Cross-regional municipal road maintenance resource scheduling method and system

Abstract

The invention discloses a cross-regional municipal road maintenance resource scheduling method and system. The scheduling method comprises the following steps: collecting road disease data, maintenance historical records and real-time traffic flow data of each region; constructing a judgment matrix by applying an analytic hierarchy process, calculating a road factor weight through a root method, calculating a maintenance emergency degree score of each region according to the quantized basic data and the weight, and dividing region maintenance grades; according to the maintenance level of each area, maintenance groups with different priorities are divided; counting the existing maintenance resources of each region, and predicting the resource demand quantity of each region by combining a maintenance group division result and taking historical maintenance resource consumption data as a benchmark; for a resource shortage region, selecting a scheduling path by adopting a greedy algorithm, and formulating a preliminary cross-region resource scheduling scheme; and constructing a multi-target cost evaluation algorithm, and iteratively optimizing the scheme in combination with a particle swarm optimization algorithm. According to the invention, efficient, accurate and low-cost scheduling of cross-regional municipal road maintenance resources is realized.

Description

Technical Field

[0001] This invention relates to the field of transportation, and more particularly to a method and system for scheduling municipal road maintenance resources across regions. Background Technology

[0002] With the development of urban agglomerations and metropolitan areas, the spatial scope of cities continues to expand. Municipal road networks break through the boundaries of single administrative regions, forming inter-regional and inter-provincial transportation interconnection systems. The traditional "separate management" maintenance model based on administrative regions is insufficient to meet the collaborative maintenance needs of inter-regional roads. Maintenance standards and resource allocation efficiency for the same road vary significantly across different administrative regions. Resource allocation is uneven across regions; economically developed areas have abundant equipment, while less developed areas lack resources, resulting in slow dispatch response during inter-regional operations. Based on this, this invention proposes a cross-regional municipal road maintenance resource scheduling method and system. Summary of the Invention

[0003] This invention provides a method for scheduling municipal road maintenance resources across regions, characterized by comprising:

[0004] S10. Collect road damage data, maintenance history records and real-time traffic flow data in various areas through IoT sensors and municipal management system interfaces, preprocess the collected data and construct a structured dataset.

[0005] S20. Based on the type, severity and traffic flow of road defects, a judgment matrix is ​​constructed using the analytic hierarchy process. The weight of each factor is calculated using the root method. Based on the quantified basic data and weights, the maintenance urgency score of each area is calculated, and the maintenance level of each area is classified.

[0006] S30. Based on the maintenance level of each area, use clustering algorithms combined with regional geographical location and road network connectivity to divide maintenance groups with different priorities, and draw a heat map of maintenance priorities.

[0007] S40. Statistically analyze the existing maintenance resources in each region, combine the maintenance group division results, and use historical maintenance resource consumption data as a benchmark to predict the resource demand in each region using regression analysis.

[0008] S50. For resource-scarce areas, a greedy algorithm is used to prioritize the scheduling path that is shortest to resource-rich areas and has complementary resource types. A resource scheduling threshold is set, and a preliminary cross-regional resource scheduling plan is formulated.

[0009] S60. Construct a multi-objective cost evaluation algorithm, and combine it with the particle swarm optimization algorithm to iteratively optimize the scheduling path and resource allocation ratio of the preliminary scheme, and select the scheme whose total cost is reduced to a set threshold as the final execution scheme.

[0010] The method for scheduling municipal road maintenance resources across regions, as described above, involves collecting road damage data, maintenance history records, and real-time traffic flow data from various regions through IoT sensors and municipal management system interfaces. The collected data is then preprocessed to construct a structured dataset. Specifically, this involves the following sub-steps:

[0011] Dynamic data on crack development are collected by distributed fiber optic sensors deployed on the road surface, and the depth and area data of potholes are obtained by scanning with lidar carried by drones.

[0012] Access the municipal management system database, extract maintenance history records for the past three years, integrate real-time traffic flow interface data from the transportation department, and collect traffic flow and speed information for various areas and road sections.

[0013] Box plot analysis was applied to the collected road defect data to identify and remove outliers. For traffic flow data, the isolated forest algorithm was used to detect sudden traffic flow anomalies.

[0014] The aforementioned cross-regional municipal road maintenance resource scheduling method involves constructing a judgment matrix based on road defect type, severity, and traffic flow using the analytic hierarchy process (AHP). The weights of each factor are calculated using the square root method. Based on the quantified baseline data and weights, a maintenance urgency score for each region is calculated, thus classifying regional maintenance levels. Specifically, it comprises the following sub-steps:

[0015] Establish a hierarchical index system of factors to assess the urgency of road maintenance and determine the key assessment factors for roads at each level.

[0016] A pairwise comparison judgment matrix is ​​constructed based on factor scoring, and the weight of each factor is calculated using the square root method. The rationality of the weights is ensured through consistency checks.

[0017] The quantified factor data for each region are multiplied and summed with their weights to calculate the maintenance urgency score for each region, thus classifying the maintenance levels of each region.

