A multi-agent collaborative demand-oriented urban parking governance evaluation and grading method

Abstract

The application provides a kind of urban parking management evaluation and grading method for multi-agent collaborative demand, constructs V, P, O, S, R, M six dimensions, 18 fine index system, covers resource state, operation pressure, system vitality, risk resistance, subject organization, management ability whole chain, fits multi-agent collaborative governance appeal, promotes parking management from single point evaluation to system global evaluation, the evaluation result is more comprehensive, more close to the real governance scene of city. At the same time, the index and strategy system is built with the multi-agent collaborative demand of managers, operators and users as the bottom logic, so that the evaluation is derived from the multi-agent demand, and the strategy serves the multi-agent landing, solving the problem of fragmented rights and responsibilities, mismatched measures and poor collaboration in traditional governance.

Description

Technical Field

[0001] This invention relates to the fields of urban traffic management, smart parking, multi-source data analysis, and refined urban management, and in particular to an evaluation and classification method for urban parking management that addresses the collaborative needs of multiple stakeholders. Background Technology

[0002] With the continuous growth of urban motor vehicle ownership, numerous governance challenges in urban parking have become increasingly prominent, manifested in a series of practical problems such as a significant gap between parking supply and demand, an imbalance in the spatial distribution of parking facilities, unscientific planning and setting of on-street parking, low utilization rate of smart parking data resources, lack of transparency in parking fee information disclosure, and poor parking experience for the public. Current mainstream parking governance evaluation systems primarily rely on single static indicators such as the total number of parking spaces, the scale of parking lot construction, and the static supply-demand ratio, which cannot accurately adapt to the differentiated governance needs of different urban areas and participating entities, resulting in significant limitations in evaluation dimensions.

[0003] Currently, smart parking systems have generally achieved basic functions such as parking space inquiry, parking guidance, and automated fee management. However, the system design often focuses on single service scenarios and has not fully considered the collaborative development needs of multiple stakeholders, including government management departments, parking operators, and parking users. Existing systems struggle to integrate multi-dimensional information such as parking resource allocation efficiency, facility operation status, safety and compliance levels, public service quality, and user experience into standardized, quantifiable, and calculable evaluation results. Furthermore, existing parking governance evaluation methods mostly employ a single weighting model or a simple scoring and ranking approach, lacking a scientific hierarchical decision-making mechanism and a precise identification mechanism for governance shortcomings. This makes it difficult to accurately adapt to the refined and differentiated governance needs of various typical governance units, such as 15-minute living circles, hospitals, commercial clusters, transportation hubs, and older residential areas, thus hindering the effective implementation of precise urban parking governance.

[0004] Therefore, it is necessary to propose a method for identifying and classifying urban parking governance demand characteristics by integrating multi-source parking data, multi-subject demand indicators, combined weighting, and hierarchical evaluation models, so as to provide technical support for prioritizing parking governance, diagnosing shortcomings, and formulating collaborative governance strategies. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides an evaluation and grading method for urban parking management that addresses the collaborative needs of multiple stakeholders.

[0006] To achieve the above objectives, the technical solution of this invention is as follows: In a first aspect, the present invention provides an evaluation and classification method for urban parking management that addresses the collaborative needs of multiple stakeholders, the method comprising: Obtain the urban area to be governed and divide it into multiple governance units; identify the relevant stakeholders in parking governance and construct a subject set including managers, operators and users; Based on the subject set and six dimensions—Vitality (V), Stress (P), Organization (O), State (S), Resilience (R), and Management (M)—a VPOSRM indicator system containing 18 parking governance demand indicators is constructed. The process involves acquiring multi-source heterogeneous raw data from various governance units and standardizing it to generate a standardized indicator matrix. The multi-source heterogeneous raw data includes parking resource survey data, smart parking platform operation data, on-site verification data, GIS spatial data, and satisfaction survey data. The three types of weight vectors for each indicator in the standardized indicator matrix are calculated using the projection pursuit method, the analytic hierarchy process (AHP), and the entropy method, respectively. Based on the preset combined weight coefficients, the calculated three types of weight vectors are combined and weighted to obtain the comprehensive weight of each indicator. Based on the comprehensive weight and the standardized indicator matrix, the parking management demand characteristic index of each governance unit is calculated. Based on the parking management demand characteristic index, the rank-sum ratio method is used to rank each management unit, and the grading threshold is calculated by the Probit regression model to divide each management unit into different management levels. According to the weight ranking of the comprehensive weight of each indicator and the management level of each management unit, collaborative management strategy texts corresponding to managers, operators and users are generated respectively.

[0007] In some embodiments, the 18 parking management requirements indicators include at least the following: indicators for managers: number of parking spaces per vehicle, reasonable level of on-street parking space allocation, density of public parking lots, reasonable level of parking facility structure, level of identification of parking conflict areas, and optimal utilization rate of parking resources; indicators for operators: parking space occupancy rate, fire safety compliance rate, level of smart parking services, parking fee arrears rate, completeness rate of parking signs and markings, and satisfaction with parking service quality; and indicators for users: level of convenience of parking and travel, completeness rate of parking and charging facilities, level of parking space design, level of smart parking information services, level of smart parking entrance and exit design, and completeness and transparency of parking fee information.

[0008] In some embodiments, the calculation formula for the manager's indicator is as follows: The calculation formula for the average number of parking spaces per vehicle is: The calculation formula for the average number of parking spaces per vehicle is: In the formula, The number of parking spaces per vehicle; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The formula for calculating the reasonable level of on-street parking spaces is as follows: (Based on the number of motor vehicles) In the formula, Set a reasonable level (%) for on-street parking spaces; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles; the formula for calculating the distribution density of public parking lots is: In the formula, Public parking lot distribution density; Number of off-street public parking lots available; Provide the number of public parking lots within the roadside; The formula for calculating the reasonable level of parking facility structure is: (The area is the built-up area). In the formula, To ensure the rational level of parking facility structure; The proportion of parking facilities to be built; The proportion of public parking facilities in the city; The percentage of on-street parking facilities; The total supply of parking facilities is the sum of the supply of parking facilities in buildings, public parking facilities in cities, and on-street parking facilities. Number of parking facilities to be provided; The number of public parking facilities available in the city; The formula for calculating the number of on-street parking facilities and the level of identification of parking conflict areas is as follows: In the formula, Level of identification for parking conflict areas; The number of evaluation items identified; the formula for calculating the optimal utilization rate of parking resources is: In the formula, D ZL To optimize the utilization rate of parking resources; R J For smart parking platform access rate; R C For time-sharing rate; N J The number of parking lots connected to the smart parking platform; N C This refers to the number of shared parking lots during off-peak hours.

[0009] In some embodiments, the formula for calculating the operator's metrics is as follows: The formula for calculating parking space occupancy rate is: In the formula, Parking space occupancy rate; This represents the actual number of vehicles occupied in the parking lot. This represents the total number of parking spaces in the parking lot; the formula for calculating the average number of parking spaces per vehicle is: In the formula, To improve fire safety compliance rate; The number of projects that meet fire safety standards; the formula for calculating the level of smart parking services is: In the formula, To improve the level of smart parking services; , , The smart parking service levels are respectively , , The number of parking lots; The number of parking lots available in the area; the formula for calculating the parking fee arrears rate is: In the formula, Parking fee arrears rate; Unpaid parking fees; The formula for calculating the completeness rate of parking signs and markings is as follows: (This is to account for parking fees.) In the formula, The completeness rate of parking signs and markings; The actual number of items that meet the verification standards; The total number of items to be verified; the formula for calculating parking service quality satisfaction is: In the formula, For service quality satisfaction; Number of satisfied users; This represents the total number of users surveyed.

[0010] In some embodiments, the formula for calculating user-specific indicators is as follows: The formula for calculating the level of parking and travel convenience is: In the formula, To improve the convenience of parking and travel; The calculation formulas are as follows: Actual walking distance; parking and charging facility completeness rate, calculated based on the proportion of charging spaces and facility dimensions; parking space design level, calculated based on garage building design specifications; and the level of smart parking information services. In the formula, Information acquisition index; To verify the number of approved items; the design level of smart parking entrances and exits, based on the display of available parking spaces, payment functions, and navigation push, the calculation formula for the completeness and transparency of parking fee information is as follows: In the formula, To ensure complete transparency of parking fee information; To meet the number of evaluation items required for the evaluation content.

