)

In summary, the correlation coefficients of the indicators are all less than 0.9, indicating that the indicators are independent. The initial set of indicators is the final infrastructure resilience assessment framework, whose correlation coefficients and weights are calculated as shown in Table 3.

3.3. Calculation of the Infrastructure Resilience Comprehensive Index

Based on the above pre-processing methods such as the data standardization, correlation screening and weight calculation, the weight grade method [52] was used to calculate the comprehensive index of infrastructure resilience for cities and towns in the TGRA with the following formula.

$$R_j={{\displaystyle \sum }}_{j=1}^n{W}_j\times X_{ij}^{\prime }$$

(4)

R_j is the resilience index of each infrastructure subsystem in the calculation; W_j is the estimated indicator weight value; and $X_{ij}^{\prime }$ is the standardized value of the jth index.

3.4. Coupling Degree and Coupling Coordination Degree Model (CCDM)

The term “coupling” originates in physics and is defined as a measure of two entities’ dependence on each other. The coupling degree measures the strength of the coupling between three systems, but it does not show if the systems are developed in a coordinated manner. As a result, the CCDM is employed to measure the degree of coupling coordination between three-dimensional systems [53,54,55], which are both expressed as follows:

$$C=3\times {\left[\frac{R_1\times R_2\times R_3}{{\left(R_1+R_2+R_3\right)}^3}\right]}^{\frac{1}{3}}$$

(5)

$$S=\sqrt{C\times T}, T=a\times R_1+b\times R_2+c\times R_3$$

(6)

In the formula, R₁ represents the resilience index of the production-oriented infrastructure subsystem, R₂ represents the resilience index of the service-oriented infrastructure subsystem, and R₃ represents the resilience index of the emerging intelligent infrastructure subsystem; T represents the three systems’ comprehensive development level; a, b, and c represent the undetermined coefficients (i.e., weighting coefficients) of production-oriented infrastructure, service-oriented infrastructure, and emerging intelligent infrastructure, respectively, and a + b + c = 1. According to the studies of Qian et al. [56,57], set a = b = c = 1/3 to reflect the equal relevance of the three infrastructural subsystems to human society. C is the degree of coupling between the three systems, and its value range is (0,1); the closer C is to 1, the higher the coupling correlation between the systems; S denotes the degree of coupling coordination between the three systems.

According to the median segmentation approach [58] and Zhang et al.’s investigation [59], the coupling degree C is graded with the coupling coordination degree S, as shown in

Table 4. Grading criteria for the coupling degree of urban infrastructure system in the TGRA.

Coupling Degree C	C = 0	0 < C ≤ 0.3	0.3 < C ≤ 0.5	0.5 < C ≤ 0.8	0.8 < C ≤ 1.0
Coupling Degree Development Stage	Disorderly State	Low-level Coupling	Antagonistic State	Medium-level Coupling	High-level Coupling

and

Table 5. Grading criteria for the coupling coordination degree of urban infrastructure system in the TGRA.

Level of Coupling Coordination	Coupling Coordination Degree S	Meaning
Low-level Coupling Coordination	0 ≤ C ≤ 0.1	Extreme imbalance
	0.1 ≤ C ≤ 0.2	Severe imbalance
	0.2 ≤ C ≤ 0.3	Moderate imbalance
	0.3 ≤ C ≤ 0.4	Mild imbalance
Medium-level Coupling Coordination	0.4 ≤ C ≤ 0.5	Approaching imbalance
	0.5 ≤ C ≤ 0.6	Reluctant coordination
	0.6 ≤ C ≤ 0.7	Preliminary coordination
High-level Coupling Coordination	0.7 ≤ C ≤ 0.8	Moderate coordination
	0.8 ≤ C ≤ 0.9	Excellent coordination
	0.9 ≤ C ≤ 1.0	High-quality coordination

.