[0018] The method for scheduling municipal road maintenance resources across regions, as described above, involves dividing maintenance into different priority groups based on the maintenance level of each region, using a clustering algorithm that combines regional geographical location and road network connectivity, and then creating a heatmap of maintenance priorities. Specifically, it consists of the following sub-steps:

[0019] Using the maintenance urgency score as the basic variable, and introducing regional area and population density as auxiliary variables, a density clustering algorithm is used to automatically identify densely connected regional clusters and determine the number of clusters.

[0020] By combining clustering results with geographic information systems, we can analyze the road network connectivity and transportation hub distribution in each region. We can prioritize merging regions with directly connected road networks and similar maintenance needs. For regions with ambiguous boundaries, we can adjust the cluster boundaries through manual intervention.

[0021] The method for scheduling municipal road maintenance resources across regions, as described above, involves statistically analyzing the existing maintenance resources in each region, combining this with the maintenance group division results, and using historical maintenance resource consumption data as a benchmark to predict the resource demand in each region through regression analysis. Specifically, it comprises the following sub-steps:

[0022] Through the municipal asset management system, the maintenance resources in each area are counted, a resource ledger is established, and the status of resource usage is recorded.

[0023] Based on historical maintenance resource consumption data and combined with future regional plans, the resource demand of each region is predicted for the next three months.

[0024] Calculate the resource supply-demand ratio for each region, identify resource-scarce areas, and further analyze the types of resource gaps.

[0025] The method for scheduling municipal road maintenance resources across regions, as described above, employs a greedy algorithm to prioritize scheduling paths that are shortest in resource-scarce areas and have complementary resource types to resource-scarce areas. A resource scheduling threshold is set, and a preliminary cross-regional resource scheduling plan is formulated.

[0026] Specifically, it is divided into the following sub-steps:

[0027] Establish a resource type matching matrix, compare the personnel skills, equipment functions, and material types of resource-scarce areas and redundant areas, calculate the matching degree score, and combine the shortest path algorithm between areas to determine potential scheduling routes.

[0028] Starting with the region pair with the highest matching degree and the shortest distance, resources are allocated step by step according to the greedy algorithm. During the allocation process, the resource status of each region is updated in real time to ensure that the remaining resources in the allocated region meet the basic maintenance needs.

[0029] Based on the resource allocation results, a detailed scheduling list is developed and a resource scheduling flowchart is drawn.

[0030] The method for scheduling municipal road maintenance resources across regions, as described above, involves constructing a multi-objective cost evaluation algorithm and combining iterative optimization of the scheduling path and resource allocation ratio of the initial plan with a particle swarm optimization algorithm. The final execution plan is selected based on the scheme whose total cost is reduced to below a set threshold. Specifically, it consists of the following sub-steps:

[0031] A multi-objective cost assessment algorithm is constructed based on transportation costs, time costs, and labor costs;

[0032] By employing the particle swarm optimization algorithm, two optimization dimensions are set: path optimization and resource allocation ratio adjustment. Through iterative calculation, the scheduling scheme is continuously updated to find the solution space with the minimum total cost.

[0033] This invention also provides a cross-regional municipal road maintenance resource scheduling system, comprising:

[0034] Data Acquisition and Processing Module: Collects road damage data, maintenance history records, and real-time traffic flow data in various areas through IoT sensors and municipal management system interfaces; preprocesses the collected data; and constructs a structured dataset.

[0035] Emergency Score Module: Based on road defect type, severity, and traffic flow, an analytic hierarchy process is used to construct a judgment matrix. The weight of each factor is calculated using the square root method. Based on the quantified basic data and weights, the maintenance urgency score of each area is calculated, and the maintenance level of each area is classified.

[0036] Maintenance group division module: Based on the maintenance level of each area, clustering algorithms are used to combine regional geographical location and road network connectivity to divide maintenance groups with different priorities, and a heat map of maintenance priority is drawn.

[0037] Demand prediction module: Statistics on existing maintenance resources in each region, combined with the results of maintenance group division, and based on historical maintenance resource consumption data, regression analysis is used to predict the resource demand in each region;

[0038] Scheduling scheme module: For resource-scarce areas, a greedy algorithm is used to prioritize the scheduling path that is shortest to resource-rich areas and has complementary resource types. Resource scheduling thresholds are set, and a preliminary cross-regional resource scheduling scheme is formulated.

[0039] Optimization Scheme Module: Construct a multi-objective cost evaluation algorithm, and combine it with the particle swarm optimization algorithm to iteratively optimize the scheduling path and resource allocation ratio of the preliminary scheme, and select the scheme whose total cost is reduced to a set threshold as the final execution scheme.

[0040] The beneficial effects achieved by this invention are as follows: This invention enables efficient, accurate, and low-cost scheduling of cross-regional municipal road maintenance resources. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0042] Figure 1This is a flowchart of a cross-regional municipal road maintenance resource scheduling method provided in Embodiment 1 of this application.

[0043] Figure 2 This is a schematic diagram of a cross-regional municipal road maintenance resource scheduling system provided in Embodiment 2 of this application. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of 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.