[0011] In some embodiments, the parking management demand characteristic index is calculated through the following steps: standardizing the positive and negative indicators in the multi-source heterogeneous raw data to obtain a standardized indicator matrix; using the projection pursuit method to reduce the dimensionality of the standardized indicator matrix to obtain the optimal projection direction vector; constructing a judgment matrix using the analytic hierarchy process (AHP) to calculate the subjective weight vector; calculating the information entropy of each indicator using the entropy method and determining the objective weight vector based on the difference coefficient; linearly weighting the three types of weight vectors based on preset combined weight coefficients to obtain the comprehensive weight of each indicator; wherein the preset combined weight coefficients are the projection pursuit weight coefficient α, the AHP weight coefficient β, and the entropy weight coefficient γ, and satisfy α+β+γ=1; and weighting the standardized indicator matrix with the corresponding comprehensive weights to obtain the parking management demand characteristic index of each management unit.

[0012] In some embodiments, based on the parking management demand characteristic index, the rank-sum ratio method is used to rank each management unit, and the classification threshold is calculated using a Probit regression model to divide each management unit into different management levels. This includes: constructing an evaluation matrix using the parking management demand characteristic index as the evaluation index value; performing a rank transformation on the evaluation matrix to generate a rank matrix; calculating the weighted rank-sum ratio of each management unit based on the rank matrix; determining the distribution of the weighted rank-sum ratio, calculating the downward cumulative frequency corresponding to each weighted rank-sum ratio value, and converting the downward cumulative frequency into a probability unit Probit value; performing linear regression with the probability unit Probit value as the independent variable and the weighted rank-sum ratio as the dependent variable to construct a regression equation; determining the Probit value interval corresponding to each level based on the standard normal deviation according to the preset number of levels, substituting it into the regression equation to calculate the corresponding weighted rank-sum ratio estimate; and using the weighted rank-sum ratio estimate as the classification threshold to classify each management unit into different levels to obtain the management level.

[0013] In some embodiments, collaborative governance strategy texts corresponding to managers, operators, and users are generated respectively: high-weight indicators with weights greater than a preset weight threshold in the comprehensive weight are matched with low-rank indicators with ranks lower than a preset rank threshold in the rank matrix. If an indicator is both high-weight and low-rank, it is determined as the priority weakness indicator of the corresponding governance unit. Based on the governance level and priority weakness indicator, collaborative governance strategy texts are called from a preset strategy library and generated.

[0014] This invention provides an evaluation and grading method for urban parking governance that addresses the collaborative needs of multiple stakeholders. It constructs a refined indicator system encompassing six dimensions (V, P, O, S, R, M) and 18 sub-indicators, covering the entire chain of resource status, operational pressure, system vitality, resilience, stakeholder organization, and management capabilities. This system aligns with the demands of multi-stakeholder collaborative governance, propelling parking governance from single-point evaluation to a system-wide evaluation, resulting in more comprehensive evaluations that better reflect real-world urban governance scenarios. Furthermore, it builds an indicator and strategy system based on the collaborative needs of managers, operators, and users, ensuring that evaluations originate from multi-stakeholder needs and strategies serve multi-stakeholder implementation. This addresses the problems of fragmented responsibilities, mismatched policies, and poor collaboration among stakeholders in traditional governance. Regarding empowerment, traditional technologies often employ singular subjective or objective empowerment methods, which are prone to strong subjective bias or detachment from actual business realities. This method integrates three algorithms: projection pursuit high-dimensional feature mining, analytic hierarchy process (AHP) empirical weighting, and entropy method data difference weighting. By combining coefficients, it achieves a deep fusion of subjective and objective approaches, respecting the business rules of the parking management industry while fully mining the inherent characteristics of the data. This results in a higher degree of matching between indicator weights and management needs, and stronger reliability of index calculation results. This invention eliminates extreme value interference through the rank-sum ratio method and adaptively generates tiering thresholds using a Probit regression model, achieving standardized tiering driven by all data and without human intervention. The tiers are balanced, with clear boundaries and stable results, adaptable to the refined tiered management needs of different cities and regions. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the urban parking management evaluation and grading method for multi-entity collaborative needs provided in this embodiment of the invention. Figure 2 This is a flowchart illustrating another urban parking management evaluation and grading method for multi-entity collaborative needs provided by an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] In the following description, references to "some embodiments" refer to a subset of all possible embodiments; however, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. Unless otherwise defined, all technical and scientific terms used in the embodiments of the invention have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments of the invention pertain. The terminology used in the embodiments of the invention is for the purpose of describing the embodiments of the invention only and is not intended to limit the invention.

[0018] This invention provides an evaluation and classification method for urban parking management that addresses the collaborative needs of multiple stakeholders. (See also...) Figure 1 , Figure 1 This is a flowchart illustrating an urban parking management evaluation and grading method for multi-stakeholder collaborative needs, provided by an embodiment of the present invention. Figure 1 The steps shown are explained.

[0019] S110: Obtain the urban area to be governed and divide it into multiple governance units; identify the relevant entities in parking governance and construct a set of entities including managers, operators and users.

[0020] Here, a governance unit refers to the smallest independent unit for governance statistics and evaluation obtained after dividing the urban area to be governed into grids and fragments according to rules such as urban area spatial boundaries, parking resource distribution characteristics, administrative jurisdiction, and road network structure. It is the basic calculation unit for data collection, indicator calculation, level classification, and strategy generation in this invention. For example, independent areas divided by urban street jurisdiction, community grids, 1km×1km geographical grids, and business districts, as well as a single community, a single business district, or a single main road control area, can all be considered as a governance unit.

[0021] Here, the relevant stakeholders who directly participate in and influence the effectiveness of urban parking management, or are affected by the results of parking management, are precisely limited in this invention to include only three core stakeholders: managers, operators, and users, excluding irrelevant third-party stakeholders.

[0022] Here, "managers" refers to government and public institutions responsible for regulating the urban parking industry, formulating policies, controlling order, conducting performance evaluations, and coordinating public resources. They are the main regulatory and leading entities in parking management. In practical applications, operators can be regulatory units such as the urban transportation bureau, urban management law enforcement bureau, public security traffic management department, street offices, and urban governance command centers.

[0023] Here, "operator" refers to a market-oriented enterprise or organization that undertakes the market-based operation of urban parking resources, the operation and maintenance of smart parking systems, parking space management, fee collection services, and equipment maintenance. They are the implementing entities for the practical operation of parking resources. In real-world applications, operators can be urban smart parking operation companies, commercial district parking lot operation companies, residential property parking operation and maintenance departments, or on-street parking management and operation organizations.

[0024] Here, users refer to the general public who use various urban parking resources on a daily basis, including private car owners and drivers of commercial vehicles. They are the end audience, experience subjects, and evaluation subjects of parking management services.

[0025] In this invention, the urban area to be governed is first divided into units to determine the smallest evaluation calculation unit; at the same time, all core entities involved in governance are defined to lock in the main objects for subsequent indicator construction and strategy output.

[0026] S120, based on the subject set and six dimensions of vitality (V), pressure (P), organization (O), state (S), resilience (R), and management (M), constructs a VPOSRM indicator system containing 18 parking management demand indicators.

[0027] Here, the VPOSRM indicator system refers to the exclusive parking management demand evaluation indicator system independently constructed by this invention. It is derived from the first letters of the six core evaluation dimensions in English, namely Vitality (V), Pressure (P), Organization (O), State (S), Resilience (R), and Management (M). The system contains a total of 18 subdivided parking management demand indicators.

[0028] Here, "vitality" refers to the activity level, circulation, and supporting vitality of parking resources within the governance unit, reflecting the dynamic potential for the use of parking resources. Examples include the average daily turnover rate of parking spaces, peak parking utilization rate, and the activation rate of idle parking spaces.

[0029] Pressure refers to the supply and demand pressure, traffic pressure, and management pressure on parking resources within the governance unit, reflecting the pain points of regional parking governance. Examples include the parking space supply-demand gap rate, the number of illegal parkings during peak hours, and the duration of parking congestion.