3.5. Grey Relational Analysis (GRA) Method

The development of a system is often influenced by many factors, each of which has a different degree of influence on the system. The relational degree measures the strength of the correlation between two systems for factors that vary over time. Grey system theory [60] is the study of uncertain systems in which “some information is known and some information is unknown.” In this system, the GRA is a method to measure the degree of correlation between factors and reflects the merit ranking of the evaluated object. If the trend of change of two factors is consistent, that is, if the degree of synchronous change is high, then the correlation between them is high. In this paper, the GRA method [61,62] is used to compare the degree of association between each resilience index and the comprehensive resilience level. Finally, the main control elements affecting the infrastructure resilience of the TGRA are studied and judged, and the formula is as follows:

$${\mathsf{\Delta }}_i\left(j\right)=\left|Z_0\left(j\right)-Z_i\left(j\right)\right|$$

(7)

$$\mathsf{\Delta }\left(max\right)=\underset{i}{\max }\underset{j}{\max }{\mathsf{\Delta }}_i\left(j\right)$$

(8)

$$\mathsf{\Delta }\left(min\right)=\underset{i}{\min }\underset{j}{\min }{\mathsf{\Delta }}_i\left(j\right)$$

(9)

$${\delta }_i\left(j\right)=\frac{\mathsf{\Delta }\left(min\right)+\rho \mathsf{\Delta }\left(max\right)}{{\mathsf{\Delta }}_i\left(j\right)+\rho \mathsf{\Delta }\left(max\right)}$$

(10)

$${\delta }_i=\frac{1}{n}{{\displaystyle \sum }}_{i=1}^n{\delta }_i\left(j\right)$$

(11)

In the formula, ${\delta }_i\left(j\right)$ is the relative difference between the comparison series Z_i and the reference series Z₀ on the jth evaluation index; ρ stands for the discrimination coefficient, which takes the value of (0,1) and will not affect the sorting results, but only the size of the value. This paper takes $\rho$= 0.5. ${\delta }_i$ is the correlation degree, and the larger its value, the greater the influence of the index on the comprehensive resilience level.

4. Results

4.1. Macro Level—The Temporal Evolution of Resilience and Coupling Coordination

Based on the Entropy–CRITIC–COV combination weighting method and the comprehensive index method, the overall resilience of the infrastructure system and the resilience of each subsystem in the middle and lower section of the TGRA were calculated according to Equations (1) and (2). Also, the temporal evolution of the resilience was analyzed graphically (Figure 4). From 2009–2020, the overall resilience showed a continuous growth trend. Excluding the service-oriented infrastructure resilience in 2020, the remaining production-oriented and service-oriented infrastructure resilience were higher than the overall resilience, whereas the emerging intelligent infrastructure resilience continued to grow at a lower level than the overall resilience.

Production-oriented and service-oriented infrastructure resilience demonstrated a complicated pattern of temporal evolution across time. Between 2010–2011 and 2017–2018, the resilience of production-oriented infrastructure suffered a modest dip before continuing to develop for the duration of the time and demonstrating an exponential growth trend between 2018–2020. The resilience of service-oriented infrastructure dropped in 2009–2011, 2012–2013, 2016–2017, and 2019–2020, while its growth rate changed continually in the following years. The temporal evolution of emerging intelligent infrastructure resilience resembled the overall resilience. From 2009 to 2018, both showed more consistent growth; from 2018 to 2020, both showed an exponential increase, with the new infrastructure resilience outpacing overall resilience.

Equations (3) and (4) were used to compute the coupling degree and coupling coordination degree among the three subsystems in the middle and lower sections of the TGRA based on the CCDM. They were also graded accordingly according to the grading criteria in Table 2 and Table 3. From 2009 to 2020, the coupling degree and the coupling coordination increased gradually (Figure 5). The coupling degree was at the high-level coupling stage and had been steadily growing, with a higher growth rate in 2009–2010 and 2017–2019. Except for a modest reduction from 2010 to 2011, the coupling coordination degree showed distinct growth rates throughout the rest of the period. The coupling coordination is in the reluctant coordination condition from 2009 to 2018, and the preliminary coordination state from 2019 to 2020. The coupling coordination degree rose significantly between 2018 and 2019, rising from 0.5981 (reluctant coordination) in 2018 to 0.6596 (preliminary coordination) in 2019.