[0045] Example 1

[0046] like Figure 1 As shown, Embodiment 1 of this application provides a method for scheduling municipal road maintenance resources across regions, including:

[0047] S10. Collect road damage data, maintenance history records, and real-time traffic flow data in various areas through IoT sensors and municipal management system interfaces. Preprocess the collected data to construct a structured dataset.

[0048] S11. Collect dynamic data on crack development by deploying distributed fiber optic sensors on the road surface, and use lidar on a drone to scan and obtain data on pothole depth and area.

[0049] Distributed fiber optic sensors are laid on the road surface at 50-meter intervals. Based on the principle of fiber Bragg gratings, they monitor road surface strain in real time. The strain data is converted into crack width, length, and propagation rate using a strain-crack relationship model. A lidar mounted on a drone scans the road surface at a resolution of 10 centimeters. After removing noise using a point cloud filtering algorithm, the depth, area, and volume of potholes are obtained through point cloud 3D reconstruction technology.

[0050] S12. Access the municipal management system database, extract maintenance history records for the past three years, integrate real-time traffic flow interface data from the transportation department, and collect traffic flow and speed information for various areas and road sections.

[0051] Access the municipal management system database to extract road maintenance records for each area over the past three years, including maintenance time, resource input, and road defect types. Resources include manpower, equipment, and materials. Call the transportation department's data interface to obtain traffic flow, average speed, number of lanes, and vehicle type distribution ratio for each road segment every 15 minutes, including the proportion of small vehicles, medium vehicles, and large vehicles, and record the collection timestamp.

[0052] S13. Apply box plot analysis to the collected road defect data to identify and remove outliers; for traffic flow data, use the isolated forest algorithm to detect sudden traffic flow anomalies.

[0053] For road defect data, box plot analysis is used to identify outliers. First, the quartiles Q3 and Q1 of the data are calculated to obtain the interquartile range IQR = Q3 - Q1. Then, a correction factor α is introduced based on the data fluctuation. Data exceeding Q3 + α × 1.5 × IQR or below Q1 - α × 1.5 × IQR are considered outliers and removed.

[0054] For traffic flow data, a dynamic threshold model is constructed based on the Isolation Forest algorithm and combined with historical traffic flow data from the same period to identify outliers. Let the traffic flow at the current data point be q. cur The historical average flow rate for the same period was The standard deviation is σ his According to the formula Calculate the standardized score when S σ When the data point exceeds the dynamic threshold, it is considered an anomaly. The dynamic threshold function simulates the variation of traffic flow at different time scales to generate a threshold that closely reflects actual traffic conditions. The specific formula is T = f(t). season ,t weekday ,t hour ), where t season For the season, t weekday For the week, t hour For hours, f represents a functional relationship that maps the three independent variables—season, weekday, and hour—to a threshold used to determine whether traffic flow data is abnormal, through specific calculation rules and logic.

[0055] For missing maintenance history data, the random forest algorithm is used to predict missing values ​​based on the maintenance patterns of similar roads.

[0056] The processed data is categorized and coded according to region, time, and data type to construct standardized data tables for easy subsequent analysis.

[0057] S20. Based on the type, severity, and traffic flow of road defects, a judgment matrix is ​​constructed using the analytic hierarchy process (AHP). The weights of each factor are calculated using the square root method. Based on the quantified basic data and weights, the maintenance urgency score of each area is calculated, and the maintenance level of each area is classified.

[0058] S21. Establish a hierarchical index system of factors to assess the urgency of road maintenance and determine the key assessment factors for roads at each level.

[0059] Assessing the urgency of road maintenance involves numerous complex factors, making direct assessment difficult and prone to confusion. By establishing a hierarchical structure, this complex problem is broken down into sub-problems at different levels. The hierarchical structure consists of three layers: the objective layer, the criterion layer, and the indicator layer. The objective layer represents the urgency of maintenance; the criterion layer is categorized into factors such as road defects, traffic flow, and economic factors; and the indicator layer includes 10 key assessment factors: severity of road defects, impact on traffic flow, regional economic importance, density of surrounding facilities, historical frequency of maintenance, road grade, accident frequency, environmental requirements, policy priority, and number of public complaints.

[0060] S22. Construct a pairwise comparison judgment matrix based on factor scores, calculate the weight of each factor using the square root method, and ensure the rationality of the weights through consistency checks.

[0061] Based on the expert's scoring using the 1-9 scale to compare each factor pairwise, a judgment matrix A = (a ij ) n×n a ij To determine the importance of the i-th evaluation factor compared to the j-th evaluation factor, the element in the i-th row and j-th column of matrix A is used, where n represents the order of the judgment matrix. The factor weights are calculated using the root method, first by calculating the product of the elements in each row of the matrix. Then calculate its nth root. Introducing the expert authority coefficient γ i ,right Weighted normalization is performed to obtain the weight vector. The reasonableness of the weights is checked using a consistency test formula. The weights are considered reasonable if the test result is less than a set threshold.