[0030] The organization refers to the collaborative mechanisms, resource allocation, and overall coordination capabilities of multiple stakeholders in parking management, reflecting the systematization level of the management work. Examples include the frequency of multi-stakeholder collaboration, the rationality of parking resource allocation, and the completeness of the management system.

[0031] Status refers to the real-time condition of parking resources, parking order, and service level within the governance unit, reflecting the quality of the basic governance status. Examples include the number of compliant parking spaces, parking complaint rate, and facility integrity rate.

[0032] Resilience refers to the ability of a regional parking system to self-adjust and recover in the face of sudden problems such as congestion, resource shortages, and equipment failures, reflecting the level of risk resistance of the system. Examples include the efficiency of managing sudden traffic flow, the capacity to allocate emergency parking spaces, and the time required to repair faults.

[0033] Management refers to the comprehensive management level of managers' oversight capabilities and operators' operational and maintenance service capabilities, reflecting the control effectiveness at the governance execution end. Examples include the efficiency of handling illegal parking, the platform's operational and maintenance qualification rate, and the completion rate of rectification implementation.

[0034] Based on the three types of collaborative needs identified above, and combined with six core evaluation dimensions, this invention establishes an 18-item exclusive parking management demand indicator system, namely the VPOSRM indicator system.

[0035] S130: Acquire multi-source heterogeneous raw data from each governance unit and perform standardization processing to generate a standardized indicator matrix; among which, the multi-source heterogeneous raw data includes parking resource census data, smart parking platform operation data, on-site verification data, GIS spatial data and satisfaction survey data.

[0036] Here, standardization refers to the process of converting various types of raw data into standardized data with the same numerical range and evaluation criteria through data cleaning and transformation methods such as normalization, dimensionless transformation, outlier removal, and missing value repair, which are used to address the problems of inconsistent dimensions, large differences in numerical ranges, and mixed positive and negative indicators in multi-source heterogeneous data.

[0037] Here, the standardized indicator matrix is ​​a two-dimensional data matrix formed by filling the standardized indicator values ​​of all governance units with the governance units as rows and the 18 parking governance demand indicators as columns. It is the core data carrier for all subsequent algorithm calculations.

[0038] S140, the three types of weight vectors of each indicator in the standardized indicator matrix are calculated using the projection pursuit method, the analytic hierarchy process, and the entropy method, respectively; based on the preset combined weight coefficients, the calculated three types of weight vectors are combined and weighted to obtain the comprehensive weight of each indicator; based on the comprehensive weight and the standardized indicator matrix, the parking management demand characteristic index of each governance unit is calculated.

[0039] Here, the weight vector is a one-dimensional array consisting of the weight values ​​of the 18 indicators, each calculated by a single algorithm. Each vector contains the single weight of all indicators and corresponds to the three independent weight results of the projection pursuit method, the analytic hierarchy process, and the entropy method.

[0040] Here, the preset combined weight coefficient refers to the preset ratio coefficient used to balance the weight results of the three types of single algorithms. It is a fixed empirical coefficient or training optimization coefficient, used to integrate subjective weights and objective weights, and avoid the limitations of a single algorithm.

[0041] Here, the comprehensive weight refers to the final and unique weight value of each indicator obtained by combining the weight coefficients of the three single weight vectors, taking into account both the objective characteristics of the data and the subjective experience of governance.

[0042] Here, the parking management demand characteristic index is a comprehensive value calculated by weighting standardized indicator values ​​and comprehensive weights. It is used to quantitatively characterize the overall parking management demand intensity and the degree of management pain points of a single management unit. The higher the index value, the more urgent the parking management demand and the more prominent the management problem in that unit.

[0043] In this invention, based on the standardized index matrix generated by S130, independent weight vectors are calculated using three classic algorithms, and then combined and weighted using preset coefficients to obtain the comprehensive weight of the index; finally, the comprehensive weight and the standardized matrix are combined to calculate the governance demand characteristic index of each unit.

[0044] S150, based on the parking management demand characteristic index, uses the rank-sum ratio method to rank each management unit, and calculates the grading threshold through the Probit regression model to divide each management unit into different management levels; according to the weight ranking of the comprehensive weight of each indicator and the management level of each management unit, it generates collaborative management strategy texts corresponding to managers, operators and users respectively.

[0045] Here, the rank-sum ratio method in this invention transforms the index data of each governance unit into ranks and calculates the rank-sum ratio value, thereby achieving an ordered ranking of the governance demand intensity of multiple governance units. It has the advantages of simple calculation, accurate ranking, and adaptability to comparison of multiple units.

[0046] The Probit regression model in this invention is used to fit the correspondence between the rank-sum ratio ranking results and the governance level. Through probability threshold calculation, it automatically divides the multi-group classification threshold, i.e. the classification threshold, to realize the automatic classification of governance units.

[0047] Here, the grading threshold is used to distinguish the critical index values ​​of different governance levels. The demand characteristic index of different ranges corresponds to different governance levels and is a quantitative judgment standard for governance grading.

[0048] Here, governance level refers to the unit governance level divided according to the urgency of governance needs and the severity of problems. It is used to distinguish different governance priorities such as key governance, routine governance, and optimization and improvement, and can provide a basis for differentiated policies.

[0049] Here, the collaborative governance strategy text is the structured governance solution text that is specifically adapted to the rights and responsibilities of the three types of subjects in the final output of this invention. It identifies the core influencing indicators based on the weight of the indicators and, combined with the unit governance level, generates differentiated and implementable collaborative governance countermeasures for managers, operators, and users respectively.

[0050] In this invention, the parking management demand characteristic index obtained from S140 is used to sort the parking management units by rank sum ratio and divide them by threshold using a Probit regression model to complete the classification of management units. Finally, the collaborative management strategy text is generated by combining the index weight sorting and the unit management level.

[0051] In some embodiments, the 18 parking management requirements indicators include at least the following: indicators for managers: number of parking spaces per vehicle, reasonable level of on-street parking space allocation, density of public parking lots, reasonable level of parking facility structure, level of identification of parking conflict areas, and optimal utilization rate of parking resources; indicators for operators: parking space occupancy rate, fire safety compliance rate, level of smart parking services, parking fee arrears rate, completeness rate of parking signs and markings, and satisfaction with parking service quality; and indicators for users: level of convenience of parking and travel, completeness rate of parking and charging facilities, level of parking space design, level of smart parking information services, level of smart parking entrance and exit design, and completeness and transparency of parking fee information.

[0052] In some embodiments, the calculation formula for the manager's indicator is as follows: The calculation formula for the average number of parking spaces per vehicle is: The calculation formula for the average number of parking spaces per vehicle is: In the formula, The number of parking spaces per vehicle; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The formula for calculating the reasonable level of on-street parking spaces is as follows: (Based on the number of motor vehicles) In the formula, Set a reasonable level (%) for on-street parking spaces; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles; the formula for calculating the distribution density of public parking lots is: In the formula, Public parking lot distribution density; Number of off-street public parking lots available; Provide the number of public parking lots within the roadside; The formula for calculating the reasonable level of parking facility structure is: (The area is the built-up area). In the formula, To ensure the rational level of parking facility structure; The proportion of parking facilities to be built; The proportion of public parking facilities in the city; The percentage of on-street parking facilities; The total supply of parking facilities is the sum of the supply of parking facilities in buildings, public parking facilities in cities, and on-street parking facilities. Number of parking facilities to be provided; The number of public parking facilities available in the city; The formula for calculating the number of on-street parking facilities and the level of identification of parking conflict areas is as follows: In the formula, Level of identification for parking conflict areas; The number of evaluation items identified; the formula for calculating the optimal utilization rate of parking resources is: In the formula, D ZL To optimize the utilization rate of parking resources; R J For smart parking platform access rate; R C For time-sharing rate; N J The number of parking lots connected to the smart parking platform; N C This refers to the number of shared parking lots during off-peak hours.