4.2. Micro Level—Analysis of Spatio-Temporal Patterns of Resilience and Coupling Coordination

The same comprehensive index approach and CCDM were employed to calculate the infrastructure resilience and coupling coordination degree for each county and district (Figure 6). Significant variances exist in the spatio-temporal evolution of resilience trends in counties and districts from 2009 to 2020. From 2009 to 2020, the remaining seven counties showed varying degrees of decline, with a fluctuating growth trend, except Yiling district, which offered a constant growth trend. Wanzhou experienced a one-year decrease (2010–2011), Zigui experienced a two-year decline (2009–2010, 2012–2013), while Xingshan, Badong, Yunyang, Fengjie, and Wushan experienced three-year declines. Furthermore, each county and district experienced significant growth in 2018–2019, with Xingshan experiencing the most significant increase of 0.1023 in 2019 compared to 2018. In Yiling, Zigui, Xingshan, and Badong in Hubei province, production-oriented infrastructure resilience is often higher than service-oriented infrastructure resilience. In contrast, the latter is generally higher than the former in Wanzhou, Yunyang, Fengjie, and Wushan in Chongqing. Except for Xingshan, the remaining counties and districts showed constant or moderate growth between 2009 and 2018, transitioning to exponential growth between 2018 and 2020. Xingshan’s new infrastructure resilience was much higher than that of the remaining counties and districts from 2009 to 2016, with a significant fall in 2016–2017 and a steady increase in 2017–2020.

From 2009 to 2020, there were only four states in the counties and districts in the TGRA’s middle and lower parts: approaching imbalance state, reluctant coordination state, preliminary coordination state, and moderate coordination state, with no low-level coupling coordination (Figure 7). This demonstrates that the degree of coupling coordination among infrastructure subsystems in each county and district was broadly consistent and growing gradually. On the other hand, the coupling coordination degree displayed various complicated spatial patterns at different time phases. Only two types of approaching imbalance and reluctant coordination states occurred in the counties from 2009 to 2014, with Yiling and Xingshan increasing slightly in the reluctant coordination state. Badong and Yunyang fluctuated and grew in the approaching imbalance state, while the other counties increased from approaching imbalance to reluctant coordination. Xingshan, Wanzhou, and Yiling moved from reluctant coordination to preliminary coordination in 2015, 2016, and 2017, respectively, while the remaining counties mainly stayed within the range of reluctant coordination. In 2019 and 2020, Yiling, Xingshan, and Wanzhou went from preliminary to moderate coordination, while the remaining counties moved from reluctant to preliminary coordination.

In conclusion, despite disparities in growth rates between counties, subsystem resilience and overall resilience showed an upward trend and, starting in 2018, the growth rate increased compared to previous years, particularly in terms of new infrastructure resilience. From 2009 to 2020, all counties and districts in the TGRA enhanced their coupling coordination degree from nearing imbalance to moderate coordination. The early post-Three Gorges period (2009–2014) displayed less variety between counties and districts and lacked a more defined spatial distribution. After 2014, the development plans of the middle reaches of the Yangtze River urban agglomeration and the Chengyu urban agglomeration have been implemented. Driven by the radiation of the two urban agglomerations, the coupling coordination degree of the Yiling district in the Hubei section and Wanzhou district in the Chongqing section increased significantly.