[0062] S23. Multiply and sum the quantified factor data of each region with their weights to calculate the maintenance urgency score of each region and classify the maintenance level of each region.

[0063] The formula for calculating the maintenance urgency score for each area is as follows: Where, δ t W represents the weighting adjustment coefficient for different time periods. i Let x be the weight of the i-th factor. i This represents the quantitative score for the i-th factor. Based on the score, the system automatically divides the area into emergency maintenance zone, routine maintenance zone, and observation zone using the natural discontinuity grading method.

[0064] S30. Based on the maintenance level of each area, use a clustering algorithm combined with the geographical location of the area and the road network connectivity to divide the maintenance groups with different priorities and draw a heat map of maintenance priorities.

[0065] S31. Using the maintenance urgency score as the basic variable, the area and population density are introduced as auxiliary variables. A density clustering algorithm is used to automatically identify densely connected area clusters and determine the number of clusters.

[0066] Using the maintenance urgency score as the base variable, and introducing regional area, population density, and economic activity as auxiliary variables, a four-dimensional data space is constructed. Economic activity is calculated by weighting indicators such as regional GDP and number of enterprises. A density clustering algorithm is employed, with a dynamic neighborhood radius set as... The minimum number of samples is MinPts, which is dynamically adjusted based on the data density. ε0 is the initial radius, and θ is the adjustment coefficient. d represents the average distance from the current data point to its neighboring points. avg The distance is the global average distance. The algorithm starts from any unvisited point and searches for points in its ε-neighborhood. If the number of neighboring points is greater than or equal to MinPts, a new cluster is formed, and the points in the neighboring area are added to the cluster. The algorithm continues to search for neighbors from unvisited points within the cluster, continuously expanding the clusters until all points have been visited. Density-connected regions are automatically identified, and the number of clusters is determined.

[0067] S32. Combine the clustering results with the geographic information system to analyze the road network connectivity and transportation hub distribution in each region, and prioritize merging regions with directly connected road networks and similar maintenance needs; for regions with ambiguous boundaries, adjust the cluster boundaries through manual intervention.

[0068] The clustering results are imported into a geographic information system, and road network analysis tools are used to calculate the road network connectivity between different regions. The connectivity calculation formula is as follows: Where C represents road network connectivity, L represents the actual number of connecting roads between regions, τ represents the road capacity correction coefficient, determined based on road grade and number of lanes, N represents the total number of regions, μ represents the weighting coefficient, and w avg w is the average width of the connecting roads max The maximum width among all roads. Prioritize merging areas with road network connectivity greater than a set threshold and similar maintenance needs; similar maintenance needs are defined as an urgency score difference of less than 10. For areas with ambiguous boundaries, precise delineation is achieved by manually reviewing satellite imagery, road planning maps, and traffic flow heatmaps, combined with ant colony optimization algorithms to adjust the boundary paths.

[0069] S40. Statistically analyze the existing maintenance resources in each region, combine the maintenance group division results, and use historical maintenance resource consumption data as a benchmark to predict the resource demand in each region using regression analysis.

[0070] S41. Through the municipal department's asset management system, collect statistics on maintenance resources in each area, establish resource ledgers, and record the status of resource usage.

[0071] Through the municipal asset management system, information tables of maintenance personnel in each area are exported, and statistics are compiled on the number of different professionals, such as road repair, machinery operation, and electrical maintenance personnel, as well as the distribution of personnel skill levels and years of work experience. Equipment ledgers are obtained, recording equipment models, purchase dates, usage durations, current status, maintenance records, and equipment availability rates. Equipment availability rates are calculated based on historical fault data and maintenance records. Statistics on the inventory quantity, warehousing time, shelf life, and material loss rate of commonly used materials, including asphalt, sand, gravel, and cement, are compiled. Material loss rates are determined based on storage conditions and time. A dynamic resource ledger is established, updating resource usage status in real time, including personnel attendance, equipment scheduling records, material outbound quantities, and resource idle rates.

[0072] S42. Based on historical maintenance resource consumption data and combined with regional future planning, predict the resource demand of each region in the next three months.

[0073] The specific prediction formula is as follows: D future This represents the predicted demand for maintenance resources in a certain region over the next three months. This represents the monthly average of maintenance resource consumption in the region over the past three years, reflecting historical resource consumption levels. ω l The weighting of the impact of new road length on resource demand is determined through regression analysis based on the correlation between new road construction and resource consumption in historical data, with values ​​ranging from 0 to 1; Δl represents the total planned new road length in the next three months, reflecting the incremental impact of new road construction projects on resource demand. total Let q be the total existing road length in the area, and Δq be the projected increase in traffic flow over the next three months, obtained from traffic flow growth forecast data from the transportation department. avg ω represents the average monthly traffic flow in the area over the past year. q The weight of the impact of traffic flow changes on resource demand is also determined based on historical data regression analysis; y is the number of other factors affecting resource demand in the region's future planning, in addition to the construction of new roads and traffic flow growth, such as the holding of large-scale events and policy requirements for road upgrades and renovations. is the correction coefficient for the i-th other influencing factor on resource demand, determined based on a comprehensive assessment of the factor's scale, urgency, etc.; a positive number indicates increased demand, and a negative number indicates decreased demand; w i Let be the weight of the i-th other influencing factor.