[0053] In some embodiments, the formula for calculating the operator's metrics is as follows: The formula for calculating parking space occupancy rate is: In the formula, Parking space occupancy rate; This represents the actual number of vehicles occupied in the parking lot. This represents the total number of parking spaces in the parking lot; the formula for calculating the average number of parking spaces per vehicle is: In the formula, To improve fire safety compliance rate; The number of projects that meet fire safety standards; the formula for calculating the level of smart parking services is: In the formula, To improve the level of smart parking services; , , The smart parking service levels are respectively , , The number of parking lots; The number of parking lots available in the area; the formula for calculating the parking fee arrears rate is: In the formula, Parking fee arrears rate; Unpaid parking fees; The formula for calculating the completeness rate of parking signs and markings is as follows: (This is to account for parking fees.) In the formula, The completeness rate of parking signs and markings; The actual number of items that meet the verification standards; The total number of items to be verified; the formula for calculating parking service quality satisfaction is: In the formula, For service quality satisfaction; Number of satisfied users; This represents the total number of users surveyed.

[0054] In some embodiments, the formula for calculating user-specific indicators is as follows: The formula for calculating parking and travel convenience level is: In the formula, To improve the convenience of parking and travel; The calculation formulas are as follows: Actual walking distance; parking and charging facility completeness rate, calculated based on the proportion of charging spaces and facility dimensions; parking space design level, calculated based on garage building design specifications; and the level of smart parking information services. In the formula, Information acquisition index; To verify the number of approved items; the design level of smart parking entrances and exits, based on the display of available parking spaces, payment functions, and navigation push, the calculation formula for the completeness and transparency of parking fee information is as follows: In the formula, To ensure complete transparency of parking fee information; To meet the number of evaluation items required for the evaluation content.

[0055] In some embodiments, the parking management demand characteristic index is calculated through the following steps: standardizing the positive and negative indicators in the multi-source heterogeneous raw data to obtain a standardized indicator matrix; using the projection pursuit method to reduce the dimensionality of the standardized indicator matrix to obtain the optimal projection direction vector; constructing a judgment matrix using the analytic hierarchy process (AHP) to calculate the subjective weight vector; calculating the information entropy of each indicator using the entropy method and determining the objective weight vector based on the difference coefficient; linearly weighting the three types of weight vectors based on preset combined weight coefficients to obtain the comprehensive weight of each indicator; wherein the preset combined weight coefficients are the projection pursuit weight coefficient α, the AHP weight coefficient β, and the entropy weight coefficient γ, and satisfy α+β+γ=1; and weighting the standardized indicator matrix with the corresponding comprehensive weights to obtain the parking management demand characteristic index of each management unit.

[0056] Here, positive indicators refer to evaluation indicators in the VPOSRM parking management indicator system where a larger value indicates a better parking management status, a higher management level, and fewer negative problems. These are benefit-type indicators and are gain-type indicators in the parking management evaluation system.

[0057] Negative indicators refer to evaluation indicators in the VPOSRM parking management indicator system where a larger value indicates a more prominent parking management problem, a worse management level, and greater supply and demand pressure. These are cost-based indicators and are loss-related indicators in the parking management evaluation system.

[0058] It should be noted that dimensionality reduction refers to extracting projection features from high-dimensional matrix data containing 18 indicators and dozens of governance units through projection dimensionality reduction, which can comprehensively represent the differences in parking governance of each unit, in order to avoid the problem of overlapping interference of multiple indicators.

[0059] Here, the optimal projection direction vector refers to the optimal weight vector obtained by iteratively solving the projection pursuit method. It is the projection direction that can maximize the mining of the internal differences of the indicator data of each governance unit and maximize the preservation of the effective information of the original data. It serves as the objective weight basis for the projection pursuit method and can accurately reflect the inherent contribution of each indicator data.

[0060] Here, the subjective weight vector is normalized after solving for eigenvalues ​​and performing consistency checks on the judgment matrix constructed by the analytic hierarchy process. The resulting one-dimensional weight array is generated entirely based on governance experience, expert knowledge, and industry standards, reflecting the importance of each indicator in the subjective human judgment.

[0061] Here, information entropy is used to measure the dispersion and information content of individual indicator data. The higher the dispersion of indicator data, the smaller the information entropy value, which means that the indicator contains more effective distinguishing information and has higher reference value for governance evaluation; conversely, the less effective information there is, the lower the evaluation value.

[0062] Here, the difference coefficient is a correction coefficient calculated based on the information entropy of the indicator. It is used to amplify the weight of indicators with high discrimination and weaken the weight of homogeneous indicators. It is the core correction parameter for generating objective weights using the entropy method and can accurately highlight the differentiated evaluation value of indicators.

[0063] Here, the objective weight vector is a weight array obtained by calculating and normalizing the index information entropy and difference coefficient. It is generated entirely based on the objective distribution characteristics of the original data, without any subjective human intervention, and truly reflects the evaluation information contained in the data itself.

[0064] This invention specifically distinguishes between positive gain and negative loss indicators and standardizes them separately, completely avoiding the calculation distortion caused by inconsistent dimensions and mixed positive and negative attributes of different indicators in multi-source heterogeneous data. Simultaneously, it combines three core algorithms to complete weight calculation: projection pursuit method for dimensionality reduction of high-dimensional data, uncovering deep-seated differential features and solving the problem of high-dimensional redundancy interference from multiple indicators; analytic hierarchy process (AHP) to incorporate industry governance experience, ensuring that weights align with actual parking management business logic and avoiding pure data calculations divorced from engineering applications; and entropy method for weighting based on data dispersion, truly restoring the objective distinguishing value of indicators. This completely solves the dual technical drawbacks of traditional solutions that rely solely on subjective weighting, leading to human bias, or rely solely on objective weighting, detached from actual governance.

[0065] In some embodiments, based on the parking management demand characteristic index, the rank-sum ratio method is used to rank each management unit, and the classification threshold is calculated using a Probit regression model to divide each management unit into different management levels. This includes: constructing an evaluation matrix using the parking management demand characteristic index as the evaluation index value; performing a rank transformation on the evaluation matrix to generate a rank matrix; calculating the weighted rank-sum ratio of each management unit based on the rank matrix; determining the distribution of the weighted rank-sum ratio, calculating the downward cumulative frequency corresponding to each weighted rank-sum ratio value, and converting the downward cumulative frequency into a probability unit Probit value; performing linear regression with the probability unit Probit value as the independent variable and the weighted rank-sum ratio as the dependent variable to construct a regression equation; determining the Probit value interval corresponding to each level based on the standard normal deviation according to the preset number of levels, substituting it into the regression equation to calculate the corresponding weighted rank-sum ratio estimate; and using the weighted rank-sum ratio estimate as the classification threshold to classify each management unit into different levels to obtain the management level.

[0066] Here, the evaluation matrix is ​​a two-dimensional data matrix constructed using each governance unit as a sample and the parking governance demand characteristic index as the sole evaluation indicator. If the area to be evaluated contains 20 governance units, a 20-row, 1-column evaluation matrix is ​​constructed, with each row's value corresponding to the parking governance demand characteristic index of one governance unit, thus centrally summarizing the quantitative results of governance demand for all units.

[0067] Here, rank transformation refers to the process of sorting the original index values ​​of continuous distribution in the evaluation matrix according to their numerical order and replacing the original values ​​with the corresponding rank numbers. This process can weaken the interference of extreme outliers on the classification results and preserve the relative differences between samples.

[0068] Here, the weighted rank-sum ratio is a statistical evaluation value calculated based on the rank matrix and the comprehensive weight of the indicators. It incorporates the weight differences of each governance indicator in the previous period, which not only preserves the differences in sample ranking, but also reflects the influence weight of the core governance indicators. It is one of the statistical quantities that characterizes the comprehensive governance demand level of a single governance unit.

[0069] Here, the downward cumulative frequency is the weighted rank-sum ratio statistical distribution for all governance units. It is the proportion of the cumulative sample size from the minimum value to the current value to the total sample size. It is used to reflect the overall proportion of governance units at or below a certain rank-sum ratio level and serves as an intermediate parameter connecting the rank-sum ratio and the Probit probability conversion.

[0070] Here, the probability unit Probit value is a normal deviation transformation value obtained by converting the cumulative frequency (probability value in the 0-1 interval) based on the standard normal distribution. It can transform the nonlinear frequency distribution into a linear regressible continuous value, solving the problem of unclear boundaries and difficulty in accurately classifying nonlinear distributions in traditional grading.