4.3. Meso Level—Main Control Factors of Resilience and Coupling Coordination Degree

The study of the factors influencing the degree of infrastructure resilience and coupling coordination in the TGRA sections of Hubei and Chongqing is critical for determining how to improve the degree of coupling coordination among “productive-service-oriented-intelligent” infrastructure subsystems in order to achieve sustainable regional development. As a result, from 2009 to 2020, the GRA model [63] was used to calculate the grey relational degree of each component and determine the ranking of elements influencing infrastructure resilience and coupling coordination degree.

This study recognized the top ten influencing factors of the grey relational degree between the Hubei and Chongqing parts as the most critical aspects affecting infrastructure resilience (

Table 6. The main control factors of infrastructure resilience between the Hubei and Chongqing sections in 2009–2020 and their grey relational degree.

Rank	Variables	Hubei	Variables	Chongqing
1	R5	0.782	R16	0.769
2	R40	0.772	R32	0.751
3	R37	0.77	R40	0.751
4	R42	0.768	R15	0.738
5	R53	0.759	R5	0.729
6	R7	0.741	R42	0.724
7	R33	0.738	R31	0.72
8	R30	0.724	R4	0.716
9	R16	0.724	R27	0.708
10	R28	0.723	R17	0.704

). The primary factor in the Hubei section is R5 (urbanization rate), which has a grey relational degree of 0.782, while the other key factors are R40 (percentage of health expenditure), R37 (number of beds in adoption-type institutions per 10,000 people), R42 (percentage of social security expenditure), R53 (percentage of science and technology expenditure), R7 (highway road network density), R33 (number of library books per 10,000 people), R30 (number of CDCs per 10,000 people), R16 (percentage of transportation expenditure), and R28 (number of village committees per 10,000 people), all with a grey relational degree more significant than 0.7. The primary factor in the Chongqing section is R16, with a grey relational degree of 0.769. In contrast, the other vital factors are R32 (number of health technicians per 10,000 population), R40, R15 (percentage of GDP in the secondary industrial sector), R5, R42, R31 (percentage of GDP in the secondary industrial sector), R4 (geological disaster potential sites), R27 (number of township governments per 10,000 people), and R17 (percentage of expenditure for urban and rural community affairs), all with the grey relational degree more significant than 0.7. In comparison, the Hubei section’s grey relational degree of critical factors is often higher than that of the Chongqing region. Meanwhile, the essential components for both parts are R5, R40, and R42. This implies that high-quality new urbanization construction and high-quality development of livelihood construction, such as medical and social security, play an essential role in improving the resilience of urban infrastructure in the TGRA.

Table 7. The grey relational degree of the coupling coordination in Hubei and Chongqing sections from 2009 to 2020.