[0074] Perform residual checks on the prediction results to ensure that the error is within an acceptable range.

[0075] S43. Calculate the resource supply-demand ratio for each region, identify resource-scarce areas, and further analyze the types of resource gaps.

[0076] The resource supply-demand ratio for each region is calculated, taking into account differences in resource timeliness and importance. The calculation formula is as follows: Where R is the overall supply-demand ratio, and D i S represents the predicted demand for resource type i. i For the existing quantity of the i-th type of resource, Let be the importance weight of the i-th resource category, and k be the number of resource categories. For regions where the ratio exceeds a set threshold, further analysis of resource gap types is conducted. By establishing a resource gap assessment matrix, considering factors such as the supply-demand ratio, resource urgency, and resource substitutability, specific resource shortage types are determined, including insufficient manpower, equipment shortages, and material scarcity. For regions with mild shortages, a four-level resource early warning system based on risk level is established, and different response strategies such as emergency procurement, regional allocation, temporary recruitment, and shared leasing are formulated accordingly.

[0077] S50. For resource-scarce areas, a greedy algorithm is used to prioritize the scheduling path that is shortest to resource-rich areas and has complementary resource types. A resource scheduling threshold is set, and a preliminary cross-regional resource scheduling plan is formulated.

[0078] S51. Establish a resource type matching matrix, compare the personnel skills, equipment functions, and material types of resource-scarce areas and redundant areas, and calculate the matching degree score; combine the shortest path algorithm between areas to determine potential scheduling routes.

[0079] Establish a multidimensional resource matching matrix M = (m ij ) b×c×e Where b represents the number of resource-scarce areas, c represents the number of resource-redundant areas, and e represents the resource category, including personnel, equipment, and materials. For personnel skill matching, a skill-task matching algorithm is constructed, considering factors such as skill level, work experience, and professional relevance, and expressed by the formula... Calculate the personnel matching score, w s For skill level weighting, p ijs Let r be the ratio of the demand for level s skilled personnel in the i-th shortage area to the j-th redundant area. ijs Let be the supply ratio of level s skilled personnel from the i-th shortage area to the j-th redundant area.

[0080] For equipment function matching, the similarity between the equipment model parameters and the required parameters is compared, and a equipment performance degradation coefficient is introduced, which is then calculated using the formula... Calculate the device matching score, n para h represents the number of device parameters. ijk Let β be the similarity of the k-th parameter between the i-th shortage region and the j-th redundant region. ijk This represents the performance degradation coefficient of the k-th parameter of the equipment in the j-th redundant region. The matching degree for material types is calculated directly based on the quantity of the same type. Matrix element mijv The matching score is calculated for the i-th shortage area and the j-th redundant area for the v-th type of resource. Based on the multi-dimensional resource matching matrix results and equipment function matching results, and combined with the shortest path algorithm between areas, the optimal path between each area is calculated using road distance, travel time, and road condition level as comprehensive weights, and potential scheduling routes are determined.

[0081] S52. Starting with the region pair with the highest matching degree and the shortest distance, resources are allocated step by step according to the greedy algorithm. During the allocation process, the resource status of each region is updated in real time to ensure that the remaining resources of the allocated region meet the basic maintenance needs.

[0082] Starting with selecting the region pairs with the highest matching score and optimal overall path from the multidimensional resource matching matrix, resource allocation is performed using a greedy algorithm. During the allocation process, the resource status of each region is updated in real time, and a dynamic resource retention threshold for the regions being moved out is set, taking into account factors such as resource transportation losses and allocation time costs. T base The basic retention ratio, η is the adjustment coefficient, and d yun For the transportation distance, d max To maximize transport distance, ensure that the originating region meets basic maintenance needs. If resource needs in a certain region are not fully met, continue allocating resources from the next best matching region, while considering resource complementarity and synergy, until the urgent needs of all shortage areas are basically met.

[0083] S53. Based on the resource allocation results, formulate a detailed scheduling list and draw a resource scheduling flowchart.

[0084] Based on the resource allocation results, a detailed scheduling list is developed, including the number and specialties of personnel to be scheduled, the model and quantity of equipment, and the types and weights of materials. Transportation vehicles are arranged, taking into account resource weight, volume, transportation time limits, and road weight and height restrictions. A vehicle scheduling optimization model based on a genetic algorithm is used to select appropriate vehicle types and transportation batches. Estimated arrival times are determined based on optimal route distance, real-time traffic conditions, vehicle speed limits, and rest periods. A resource scheduling flowchart is drawn, marking key nodes such as resource origin, transportation route, transfer station, and destination, clearly defining the responsible persons for each node, task completion time requirements, resource handover procedures, and quality acceptance standards.