[0071] This invention directly uses the original index values ​​for classification. First, it converts absolute values ​​into relative ranks through rank transformation, thereby reducing the interference from extremely high or low scores of individual governance units. This avoids the problems of traditional direct threshold classification being susceptible to noise and exhibiting large fluctuations in classification results, thus improving the stability and robustness of parking governance level classification. Furthermore, it employs a weighted rank-sum ratio calculation, incorporating the comprehensive weight differences from the previous VPOSRM indicator system. This allows core governance indicators (i.e., high-weight indicators) to play a stronger dominant role in the classification results, while the role of secondary indicators is appropriately weakened. The resulting classification results are no longer simply mathematical rankings, but rather professional results that truly reflect the weight distribution of parking governance pain points, better aligning with the logic of refined urban governance.

[0072] In some embodiments, generating collaborative governance strategy texts corresponding to managers, operators, and users respectively includes: matching high-weight indicators with weights greater than a preset weight threshold in the comprehensive weight with low-rank indicators with ranks lower than a preset rank threshold in the rank matrix; if an indicator is both high-weight and low-rank, then the indicator is determined as the priority weakness indicator of the corresponding governance unit; and calling and generating collaborative governance strategy texts from a preset strategy library according to the governance level and priority weakness indicators.

[0073] The following will describe an exemplary application of the embodiments of the present invention in a practical application scenario.

[0074] This embodiment provides another method for evaluating and classifying urban parking management to meet the collaborative needs of multiple stakeholders, such as... Figure 2 As shown, it includes the following steps: S1. Determine the targets of governance and the multi-scale governance areas: The target area is divided into governance units such as district and county-level administrative regions, key governance areas, and 15-minute living circles.

[0075] S2. Identify key stakeholders in parking management: Identify the entities responsible for parking management, with managers, operators, and users as the core stakeholders.

[0076] S3. Constructing the VPOSRM indicator system: Using six subsystems—Vitality, Pressure, Organization, State, Resilience, and Management—as indicator dimensions, and corresponding to three types of subjects—managers, operators, and users—18 parking management demand indicators are formed.

[0077] S4. Collect and preprocess multi-source parking management data: Collect and process multi-source data, including parking resource census data, smart parking platform data, on-site verification data, GIS spatial data, and satisfaction survey data, and standardize the indicators.

[0078] S5. Calculate 18 parking management demand indicators: The weights of the indicators were calculated using the projection pursuit method, the analytic hierarchy process (AHP) and the entropy method, and the parking management demand characteristic index (PGDCI) was constructed by combining the weights.

[0079] S6. Ranking and categorizing governance regions based on the RSR model: Based on the PGDCI results, the rank-sum ratio method is used to sort and classify each governance unit, and output the governance level, key weakness indicators and governance priority.

[0080] S7. Output the collaborative governance strategy text: Suggestions for managers to supplement parking resources, optimize the layout of public parking lots, and adjust on-street parking; suggestions for operators to upgrade smart parking, rectify safety and compliance issues, and improve service quality; and suggestions for users to improve walking distances, supplement charging facilities, disclose fee information, and optimize entrances and exits.

[0081] In this embodiment, the 18 parking management requirements indicators include at least the following: Indicators for managers: average number of parking spaces per vehicle, reasonable level of on-street parking space allocation, density of public parking lots, reasonable level of parking facility structure, level of identification of parking conflict areas, and optimal utilization rate of parking resources; Indicators for operators: parking space occupancy rate, fire safety compliance rate, level of smart parking services, parking fee arrears rate, completeness rate of parking signs and markings, and satisfaction with parking service quality; Indicators for users: convenience of parking and travel, completeness rate of parking and charging facilities, level of parking space design, level of smart parking information services, level of smart parking entrance and exit design, and completeness and transparency of parking fee information. The value range and data sources of these indicators are shown in Table 1.

[0082] Table 1. Indicator Categories, Data Collection Types, and Value Ranges 1. Construction of a demand characteristic index system for urban parking management based on the VPOSRM framework 1.1 Indicator System for Urban Parking Management (1) Number of parking spaces per vehicle To reflect the vitality of urban parking facilities, the average number of parking spaces per vehicle is selected as an indicator of the manager's vitality subsystem. The average number of parking spaces per vehicle is the average number of parking spaces available per motor vehicle. The calculation formula is as follows: ; In the formula, The number of parking spaces per vehicle; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles.

[0083] (2) Parking spaces on the road are set up at a reasonable level. According to the "Specifications for Setting Up On-Street Parking Spaces in Urban Roads" (GA / T850-2021), road width, peak hour V / C ratio, and average vehicle travel speed are used as compliance criteria to assess the compliance and supplementary role of on-street parking space settings. A reasonable level of on-street parking space settings refers to the proportion of parking lots that meet the conditions for setting up on-street parking spaces, calculated using the following formula: ; In the formula, Set a reasonable level (%) for on-street parking spaces; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles.

[0084] (3) Distribution density of public parking lots To reflect the balanced spatial distribution of urban public parking lots, public parking lot density is selected as an indicator for the management organization subsystem. Public parking lot density refers to the ratio of the number of public parking lots or the total number of parking spaces to the area of ​​a specific region. The calculation formula is as follows: ; In the formula, The number of parking spaces per vehicle; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles.

[0085] (4) The rationality of the parking facility structure To reflect the supply status of urban parking facilities, the rationality level of parking facility structure is selected as a status subsystem indicator for parking management managers. The calculation formula is as follows: ; In the formula, To ensure the rational level of parking facility structure; The proportion of parking facilities to be built; The proportion of public parking facilities in the city; The percentage of on-street parking facilities; The total supply of parking facilities is the sum of the supply of parking facilities in buildings, public parking facilities in cities, and on-street parking facilities. Number of parking facilities to be provided; The number of public parking facilities available in the city; The number of parking facilities supplied on the road.

[0086] (5) Level of identification of parking conflict areas To reflect the resilience of urban parking systems and assess the ability of managers to identify areas with prominent parking conflicts, the level of identification of parking conflict areas is selected as an indicator of the manager's resilience subsystem. The evaluation criteria include: ① hospitals; ② transportation hubs; ③ old residential areas; ④ commercial areas; and ⑤ tourist attractions. The calculation formula is as follows: ; In the formula, Level of identification for parking conflict areas; The number of evaluation items identified.

[0087] (6) Parking resource optimization utilization rate To reflect the management level of urban parking resources, the optimal utilization rate of parking resources is selected as an indicator for the parking management subsystem. Compared with the composite algorithm of "sharing rate × turnover rate" in Japan's Parking Sharing Act, the optimal utilization rate of parking resources is the sum of the smart parking platform access rate and the off-peak sharing rate, calculated as follows: ; In the formula, DZL represents the optimized utilization rate of parking resources; RJ represents the access rate of the smart parking platform; RC represents the off-peak sharing rate; NJ represents the number of parking lots connected to the smart parking platform; and NC represents the number of off-peak shared parking lots.

[0088] 1.2 Urban Parking Management Operator Indicator System (1) Parking space occupancy rate To reflect the vitality of the urban parking system and assess the balance of use among various types of parking lots, parking space occupancy rate is selected as an indicator of operator vitality. Parking space occupancy rate refers to the ratio of the actual cumulative number of vehicles parked at a given time (time period) to the supply of parking facilities. The calculation formula is as follows: ; In the formula, Parking space occupancy rate; This represents the actual number of vehicles occupied in the parking lot. This represents the total number of parking spaces in the parking lot.

[0089] (2) Fire safety compliance rate To measure whether the fire protection facilities and measures of a parking lot comply with national fire safety standards and to reflect the operational pressure on operators, a fire safety compliance rate indicator is included. The calculation formula is as follows: ; In the formula, To improve fire safety compliance rate; The number of projects that meet fire safety standards.

[0090] (3) Smart parking service level Improving the level of smart parking management services can provide support for government planning and resource allocation

[17] . According to the Shaanxi Province "Technical Standard for Traffic Design of Smart Parking Lots (Garages)", smart parking services are divided into three levels: S1, S2, and S3. In order to reflect the operator's organizational capabilities and the level of smart operation, a smart parking service level indicator is set. The calculation formula is as follows: ; In the formula, To improve the level of smart parking services; , , The smart parking service levels are respectively , , The number of parking lots; The number of parking lots supplied within the area.