R	2009–2014				2015–2018				2019–2020
	Hubei	Rank	Chongqing	Rank	Hubei	Rank	Chongqing	Rank	Hubei	Rank	Chongqing	Rank
R1	0.545	28	0.637	12	0.648	15	0.677	10	0.74	8	0.765	3
R2	0.589	18	0.477	35	0.545	31	0.451	38	0.689	13	0.543	31
R3	0.473	39	0.677	6	0.502	39	0.522	31	0.554	33	0.606	22
R4	0.496	35	0.781	1	0.525	34	0.736	5	0.583	26	0.63	19
R5	0.548	26	0.490	32	0.666	12	0.594	21	0.683	14	0.664	16
R6	0.593	17	0.505	29	0.516	36	0.418	46	0.398	52	0.379	52
R7	0.461	42	0.702	4	0.566	27	0.831	1	0.604	23	0.793	2
R8	0.382	49	0.657	8	0.376	49	0.608	18	0.4	51	0.572	25
R9	0.406	48	0.637	12	0.438	47	0.637	13	0.511	40	0.522	35
R10	0.692	4	0.453	38	0.704	6	0.477	33	0.719	11	0.476	40
R11	0.455	44	0.530	25	0.452	45	0.612	17	0.501	41	0.551	30
R12	0.579	21	0.497	31	0.548	30	0.585	23	0.557	32	0.656	18
R13	0.549	25	0.472	36	0.653	14	0.542	28	0.722	9	0.497	37
R14	0.675	6	0.433	39	0.682	9	0.477	33	0.822	4	0.521	36
R15	0.599	15	0.572	20	0.641	17	0.642	12	0.474	44	0.469	42
R16	0.566	22	0.504	30	0.668	10	0.533	30	0.783	6	0.487	39
R17	0.429	46	0.531	24	0.505	37	0.574	26	0.415	48	0.554	28
R18	0.356	50	0.360	50	0.352	50	0.369	50	0.721	10	0.731	7
R19	0.486	37	0.415	41	0.494	42	0.444	42	0.566	30	0.527	34
R20	0.484	38	0.392	46	0.539	32	0.447	40	0.587	25	0.532	32
R21	0.462	41	0.549	22	0.493	43	0.580	25	0.553	34	0.664	16
R22	0.607	14	0.653	9	0.700	7	0.686	9	0.679	15	0.704	13
R23	0.759	2	0.721	3	0.738	4	0.783	4	0.628	19	0.73	8
R24	0.636	10	0.660	7	0.616	22	0.585	23	0.536	36	0.626	20
R25	0.582	20	0.482	34	0.764	1	0.602	19	0.744	7	0.76	5
R26	0.521	31	0.509	28	0.635	18	0.488	32	0.661	16	0.561	26
R27	0.491	36	0.545	23	0.498	41	0.587	22	0.511	39	0.561	26
R28	0.532	29	0.459	37	0.603	23	0.446	41	0.606	22	0.42	48
R29	0.447	45	0.391	47	0.421	48	0.420	45	0.407	50	0.435	45
R30	0.528	30	0.397	45	0.518	35	0.401	48	0.524	38	0.388	50
R31	0.641	9	0.591	17	0.762	2	0.815	3	0.84	2	0.819	1
R32	0.749	3	0.574	19	0.753	3	0.660	11	0.86	1	0.72	10
R33	0.642	8	0.406	44	0.602	24	0.465	36	0.61	21	0.464	43
R34	0.561	23	0.530	25	0.499	40	0.827	2	0.476	43	0.738	6
R35	0.771	1	0.486	33	0.727	5	0.538	29	0.532	37	0.529	33
R36	0.636	10	0.582	18	0.647	16	0.624	16	0.803	5	0.681	15
R37	0.625	12	0.630	14	0.629	19	0.421	44	0.58	27	0.443	44
R38	0.621	13	0.752	2	0.683	8	0.718	7	0.834	3	0.763	4
R39	0.681	5	0.558	21	0.504	38	0.430	43	0.484	42	0.432	46
R40	0.499	34	0.606	15	0.668	10	0.633	14	0.574	28	0.713	11
R41	0.595	16	0.645	11	0.592	26	0.706	8	0.455	47	0.725	9
R42	0.546	27	0.698	5	0.666	12	0.729	6	0.633	17	0.706	12
R43	0.516	32	0.646	10	0.449	46	0.553	27	0.548	35	0.598	23
R44	0.586	19	0.412	42	0.558	28	0.449	39	0.593	24	0.553	29
R45	0.515	33	0.594	16	0.554	29	0.599	20	0.632	18	0.682	14
R46	0.356	50	0.360	50	0.352	50	0.369	50	0.573	29	0.429	47
R47	0.466	40	0.381	49	0.594	25	0.405	47	0.619	20	0.409	49
R48	0.356	50	0.360	50	0.352	50	0.369	50	0.395	53	0.476	40
R49	0.356	50	0.360	50	0.352	50	0.369	50	0.409	49	0.368	53
R50	0.410	47	0.424	40	0.619	21	0.626	15	0.709	12	0.608	21
R51	0.459	43	0.410	43	0.468	44	0.471	35	0.47	46	0.494	38
R52	0.551	24	0.383	48	0.531	33	0.397	49	0.472	45	0.385	51
R53	0.667	7	0.529	27	0.628	20	0.460	37	0.561	31	0.58	24