[0085] S60. Construct a multi-objective cost evaluation algorithm, and combine it with the particle swarm optimization algorithm to iteratively optimize the scheduling path and resource allocation ratio of the preliminary scheme, and select the scheme whose total cost is reduced to a set threshold as the final execution scheme.

[0086] S61. Construct a multi-objective cost evaluation algorithm based on transportation costs, time costs, and labor costs.

[0087] Transportation costs are divided into fuel costs, toll fees, and vehicle wear and tear costs. Fuel costs are calculated based on transportation distance, fuel consumption per unit distance of the vehicle, fuel price per unit, and the smoothness of the route. Toll fees are calculated based on the number of toll sections, toll standards, and toll discounts. Vehicle wear and tear costs are determined based on vehicle malfunctions.

[0088] Time costs are determined based on basic transportation time, road congestion time, and the urgency of transportation.

[0089] Labor costs are determined based on the number of workers, working hours, hourly wages, and overtime pay.

[0090] S62. Using the particle swarm optimization algorithm, two optimization dimensions are set: path optimization and resource allocation ratio adjustment. Through iterative calculation, the scheduling scheme is continuously updated to find the solution space with the minimum total cost.

[0091] The resource allocation ratio and transportation route selection in the scheduling scheme are used as the position parameters of the particles. The allocation ratio refers to the proportion of personnel, equipment, and materials allocated to different areas, and the transportation route selection refers to the specific combination of transportation routes between areas. The particle swarm size is set to N. Each particle initializes its position and velocity in a multi-dimensional space, which includes resource allocation and path selection dimensions. The value range of the position parameter is determined based on the actual total resources and the regional road network, while the velocity parameter is set within a reasonable range to control the step size of the particle search.

[0092] The total cost C calculated by the multi-objective cost evaluation algorithm is used as the fitness function of the particles; that is, the lower the total cost of the scheduling scheme corresponding to the particle position, the higher the fitness. The total cost corresponding to each particle position is calculated in each iteration to evaluate the particle's performance.

[0093] In each iteration, each particle updates its velocity and position based on its own historical best position and the group's historical best position. During the update process, the total cost corresponding to the new position is calculated in real time, and the particle's own historical best position and the group's historical best position are updated accordingly. This process continues until the maximum number of iterations is reached or the total cost converges to a satisfactory range, ultimately yielding the particle position with the minimum total cost, i.e., the optimal scheduling scheme.

[0094] The optimized scheduling scheme is then substituted into the cost evaluation algorithm to recalculate the total cost. The scheme whose total cost is reduced by more than a set threshold compared to the initial scheme is taken as the final scheme. If no scheme fully meets the standard, the optimization algorithm parameters are adjusted or the cost evaluation algorithm is re-examined, and optimization and verification are performed again until a reliable scheduling scheme that meets the requirements is obtained.

[0095] Example 2

[0096] like Figure 2As shown, Embodiment 2 of this application provides a cross-regional municipal road maintenance resource scheduling system, including:

[0097] The data acquisition and processing module collects road condition data, maintenance history records, and real-time traffic flow data from various areas through IoT sensors and municipal management system interfaces. It preprocesses the collected data to construct a structured dataset. Specifically, it consists of the following sub-modules:

[0098] Road data submodule: Collects dynamic data on crack development through distributed fiber optic sensors deployed on the road surface, and obtains data on pothole depth and area using lidar on a drone.

[0099] Traffic data submodule: Access the municipal management system database, extract maintenance history records for the past three years, integrate real-time traffic flow interface data from the transportation department, and collect traffic flow and speed information for various areas and road sections.

[0100] The anomaly detection and removal submodule applies box plot analysis to the collected road defect data to identify and remove outliers; for traffic flow data, the isolated forest algorithm is used to detect sudden traffic flow anomalies.

[0101] Emergency Score Module: Based on road defect type, severity, and traffic flow, an analytic hierarchy process (AHP) is used to construct a judgment matrix. The weights of each factor are calculated using the square root method. Based on the quantified baseline data and weights, the emergency score for each area is calculated, classifying the area's maintenance level. Specifically, it is divided into the following sub-modules:

[0102] Factor Determination Submodule: Establish a hierarchical relationship index system for factor analysis to assess the urgency of road maintenance and determine the key assessment factors for roads at each level.

[0103] Factor weighting submodule: Constructs a pairwise comparison judgment matrix based on factor scores, calculates the weight of each factor using the square root method, and ensures the rationality of the weights through consistency checks.

[0104] Sub-module division: Multiply and sum the quantified factor data of each region with the weights to calculate the maintenance urgency score of each region and classify the maintenance level of each region.

[0105] Maintenance Group Division Module: Based on the maintenance level of each area, clustering algorithms are used in conjunction with regional geographical location and road network connectivity to divide maintenance into maintenance groups with different priorities, and a heat map of maintenance priorities is generated. This is specifically divided into the following sub-modules:

[0106] Clustering submodule: Based on the maintenance urgency score as the base variable, regional area and population density are introduced as auxiliary variables. The density clustering algorithm is used to automatically identify densely connected regional clusters and determine the number of clusters.