[0091] (4) Parking fee arrears rate To reflect the revenue status of urban parking operators, a parking fee arrears rate indicator is included. Calculation formula: ; In the formula, Parking fee arrears rate; Unpaid parking fees; Parking fees due.

[0092] (5) Completeness rate of parking signs and markings The integrity of parking signs and markings directly affects the reliability of parking guidance systems. According to GB5768 "Road Traffic Signs and Markings" and Shaanxi Province standards, parking signs and markings should meet multiple requirements to ensure the orderly operation of parking lots and the convenience of users. This refers to the percentage of parking guidance signs, markings, and other traffic signage systems within a parking lot that meet integrity and functionality standards. The calculation formula is as follows: ; In the formula, The completeness rate of parking signs and markings; The actual number of items that meet the verification standards; The total number of items to be verified.

[0093] (6) Parking service quality satisfaction To comprehensively reflect the overall level of service management, parking service quality satisfaction was selected as a management system indicator for operators. Parking service quality satisfaction data was obtained through questionnaires, employing a combination of survey methods. Online, questionnaires were distributed through professional survey platforms and widely disseminated via social media and parking-related apps, covering user groups with different regions and travel habits. Offline, paper questionnaires were randomly distributed in high-traffic areas such as parking lots, communities, and commercial centers to ensure sample diversity.

[0094] 1.3 User Indicator System for Urban Parking Management (1) Convenience level of parking and travel To reflect the ease with which users can access parking services and to demonstrate the vitality of parking facilities, the level of convenience in parking and travel is selected as an indicator of the user vitality subsystem. The calculation formula is as follows: ; In the formula, To improve the convenience of parking and travel; This represents the actual walking distance.

[0095] (2) Parking charging facility completeness rate In accordance with the State Council's "New Energy Vehicle Industry Development Plan (2021-2035)," it is clearly required that by 2025, 100% of newly built residential parking spaces must be equipped with charging facilities or have reserved installation conditions; the Ministry of Housing and Urban-Rural Development's "Urban Parking Facility Planning Guidelines" stipulates that the proportion of charging spaces in public parking lots shall not be less than 10%, and in core business districts, transportation hubs and other areas, it shall reach more than 30%; and the Shaanxi Province's "Technical Standards for Traffic Design of Smart Parking Lots (Garages)" stipulates the requirements for charging facilities in traffic design. In order to reflect the pressure of energy transition, the parking charging facility completeness rate index is added as a user pressure subsystem index.

[0096] (3) Parking space design level According to the "Code for Design of Parking Garages" (JGJ 100-2015) and the "Technical Standard for Traffic Design of Smart Parking Lots (Garages)" of Shaanxi Province, the design of parking spaces should meet the corresponding requirements as shown in the table. In order to evaluate the rationality of the physical organization of parking facilities, the level of parking space design is selected as the subsystem index of user organization. (4) Level of Smart Parking Information Service To assess the operational status of the information service system, a smart parking information service level indicator was added as a user status subsystem indicator. The following 10 facilities were checked: total parking space display, real-time available parking space display, vehicle payment QR code scanning device, mini-program parking space query function, roadside parking vacancy indicator screen, entrance / exit fee display board, parking space reservation function, parking bill details display, dynamic navigation information push, and parking fee transparency survey.

[0097] ; In the formula, Information acquisition index; To verify the number of items that passed.

[0098] (5) Design level of smart parking entrances and exits To assess the facility's ability to respond to emergencies and its daily usability, and to reflect the resilience of parking lot design, the design level of smart parking entrances and exits is selected as an indicator of the user resilience subsystem. The size of smart parking lots (garages) should be categorized into extra-large, large, medium, and small based on the number of parking spaces.

[0099] (6) Completeness and transparency of parking fee information To assess the completeness of parking lot fees and key information disclosed to the public, and to reflect the transparency and standardization of the management process, the completeness and transparency of parking fee information is selected as an indicator for the user management subsystem. The evaluation criteria include 10 items: pricing method, parking area classification, fee collection entity, fee items, fee standards, fee periods, billing method, fee basis, preferential policies, and complaint hotline. The calculation formula is as follows: ; In the formula, To ensure complete transparency of parking fee information; To meet the number of evaluation items required for the evaluation content.

[0100] 2. Methods for Demand Characteristics Analysis of Urban Parking Management 2.1 Weighting Method Based on Projection Pursuit (1) Data normalization processing To ensure consistency in the dimensions of the evaluation index values, preprocessing of each evaluation index value is required. The set of evaluation index values ​​is as follows: ,in For the first The first region Each evaluation indicator value, and The indicator values ​​are normalized based on the number of regions and the number of evaluation indicators, respectively. For positive indicators (the higher the calculated value of the indicator, the higher the level of parking management): ; For negative indicators (the higher the calculated value of the indicator, the lower the level of parking management): ; In the formula, Indicates the first in the region The minimum value of each evaluation indicator; Indicates the first in the region The maximum value of each evaluation indicator; This represents the normalized data.

[0101] (2) Constructing the projection index function Will 3D data transformed into One-dimensional projection value of the projection direction .

[0102] ; In the formula, It is a unit length vector.

[0103] When projecting, the projected value is required. The dispersion degree is a pattern where the whole is as dispersed as possible, while the local areas are as dense as possible. Therefore, the projection index function can be expressed as: ; ; ; In the formula, Indicates projection value Standard deviation; Indicates projection value Local density; Indicates projection value Expectations; Indicates the distance between samples. ; The window radius representing the local density; Represents a unit function that exceeds the order, when When, its function value is 1, when When , its function value is 0.

[0104] (3) Optimize the projection index function Given a set of evaluation index values ​​for a region, the projection direction 'a' is the only influencing variable of the projection index function Q(a). The optimal projection direction is the projection direction that best represents a certain type of feature structure in high-dimensional data. Therefore, solving the optimal projection direction optimization problem can be transformed into solving the problem of maximizing the projection index function using an optimization algorithm.

[0105] ; The above is a... To optimize complex nonlinear optimization problems with variable optimization, a genetic algorithm is applied to find the optimal solution.

[0106] (4) Calculation of projection weights The optimal projection direction value obtained in step (3) is the weight of each evaluation index.

[0107] Analytic Hierarchy Process (AHP) Weighting (1) Construction of factor set and evaluation set Let the factor set be: ; In the formula, These are the various influencing indicators of the evaluation objectives. Based on the comprehensive evaluation indicator system, they will be... The factors are divided into two layers: the first layer is the scheme layer, and the second layer is the criterion layer.

[0108] Based on a defined set of factors, an evaluation set of indicators can be constructed, which can be represented as: ; In the formula, It is a possible judgment result made on the evaluation indicators.

[0109] The evaluation set is defined as five levels: Excellent, Good, Average, Satisfactory, and Needs Improvement. Each level corresponds to a score of 0.9, 0.8, 0.7, 0.6, and 0, respectively. Therefore, the evaluation set vector representation is: .

[0110] (2) Weight allocation of evaluation factors The Analytic Hierarchy Process (AHP) is a multi-criteria decision-making and evaluation method that combines qualitative and quantitative approaches. It decomposes the relevant elements of a decision into an objective layer, a criterion layer, and an alternative layer. Experts then rank the merits of the alternatives based on their judgment. Qualitative and quantitative analyses are performed on this basis, and weights are assigned. Specific steps: 1) Establish a hierarchical structure model Based on an in-depth analysis of the practical problems, the relevant factors are decomposed into several levels from top to bottom according to their different attributes. The factors in the same level are subordinate to the factors in the upper level, while also dominating or being affected by the factors in the lower level, and the factors in the same level are as independent as possible from each other.

[0111] 2) Constructing pairwise comparison matrices Starting from the second level of the hierarchical model, for factors belonging to each factor in the previous level, a pairwise comparison matrix is ​​constructed using pairwise comparison and a 1-9 comparison scale, up to the lowest level. A 1-9 scale is used to represent the importance between two indicators. As the number increases, the difference in influence between the m-th and n-th factors gradually increases, with the lowest score of 1 indicating equal influence and the highest score of 9 representing an absolutely stronger influence of the m-th factor than the n-th factor.