shows the rankings of all grey relational degrees between the influencing elements of the Hubei and Chongqing sections from 2009–2014, 2015–2018, and 2019–2020. The critical determinants for the three time periods in the Hubei section were R35 (number of general secondary school faculty per 10,000 people), R25 (average regional ambient noise), and R32. Between 2009 and 2014, only three contributing factors had grey relational degrees more than 0.70: R35 and R23 (average annual rainfall); and R32. R25, R31, R32, R23, R35, and R10 (yearly water supply per 10,000 people). Additionally, R22 (urban sewage treatment rate) had grey relational degrees of more than 0.7 from 2015 to 2018. From 2019 to 2020, there were 12 grey relational degrees greater than 0.7, five of which were more significant than 0.8: R32, R31, and R38 (percentage of GDP in the tertiary industrial sector); R14 (number of emergency shelters per 10,000 people); and R36 (greening coverage rate of built-up areas). Furthermore, the final impact factors for both 2009–2014 and 2015–2018 were four: R18 (percentage of expenditure for disaster prevention and emergency management), R46 (number of big data centers per 10,000 people), R48 (number of 5G base stations per 10,000 people), and R49 (number of new energy vehicle charging piles per 10,000 people). On the other hand, there was only one last effect factor from 2019 to 2020: R48, with a grey correlation of 0.395.

The critical factors in the Chongqing section for the three periods were R4, R7, and R31, respectively. Between 2009 and 2014, only four grey relational degrees were more significant than 0.70 among all contributing factors: R4, R38, R23, and R7. From 2015 to 2018, there were eight grey relational degrees more significant than 0.7, with three of them above 0.8: R7, R34, and R31. In addition, 13 grey relational degrees were more significant than 0.7 from 2019 to 2020, including R31, which was greater than 0.8. Furthermore, the most recent impact factors for 2009–2014 and 2015–2018 were the same as in the Hubei section: R18, R46, R48, and R49. However, there was only one last effect factor from 2019 to 2020: R49, with a grey relational degree of 0.368.

The factors influencing the coupling coordination of infrastructure resilience in the Hubei and Chongqing sectors varied within the same period, as indicated in Table 7. The TGRA has just begun the first five-year plan in the post-Three Gorges phase, which runs from 2009 to 2014. Among these, the Hubei section’s grey relational degree of the primary factor R35 was 0.771, which ranked 33rd in the Chongqing section with a grey relational degree of 0.486; the Chongqing section’s grey relational degree of the primary factor R4 was 0.781, which ranked 35th in the Hubei section with a grey relational degree of 0.496. The post-Three Gorges period entered the second five-year plan from 2015 to 2018, and the central government began to push the effective development of the Yangtze River Economic Belt. Furthermore, the factors influencing the degree of coupling coordination of the two large metropolitan agglomerations differed significantly. The indicator R25, which ranked first in the Hubei section, had a grey relational degree of 0.602 in the Chongqing section and ranked 19th during this period; the indicator R25, which ranked first in the Chongqing section, had a grey relational degree of 0.566 in the Hubei section and ranked 27th during this period. The Yangtze River Economic Belt began to shift to high-quality development in 2019–2020, compared to 2009–2014 and 2015–2018, and the construction of an innovative and integrated infrastructure system began to accelerate, with new infrastructure-level factors playing a more significant role in enhancing coupling and coordination. For example, the Hubei and Chongqing sections’ R50 (degree of the comprehensive building of emerging intelligent infrastructure) improved from 0.410 (47th) and 0.424 (40th) in 2009-2014 to 0.709 (12th) and 0.608 (21st) in 2019–2020, respectively.