[0107] Grouping sub-modules: Combine clustering results with geographic information system to analyze the road network connectivity and transportation hub distribution of each region, and prioritize merging regions with directly connected road networks and similar maintenance needs; for regions with ambiguous boundaries, adjust the cluster boundaries through manual intervention.

[0108] Demand Prediction Module: This module analyzes existing maintenance resources in each region, combines them with maintenance group division results, and uses historical maintenance resource consumption data as a benchmark to predict resource demand in each region using regression analysis. It is specifically divided into the following sub-modules:

[0109] The statistical resource submodule uses the municipal asset management system to collect statistics on maintenance resources in each area, establish resource ledgers, and record the status of resource usage.

[0110] Forecasting Submodule: Based on historical maintenance resource consumption data and combined with future regional plans, forecast the resource demand of each region over the next three months.

[0111] The "Determine Shortage Areas" submodule calculates the resource supply-demand ratio for each region, identifies resource-scarce areas, and further analyzes the types of resource gaps.

[0112] Scheduling scheme module: For resource-scarce areas, a greedy algorithm is used to prioritize scheduling paths that are shortest to resource-rich areas and have complementary resource types. Resource scheduling thresholds are set, and a preliminary cross-regional resource scheduling scheme is formulated. Specifically, it is divided into the following sub-modules:

[0113] Matching submodule: Establishes a resource type matching matrix, compares the personnel skills, equipment functions, and material types of resource-scarce areas with redundant areas, calculates the matching degree score, and determines potential scheduling routes by combining the shortest path algorithm between areas.

[0114] Allocation submodule: Starting with the region pair with the highest matching degree and the shortest distance, resources are allocated step by step according to a greedy algorithm. During the allocation process, the resource status of each region is updated in real time to ensure that the remaining resources in the allocated region meet the basic maintenance needs.

[0115] Scheduling submodule: Based on the resource allocation results, a detailed scheduling list is generated and a resource scheduling flowchart is drawn.

[0116] The optimization module constructs a multi-objective cost evaluation algorithm, which, combined with particle swarm optimization, iteratively optimizes the scheduling path and resource allocation ratio of the initial plan, selecting the plan whose total cost is reduced to a set threshold as the final execution plan. Specifically, it is divided into the following sub-modules:

[0117] Cost assessment submodule: Constructs a multi-objective cost assessment algorithm based on transportation costs, time costs, and labor costs.

[0118] Iterative optimization submodule: Using the particle swarm optimization algorithm, it sets two optimization dimensions: path optimization and resource allocation ratio adjustment. Through iterative calculation, it continuously updates the scheduling scheme and finds the solution space with the minimum total cost.

[0119] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method and system for scheduling municipal road maintenance resources across regions, characterized in that, include: S10. Collect road damage data, maintenance history records and real-time traffic flow data in various areas through IoT sensors and municipal management system interfaces, preprocess the collected data and construct a structured dataset. S20. Based on the type, severity and traffic flow of road defects, a judgment matrix is ​​constructed using the analytic hierarchy process. The weight of each factor is calculated using the root method. Based on the quantified basic data and weights, the maintenance urgency score of each area is calculated, and the maintenance level of each area is classified. S30. Based on the maintenance level of each area, use clustering algorithms combined with regional geographical location and road network connectivity to divide maintenance groups with different priorities, and draw a heat map of maintenance priorities. S40. Statistically analyze the existing maintenance resources in each region, combine the maintenance group division results, and use historical maintenance resource consumption data as a benchmark to predict the resource demand in each region using regression analysis. S50. For resource-scarce areas, a greedy algorithm is used to prioritize the scheduling path that is shortest to resource-rich areas and has complementary resource types. A resource scheduling threshold is set, and a preliminary cross-regional resource scheduling plan is formulated. S60. Construct a multi-objective cost evaluation algorithm, and combine it with the particle swarm optimization algorithm to iteratively optimize the scheduling path and resource allocation ratio of the preliminary scheme, and select the scheme whose total cost is reduced to a set threshold as the final execution scheme.

2. The method for scheduling municipal road maintenance resources across regions as described in claim 1, characterized in that, Data on road defects, maintenance history, and real-time traffic flow in various areas are collected through IoT sensors and municipal management system interfaces. The collected data is preprocessed to construct a structured dataset. This process involves the following sub-steps: Dynamic data on crack development are collected by distributed fiber optic sensors deployed on the road surface, and the depth and area data of potholes are obtained by scanning with lidar carried by drones. Access the municipal management system database, extract maintenance history records for the past three years, integrate real-time traffic flow interface data from the transportation department, and collect traffic flow and speed information for various areas and road sections. Box plot analysis was applied to the collected road defect data to identify and remove outliers. For traffic flow data, the isolated forest algorithm was used to detect sudden traffic flow anomalies.