[0112] 3) Calculate the weight vector First, calculate the feature vector values ​​to obtain the indicator weights. At the same time, the value of the largest eigenvalue is obtained. This is used for the next step of consistency verification.

[0113] 4) Consistency check For each pairwise comparison matrix, calculate the largest eigenvalue and its corresponding eigenvector, and perform a consistency test using the consistency index, random consistency index, and consistency ratio. If the test passes, the eigenvector becomes the weight vector; if it fails, the pairwise comparison matrix needs to be reconstructed. The consistency test uses... Analyze the values. A value less than 0.1 indicates that the consistency test has been passed, while a value greater than 0.1 indicates that the consistency test has not been passed.

[0114] Calculate the consistency index : ; Based on the order n of the judgment matrix, the corresponding RI value is selected. The random consistency index values ​​are shown in the table. Calculate the random consistency ratio of the judgment matrix : ; The condition for judgment is: if Then the judgment matrix needs to be reconstructed until... .

[0115] 5) Calculate the combined weight vector and perform a combination consistency test. Calculate the combined weight vector of the lowest level relative to the target and perform a combined consistency test. If the test passes, the decision can be made based on the results represented by the combined weight vector; otherwise, the model needs to be reconsidered or the consistency ratios need to be reconstructed. Larger pairwise comparison arrays.

[0116] 2.3 Entropy method for weighting (1) Dimensionless processing of indicator data The original data of 18 parking facility evaluation indicators from n evaluation areas in the city are used to construct an evaluation indicator matrix B = (bij)n×18 with n rows and 18 columns.

[0117] To ensure the comparability of evaluation indicator data, the data needs to be dimensionless. Different processing methods are used for positive and negative indicators. Positive indicators indicate that the higher the calculated value of the indicator, the higher the level of parking facility development; negative indicators indicate that the higher the calculated value of the indicator, the lower the level of parking facility development.

[0118] For positive indicators Process it according to the following formula: ; In the formula, refer to Each region ; Refers to 18 indicators, ; For the first The first scheme Item indicator value.

[0119] For negative indicators Process it according to the following formula: ; (2) Determine the indicator weights, that is, calculate the weights of the indicators. The first item under the indicator The proportion of each option in this indicator ; In the formula, Refers to the first The first item under the indicator The proportion of each option in this indicator.

[0120] (3) Calculate the first Entropy value of the item index ; In the formula, Refers to the first Information entropy of the indicator.

[0121] (4) Calculate the first Coefficient of difference of the items ; In the formula, Refers to the first The coefficient of difference for the indicator. For the first... Item, indicator value The greater the difference, the greater its impact on the evaluation of the scheme, and the smaller the entropy value.

[0122] (5) Calculate the weight of each indicator. ; In the formula, For the first The weight of each indicator.

[0123] 2.4 Construction of Parking Management Demand Characteristic Index (PGDCI) The Parking Demand Characteristic Index (PGDCI) is calculated using the following formula: ; In the formula: PGDCI is the parking management demand characteristic index, α is the projection pursuit weight, β is the analytic hierarchy process weight, and γ is the entropy method weight. The weights assigned by experts to consider the characteristics of governance data, policy inclinations, and the needs of governance stakeholders in specific parking management scenarios. .

[0124] In practical analysis, this can be improved in scenarios with large data sample sizes and highly quantifiable indicators. The value increases when policy inclinations or specific needs of the subject are involved. The value is improved when dynamic optimization is required. value.

[0125] Combining weights enhances the model's adaptability to different scenarios. By flexibly adjusting the weight coefficients, PGDCI can better adapt to the parking management needs of different scenarios such as commercial areas, residential areas, and hospitals. The combined weight mechanism reduces the limitations of single methods and reduces errors caused by data noise or changes in scenarios. PGDCI can be continuously optimized as parking management data accumulates and business needs change, providing a basis for policy formulation.

[0126] 2.5 PGDCI-Rank Ratio Method Analysis of Parking Management Demand Characteristics The rank-sum ratio (RSR) ranking method is used to construct an evaluation matrix based on the values ​​of each evaluation index. After rank transformation of the matrix, the dimensionless rank-sum ratio is obtained, and the distribution of RSR values ​​is used to directly rank the evaluation objects according to their merits. The basic steps are as follows: (1) The original data of the 18 parking facility evaluation indicators in the 9 evaluation areas are used to construct an evaluation indicator matrix B = (bij)9×18 with 9 rows and 18 columns.

[0127] (2) Use the non-integer rank method to compile the rank of each indicator in each region.

[0128] (3) Rank the evaluation objects directly according to their RSR values. The calculation formula is: In the formula: , , Indicates the first row and number The rank of the elements in the column. Adding the weights obtained from the entropy method, the weighted rank sum ratio (WRSR) is calculated using the following formula: In the formula: Indicates the first row and number The rank of the elements in the column; For the first The weights of each evaluation indicator, among which The RSR value is dimensionless. .

[0129] (4) Determine the distribution of RSR.

[0130] The distribution of RSR refers to the downward cumulative probability of a specific RSR value expressed in probit units. The method involves compiling a ranked RSR frequency distribution table and accumulating the frequencies for each group. Calculate the cumulative frequency f for each group; determine the rank of the RSR for each group. And the average rank R; calculate the downward cumulative frequency. Convert the percentage to the probability unit Probit, where Probit is the standard normal deviation corresponding to the percentage p. .

[0131] (5) Calculate the regression equation.

[0132] Using the probability unit value Probit corresponding to the cumulative frequency as the independent variable and the RSR value as the dependent variable, the regression equation is calculated as follows: .

[0133] (6) Sorting by category.

[0134] according to The evaluation area is categorized into five tiers based on the standard normal deviation u. The evaluation objects are then further categorized according to the Probit value for each tier, calculated using the regression equation, and ranked accordingly.

[0135] (7) Verification of the optimal grading.

[0136] After categorizing parking facility levels across districts according to their RSR values, a variance consistency test is performed to determine if the categorization is optimal. This method is suitable for comparing multiple RSR values. Optimal categorization means that the variances of each categorization are consistent, and the differences in RSR values ​​are significant.

[0137] The above process allows all participating regions to be ranked according to the original data, and the development level of parking facilities in each region to be evaluated. Through annual data updates, an annual evaluation cycle is achieved, analyzing the reasons for insufficient development in regions and promoting parking facility development in the following year.

[0138] 2.6 Rules for Transforming Calculation Results into Governance Strategies The model output goes beyond just score ranking; it interprets both "indicator weights" and "regional tiers" together. If an indicator has a high weight and a region has a low rank for that indicator, then that indicator should be identified as a priority weakness for that region. If a region has a low tier and lags behind in multiple high-weight indicators, then that region should be included in the annual list of key parking management areas.

[0139] This embodiment constructs a refined indicator system encompassing six dimensions (V, P, O, S, R, M) and 18 items, covering the entire chain of resource status, operational pressure, system vitality, risk resistance and resilience, entity organization, and management capabilities. It aligns with the demands of multi-stakeholder collaborative governance, propelling parking governance from single-point evaluation to system-wide evaluation, resulting in more comprehensive evaluation results that better reflect real-world urban governance scenarios. Simultaneously, it builds an indicator and strategy system based on the collaborative needs of managers, operators, and users, ensuring that evaluation originates from multi-stakeholder needs and strategies serve multi-stakeholder implementation, addressing the problems of fragmented responsibilities, mismatched policies, and poor coordination among stakeholders in traditional governance. Regarding weighting, traditional techniques often employ single subjective or single objective weighting methods, easily leading to strong subjective bias or detachment from business realities. This method integrates three algorithms: projection pursuit high-dimensional feature mining, analytic hierarchy process (AHP) experience-based weighting, and entropy-based data difference weighting. By combining coefficients, it achieves deep integration of subjective and objective factors, respecting the business rules of the parking governance industry while fully mining the inherent characteristics of the data. This results in a higher degree of matching between indicator weights and governance needs, and stronger credibility of index calculation results.

[0140] The above description is merely an embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and scope of the present invention are included within the scope of protection of the present invention.