In conclusion, while the factors influencing infrastructure resilience and coupling coordination degree differ between the Hubei and Chongqing sections from 2009 to 2020, the degree of association of each component with resilience and coupling coordination typically indicates an increasing trend. Regarding the coupling coordination degree, four indicators, R23 (average annual rainfall), R31 (number of medical institution beds per 10,000 people), R32 (health technicians per 10,000 people), and R38 (percentage of GDP in the tertiary industrial sector), have all appeared more than four times in the top 10. This implies that dealing with heavy or continuous rainfall to reduce infrastructure disruptions, accelerating the construction of health care infrastructure, focusing on the coupled coordination development of urbanization and equalization of health care services, optimizing the production structure, and vigorously developing tertiary industry are all critical for improving the TGRA’s coupled and coordinated infrastructure.

5. Discussion

5.1. Innovation Potential

The resilient city has attracted extensive attention as an essential component of sustainable urban development and a frontier area of modern urbanization research. As the lifeline project of the urban system, the coupling coordination between the production-oriented infrastructure, the service-oriented infrastructure and the emerging intelligent infrastructure is a large and complex organic whole. Realizing coordinated development among the three subsystems is of tremendous theoretical and practical importance. While there are many studies on comprehensive urban resilience assessment [64,65,66], urban “natural-physical-economic-social” multi-system resilience assessment [57,67], and the coupling relationship between urban resilience and urbanization [68,69,70], there are fewer studies that exist on the coupling relationship between “production-oriented, service-oriented, emerging intelligent” infrastructure in urban physical systems. Therefore, based on the current research findings, this study selected the middle and lower sections of the TGRA as the specific region. It analyzed the interaction mechanisms between subsystems within the infrastructure system and examined the main control elements for the improvement of the multidimensional coupling resilience. This provides theoretical support for the coordinated development of the modern infrastructure system for the Yangtze River Economic Belt and, to some extent, improves the academic system of existing research.

In line with the previous study of resilience and sustainable development of critical infrastructure [71,72,73], our study builds a framework for the modern infrastructure system with multi-dimensional coupled resilience, which is more systematic, scientific and comprehensive. Moreover, it assesses the disaster-adapted resilience of the lifeline works of urban systems (i.e., “production-oriented, service-oriented, emerging intelligent” infrastructure). The study expands the methodological system for the high-quality integrated development of the Yangtze River Economic Belt and the construction of resilient cities. Also, it provides ideas and decision support for the smooth promotion of the Yangtze River Protection and the Yangtze River Economic Belt Strategy.

5.2. Limitations

However, this paper has certain drawbacks. On the one hand, this study assesses and analyzes the coupling impact relationship of production-oriented infrastructure, service-oriented infrastructure, and emerging intelligent infrastructure resilience from macro-, meso-, and micro perspectives, with some subjectivity. As a result of this assessment, future work should conduct multi-scenario perturbation simulations for various critical infrastructures and present the best coupling augmentation scenario for infrastructure system resilience. In contrast, the degree of accessibility and quality of data for each indicator vary across counties and districts, resulting in minor departures from the actual situation. For example, the notion of emerging intelligent infrastructure was just presented in 2018 and has only recently begun to be implemented systematically, resulting in decentralized construction data in 2018 not being counted. As a result, the issue of effectively using varied forms of data and establishing an ideal database to give more accurate and diverse data support must be addressed in future research on resilience evaluation and coupling coordination analysis of urban infrastructure.

6. Conclusions and Policy Recommendation

6.1. Conclusions

The primary goal of this research is to investigate the spatio-temporal characteristics of urban infrastructure resilience and coupling coordination and to investigate the direct and key factors influencing each subsystem’s level of coordinated development in order to provide scientific ideas for infrastructure construction and sustainable development. To accomplish these objectives, infrastructure was divided into “production-oriented infrastructure, service-oriented infrastructure, and emerging intelligent infrastructure,” and a coupled and coordinated resilience evaluation system for modern infrastructure was established. From 2009 to 2020, the comprehensive index method was used to calculate overall resilience and subsystem resilience in the TGRA’s middle and lower parts, as well as to examine the temporal evolution of resilience. Then, using the CCDM, the temporal evolution characteristics of its coupling degree and coupling coordination degree were analyzed; from the perspective of the counties and districts, the coupling coordination degree was calculated and classified from 2009 to 2020, and the changes in its spatial pattern were also analyzed. Finally, the critical control components affecting infrastructure resilience and coupling coordination in the Hubei and Chongqing sections were compared using GRA.