3. The method for scheduling municipal road maintenance resources across regions as described in claim 1, characterized in that, Based on road defect type, severity, and traffic flow, an analytic hierarchy process (AHP) is used to construct a judgment matrix. The weights of each factor are calculated using the root mean square method. Based on the quantified baseline data and weights, a maintenance urgency score for each area is calculated, thus classifying the area into maintenance levels. The specific steps are as follows: Establish a hierarchical index system of factors to assess the urgency of road maintenance and determine the key assessment factors for roads at each level. A pairwise comparison judgment matrix is ​​constructed based on factor scoring, and the weight of each factor is calculated using the square root method. The rationality of the weights is ensured through consistency checks. The quantified factor data for each region are multiplied and summed with their weights to calculate the maintenance urgency score for each region, thus classifying the maintenance levels of each region.

4. The method for scheduling municipal road maintenance resources across regions as described in claim 1, characterized in that, Based on the maintenance level of each region, a clustering algorithm is used, combined with regional geographical location and road network connectivity, to divide maintenance into groups with different priorities, and a heat map of maintenance priorities is drawn. Specifically, this involves the following sub-steps: Using the maintenance urgency score as the basic variable, and introducing regional area and population density as auxiliary variables, a density clustering algorithm is used to automatically identify densely connected regional clusters and determine the number of clusters. By combining clustering results with geographic information systems, we can analyze the road network connectivity and transportation hub distribution in each region. We can prioritize merging regions with directly connected road networks and similar maintenance needs. For regions with ambiguous boundaries, we can adjust the cluster boundaries through manual intervention.

5. The method for scheduling municipal road maintenance resources across regions as described in claim 1, characterized in that, The existing maintenance resources in each region are statistically analyzed. Combined with the maintenance group division results, and based on historical maintenance resource consumption data, regression analysis is used to predict the resource demand in each region. This process is divided into the following sub-steps: Through the municipal asset management system, the maintenance resources in each area are counted, a resource ledger is established, and the status of resource usage is recorded. Based on historical maintenance resource consumption data and combined with future regional plans, the resource demand of each region is predicted for the next three months. Calculate the resource supply-demand ratio for each region, identify resource-scarce areas, and further analyze the types of resource gaps.

6. The method for scheduling municipal road maintenance resources across regions as described in claim 1, characterized in that, For resource-scarce areas, a greedy algorithm is used to prioritize scheduling paths that are shortest to resource-rich areas and have complementary resource types. A resource scheduling threshold is set, and a preliminary cross-regional resource scheduling scheme is formulated. This involves the following sub-steps: Establish a resource type matching matrix, compare the personnel skills, equipment functions, and material types of resource-scarce areas and redundant areas, calculate the matching degree score, and combine the shortest path algorithm between areas to determine potential scheduling routes. Starting with the region pair with the highest matching degree and the shortest distance, resources are allocated step by step according to the greedy algorithm. During the allocation process, the resource status of each region is updated in real time to ensure that the remaining resources in the allocated region meet the basic maintenance needs. Based on the resource allocation results, a detailed scheduling list is developed and a resource scheduling flowchart is drawn.

7. The method for scheduling municipal road maintenance resources across regions as described in claim 1, characterized in that, A multi-objective cost evaluation algorithm is constructed, which is combined with particle swarm optimization to iteratively optimize the scheduling path and resource allocation ratio of the initial plan, and the plan with the total cost reduced to a set threshold is selected as the final execution plan. Specifically, this involves the following sub-steps: A multi-objective cost assessment algorithm is constructed based on transportation costs, time costs, and labor costs; By employing the particle swarm optimization algorithm, two optimization dimensions are set: path optimization and resource allocation ratio adjustment. Through iterative calculation, the scheduling scheme is continuously updated to find the solution space with the minimum total cost.

8. A cross-regional municipal road maintenance resource scheduling system, characterized in that, include: Data Acquisition and Processing Module: Collects road damage data, maintenance history records, and real-time traffic flow data in various areas through IoT sensors and municipal management system interfaces; preprocesses the collected data; and constructs a structured dataset. Emergency Score Module: Based on road defect type, severity, and traffic flow, an analytic hierarchy process is used to construct a judgment matrix. The weight of each factor is calculated using the square root method. Based on the quantified basic data and weights, the maintenance urgency score of each area is calculated, and the maintenance level of each area is classified. Maintenance group division module: Based on the maintenance level of each area, clustering algorithms are used to combine regional geographical location and road network connectivity to divide maintenance groups with different priorities, and a heat map of maintenance priority is drawn. Demand prediction module: Statistics on existing maintenance resources in each region, combined with the results of maintenance group division, and based on historical maintenance resource consumption data, regression analysis is used to predict the resource demand in each region; Scheduling scheme module: For resource-scarce areas, a greedy algorithm is used to prioritize the scheduling path that is shortest to resource-rich areas and has complementary resource types. Resource scheduling thresholds are set, and a preliminary cross-regional resource scheduling scheme is formulated. Optimization Scheme Module: Construct a multi-objective cost evaluation algorithm, and combine it with the particle swarm optimization algorithm to iteratively optimize the scheduling path and resource allocation ratio of the preliminary scheme, and select the scheme whose total cost is reduced to a set threshold as the final execution scheme.

Images