[0141] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of the invention, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the invention. The sequence numbers of the above-described embodiments of the invention are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0142] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, or apparatus. Without further limitations, an element defined by the phrase "comprising a..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. In the several embodiments provided by this invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components may be combined, or integrated into another system, or some features may be ignored or not performed.

[0143] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for evaluating and classifying urban parking management based on the collaborative needs of multiple stakeholders, characterized in that, The method includes: Obtain the urban area to be governed and divide it into multiple governance units; identify the relevant stakeholders in parking governance and construct a subject set including managers, operators and users; Based on the subject set and six dimensions—Vitality (V), Stress (P), Organization (O), State (S), Resilience (R), and Management (M)—a VPOSRM indicator system containing 18 parking governance demand indicators is constructed. The process involves acquiring multi-source heterogeneous raw data from various governance units and standardizing it to generate a standardized indicator matrix. The multi-source heterogeneous raw data includes parking resource survey data, smart parking platform operation data, on-site verification data, GIS spatial data, and satisfaction survey data. The three types of weight vectors for each indicator in the standardized indicator matrix are calculated using the projection pursuit method, the analytic hierarchy process (AHP), and the entropy method, respectively. Based on the preset combined weight coefficients, the calculated three types of weight vectors are combined and weighted to obtain the comprehensive weight of each indicator. Based on the comprehensive weight and the standardized indicator matrix, the parking management demand characteristic index of each governance unit is calculated. Based on the parking management demand characteristic index, the rank-sum ratio method is used to rank each management unit, and the grading threshold is calculated by the Probit regression model to divide each management unit into different management levels. According to the weight ranking of the comprehensive weight of each indicator and the management level of each management unit, collaborative management strategy texts corresponding to managers, operators and users are generated respectively.

2. The method according to claim 1, characterized in that, The 18 parking management requirements indicators include at least: Indicators for managers: number of parking spaces per vehicle, reasonable level of on-street parking space allocation, density of public parking lots, reasonable level of parking facility structure, level of identification of parking conflict areas, and optimal utilization rate of parking resources; Indicators for operators include: parking space occupancy rate, fire safety compliance rate, smart parking service level, parking fee arrears rate, parking sign and marking completeness rate, and parking service quality satisfaction. Indicators for users include: convenience of parking and travel, completeness of parking and charging facilities, design level of parking spaces, level of smart parking information services, design level of smart parking entrances and exits, and completeness and transparency of parking fee information.

3. The method according to claim 2, characterized in that, The formula for calculating the performance indicators for managers is as follows: The formula for calculating the average number of parking spaces per vehicle is: ； In the formula, The number of parking spaces per vehicle; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles; The formula for calculating a reasonable level of on-street parking spaces is: ； In the formula, Set a reasonable level (%) for on-street parking spaces; The total supply of parking facilities is the sum of the supply of parking facilities built within districts and counties, urban public facilities, and on-street parking facilities; The number of motor vehicles; The formula for calculating the distribution density of public parking lots is: ； In the formula, Public parking lot distribution density; Number of off-street public parking lots available; Provide the number of public parking lots within the roadside; The built-up area; The formula for calculating the rational level of parking facility structure is: ； In the formula, To ensure the rational level of parking facility structure; The proportion of parking facilities to be built; The proportion of public parking facilities in the city; The percentage of on-street parking facilities; The total supply of parking facilities is the sum of the supply of parking facilities in buildings, public parking facilities in cities, and on-street parking facilities. Number of parking facilities to be provided; The number of public parking facilities available in the city; Provide the number of on-street parking facilities; The formula for calculating the level of parking conflict area identification is as follows: ； In the formula, Level of identification for parking conflict areas; The number of evaluation items identified; The formula for calculating the optimal utilization rate of parking resources is: ； In the formula, D ZL To optimize the utilization rate of parking resources; R J For smart parking platform access rate; R C For time-sharing rate; N J The number of parking lots connected to the smart parking platform; N C This refers to the number of shared parking lots during off-peak hours.

4. The method according to claim 2, characterized in that, The formulas for calculating the metrics for operators are as follows: The formula for calculating parking space occupancy rate is: ； In the formula, Parking space occupancy rate; This represents the actual number of vehicles occupied in the parking lot. This represents the total number of parking spaces in the parking lot. The formula for calculating the average number of parking spaces per vehicle is: ； In the formula, To improve fire safety compliance rate; The number of items that meet fire safety standards; The formula for calculating the level of smart parking services is: ； In the formula, To improve the level of smart parking services; , , The smart parking service levels are respectively , , The number of parking lots; The number of parking lots to be supplied within the area; The formula for calculating the parking fee arrears rate is: ； In the formula, Parking fee arrears rate; Unpaid parking fees; Parking fees due; The formula for calculating the completeness rate of parking signs and markings is: ； In the formula, The completeness rate of parking signs and markings; The actual number of items that meet the verification standards; The total number of items to be verified; The formula for calculating parking service quality satisfaction is: ； In the formula, For service quality satisfaction; Number of satisfied users; This represents the total number of users surveyed.

5. The method according to claim 2, characterized in that, The formula for calculating user metrics is as follows: The formula for calculating the level of parking convenience is: ； In the formula, To improve the convenience of parking and travel; This is the actual walking distance; The completeness rate of parking charging facilities is calculated based on the ratio of charging parking spaces and the size of the facilities. The design level of parking spaces is verified and calculated based on the architectural design code for parking garages. The formula for calculating the level of intelligent parking information services is: ； In the formula, Information acquisition index; To verify the number of items that passed; The design level of smart parking entrances and exits is based on the display of available parking spaces, payment functions, and navigation push verification calculations. The formula for calculating the completeness and transparency of parking fee information is: ； In the formula, To ensure complete transparency of parking fee information; To meet the number of evaluation items required for the evaluation content.

6. The method according to claim 1, characterized in that, The parking management demand characteristic index is calculated through the following steps: The positive and negative indicators in the multi-source heterogeneous raw data are standardized to obtain a standardized indicator matrix; The standardized index matrix is ​​reduced in dimension using the projection pursuit method to obtain the optimal projection direction vector; the judgment matrix is ​​constructed using the analytic hierarchy process (AHP) to calculate the subjective weight vector; the information entropy of each index is calculated using the entropy method, and the objective weight vector is determined based on the difference coefficient. The three types of weight vectors are linearly weighted based on the preset combined weight coefficients to obtain the comprehensive weight of each indicator; wherein the preset combined weight coefficients are the projection pursuit weight coefficient α, the analytic hierarchy process weight coefficient β, and the entropy method weight coefficient γ, and satisfy α+β+γ=1. The parking management demand characteristic index of each governance unit is obtained by weighting and summing the standardized index matrix with the corresponding comprehensive weights.

7. The method according to claim 1, characterized in that, Based on the parking management demand characteristic index, the rank-sum ratio method is used to rank each management unit, and the grading threshold is calculated using a Probit regression model to divide each management unit into different management levels, including: An evaluation matrix is ​​constructed using the parking management demand characteristic index as the evaluation index value. The evaluation matrix is ​​then rank-transformed to generate a rank matrix. Calculate the weighted rank sum ratio of each governance unit based on the rank matrix; Determine the distribution of the weighted rank sum ratio, calculate the downward cumulative frequency corresponding to each weighted rank sum ratio value, and convert the downward cumulative frequency into a probability unit Probit value; A linear regression equation was constructed using the probability unit Probit as the independent variable and the weighted rank sum ratio as the dependent variable. Based on the preset number of tiers, the Probit value range corresponding to each tier is determined based on the standard normal deviation. The corresponding weighted rank-sum ratio estimate is then calculated by substituting it into the regression equation. The weighted rank-sum ratio estimate is used as the tiering threshold to classify each governance unit and obtain the governance level.

8. The method according to claim 7, characterized in that, Generate collaborative governance strategy texts corresponding to managers, operators, and users respectively: The high-weight indicators with a weight greater than a preset weight threshold in the comprehensive weight are matched with the low-rank indicators with a rank lower than a preset rank threshold in the rank matrix. If an indicator is both high-weight and low-rank, then the indicator is determined as the priority weakness indicator of the corresponding governance unit. Based on the governance level and priority weakness indicators, collaborative governance strategy texts are retrieved from a pre-defined strategy library and generated.

Images