The following are the key conclusions. (1) At the macro-level, both the overall resilience and the following subsystem resilience are key; however, there were significant disparities in annual growth rates, with the overall resilience increasing more in 2018–2020 than in any other period. Meanwhile, from 2009 to 2020, its coupling degree and coupling coordination degree increased gradually, with the coupling coordination degree improving from the reluctant coordination stage to the preliminary coordination stage.

(2) At the micro-level, there was substantial regional heterogeneity in the evolution of resilience tendencies; for each county and district, there was a growing trend. From 2010 to 2013, the spatial pattern of coupling coordination degree did not display a more noticeable spatial differentiation. However, from 2014 to 2020, driven by the radiation of the two major urban agglomerations, the resilience and coupling coordination of Yiling and Wanzhou both show a trend of more substantial increase, and the rest of the counties have a small increase.

(3) At the meso-level, the correlation of each component with resilience and coupling coordination usually indicated a rising trend. Still, there were discrepancies in their key control variables due to the unequal development of the Yangtze River Economic Belt. Meanwhile, seven factors have a more significant impact on the coupled and coordinated development of urban infrastructure than other indicators, including urbanization rate, average annual rainfall, the number of medical institution beds per 10,000 people, the number of health technicians per 10,000 people, and the percentage of GDP in the tertiary industrial sector.

6.2. Policy Recommendation

Based on these results, our relevant recommendations are as follows.

(1) In the face of uncertain disturbances, we need to research the disaster-adapted resilience of modern infrastructure systems in cities and towns. On the one hand, the government should provide appropriate policy and financial support as well as promote the construction of demonstration zones; on the other hand, scholars should conduct more comprehensive and scientific academic research from multiple perspectives and disciplines to promote theoretical and methodological innovation of modern infrastructure systems.

(2) Comprehensively promote the high-quality integrated development of the Yangtze River Economic Belt, break down administrative barriers and achieve regional integration and development. The study shows that the areas with high coupling coordination are located in the Wanzhou and Yiling districts, which are dependent on the development of the main urban areas. Nevertheless, the areas of low coupling coordination are in ecologically sensitive areas such as Wushan and Badong counties, where fragile geological conditions affect infrastructure development. Given the regional differences in the middle and lower reaches of the TGRA, we should analyze in depth the natural geographical conditions and the current urbanization situation, clarify the relevant influencing factors, and take corresponding measures in a targeted manner. More importantly, Hubei province and Chongqing city should coordinate with each other regarding policies, funding, and projects for sustainable development.

(3) In response to the different main control factors between the Chongqing and the Hubei section, the government should scientifically adjust its policies on the development of various infrastructures, taking the resource-carrying capacity and socioeconomic development levels of different regions into account. For example, in the Hubei section, the key factors of infrastructure resilience and coupling coordination are primarily service-oriented infrastructure, and the policies may be more inclined toward the construction of livelihoods such as education, medical care, and social security. Whereas in the Chongqing section, the key factors are primarily production-oriented infrastructure, and the policies may be more inclined toward the construction of spatial facilities such as transportation and water supply. Therefore, as the connecting link between the urban agglomeration in the middle reaches of the Yangtze River and the Chengyu urban agglomeration, how to vigorously promote the integrated development of the Yangtze River Economic Belt and expand the scope of the radiation zone of the two major urban agglomerations in the TGRA, plays an essential role in comprehensively enhancing infrastructure resilience and promoting sustainable development.