Municipal facility inspection scheduling method and system based on city steward

Abstract

The invention relates to the technical field of urban municipal facility management, and discloses a municipal facility inspection scheduling method and system based on urban large housekeepers. The method comprises the steps of obtaining municipal facility data and performing structured processing, constructing a space-time grid coordinate system, mapping and generating facility space-time coordinate information, generating an initial inspection path and performing optimization according to traffic and fault probability, splitting tasks and performing reasonable distribution, acquiring progress in real time and then optimizing the path, issuing an instruction and receiving feedback, and updating data. The system comprises a plurality of modules for data acquisition and processing and the like to realize the method. According to the invention, facility data can be integrated, an inspection path can be optimized, tasks can be reasonably distributed, sudden faults can be quickly responded, fault prediction accuracy can be improved according to feedback, municipal facility inspection scheduling efficiency and quality can be effectively improved, and stable operation of urban facilities can be guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of urban municipal facility management, and in particular to a municipal facility inspection and dispatching method and system based on a city steward. Background Art

[0002] With the acceleration of urbanization, the scale of cities continues to expand, and the number and types of municipal facilities have increased dramatically. These facilities include roads, bridges, drainage systems, lighting equipment, etc., which are the basic guarantee for the normal operation of cities. However, the traditional inspection and dispatching methods of municipal facilities have gradually exposed many problems and are difficult to meet the needs of modern urban management.

[0003] In terms of data management, the basic information and real-time status data of municipal facilities are often scattered in different systems and departments, lacking effective integration and structured processing. For example, the geographic coordinate information of the facility may be managed by the surveying and mapping department, while the maintenance cycle information is controlled by the operation and maintenance department. The data formats and standards of each department are not unified, making it difficult to share and comprehensively utilize data. This makes it impossible to quickly and accurately obtain comprehensive facility information during inspection and scheduling, affecting the scientificity and timeliness of decision-making.

[0004] In terms of inspection route planning, traditional methods are mostly based on experience or simple geographical order, without fully considering dynamic factors such as real-time traffic conditions and the probability of facility failure. During traffic congestion, inspections along fixed routes will lead to low inspection efficiency, waste a lot of time on the road, and fail to reach the facilities that need to be inspected in time. At the same time, due to ignoring the probability of facility failure, some areas with frequent failures cannot receive timely attention, increasing the risk of sudden failures of facilities and affecting the normal operation of the city.

[0005] In the task allocation process, the traditional method is usually to divide the tasks according to the general area of ​​the inspection personnel or equipment, without fully considering the personnel's skill level, the equipment load status and the urgency of the task. This may lead to unskilled personnel taking on complex inspection tasks and being unable to discover and deal with facility problems in a timely manner; or the equipment is overloaded, affecting the inspection quality and efficiency, and even causing equipment damage. Moreover, the inability to respond to emergency tasks in a timely manner delays the best time for facility maintenance.

[0006] In addition, when encountering sudden facility failures, the traditional inspection and dispatching system lacks a fast and effective emergency response mechanism. The inability to quickly adjust the inspection plan and handle the failure in a timely manner may cause the scope of the failure to expand, causing greater losses to the city. At the same time, due to the lack of feedback and analysis on the inspection results, it is difficult to optimize and improve the inspection and dispatching methods, which makes the problem persist for a long time and cannot be fundamentally solved. Summary of the invention

[0007] The purpose of the present invention is to provide a municipal facility inspection and scheduling method and system based on the urban steward to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions: A municipal facility inspection and scheduling method based on the urban steward, the method includes: Obtain the basic information and real-time status data of municipal facilities, structurally process the basic information according to preset attribute dimensions, and generate facility feature data; Construct a spatio-temporal grid coordinate system, divide grid cells based on the urban geographical area, define the X-axis of the spatio-temporal grid coordinate system as the time axis, the Y-axis as the space axis, the Z-axis as the facility type axis, and assign dynamic weights to each grid cell; Map the facility feature data into the corresponding spatio-temporal grid coordinate system to generate facility spatio-temporal coordinate information; Based on the preset inspection task requirements, extract the type, geographical location and priority of the inspection target facilities to generate an initial inspection path; According to the real-time traffic flow data and the facility failure probability prediction model, dynamically adjust the weights of each grid cell in the initial inspection path, and generate an optimized inspection path based on the greedy algorithm; Split the optimized inspection path into multiple inspection subtasks, and allocate the inspection subtasks according to the real-time location and load status of the inspection personnel or equipment; Real-time collect the execution progress data of the inspection subtasks, and combine with the dynamic weights of the spatio-temporal grid coordinate system to re-optimize the paths of the uncompleted inspection subtasks; Send the re-optimized inspection path and task allocation instructions to the corresponding inspection terminals, and receive the inspection result data feedback from the terminals; Update the facility feature data and the dynamic weights of the spatio-temporal grid coordinate system according to the inspection result data.

[0009] Preferably, the preset attribute dimensions include facility type, geographical coordinates, maintenance cycle and historical failure records.

[0010] Preferably, the dynamically adjusting the weights of each grid cell in the initial inspection path includes: Obtain the real-time traffic congestion index and weather data, and calculate the travel time cost of each grid cell; Predict the failure probability of each facility based on the facility historical failure records and the machine learning model; Dynamically update the weights of the grid cells according to the product of the travel time cost and the failure probability.

[0011] Preferably, the machine learning model is a LSTM network based on time series, the input data includes facility operating parameters, environmental data and historical maintenance records, and the output is the failure probability within a preset time period in the future.

[0012] Preferably, the allocating the inspection subtasks includes: Construct a priority scoring matrix based on the skill level, current location and urgency of the inspection personnel or equipment; The Hungarian algorithm is used to optimally match the scoring matrix to determine the allocation results of the inspection subtasks.

[0013] Preferably, the method further comprises: Collaborative planning is performed for multiple inspection tasks in the same geographical area. The path intersections between multiple tasks are optimized by introducing the ant colony algorithm to reduce repeated inspection areas.

[0014] Preferably, the method further comprises: A multi-objective optimization model is constructed with minimizing the total inspection time, balancing the load distribution and maximizing the fault coverage as the optimization goals. The NSGA-II algorithm is used to solve the Pareto optimal solution set, and the final inspection plan is selected based on the preset rules.

[0015] Preferably, the method further comprises: When a sudden facility failure is detected, the current inspection route is interrupted, an emergency inspection task is inserted based on the time-space coordinates and priority of the faulty facility, and the optimized route is recalculated.

[0016] Preferably, the method further comprises: An inspection effect feedback strategy is established to dynamically correct the parameters of the facility failure probability prediction model based on the correlation between historical inspection data and facility status changes.

[0017] Preferably, the present invention also includes a municipal facility inspection and dispatching system based on the city's big housekeeper, which is used to implement the above-mentioned municipal facility inspection and dispatching method based on the city's big housekeeper, and the system includes: Data collection and processing module: used to obtain basic information and real-time status data of municipal facilities, structure the basic information according to preset attribute dimensions, and generate facility feature data; Space-time grid construction module: construct a space-time grid coordinate system, divide the grid units based on the urban geographical area, define the X-axis of the space-time grid coordinate system as the time axis, the Y-axis as the space axis, and the Z-axis as the facility type axis, and assign dynamic weights to each grid unit; Mapping module: maps the facility feature data to the corresponding space-time grid coordinate system to generate facility space-time coordinate information; Initial path generation module: Based on the requirements of a preset inspection task, extract the type, geographical location, and priority of the inspection target facilities, and generate an initial inspection path; Path optimization module: According to the real-time traffic flow data and the facility failure probability prediction model, dynamically adjust the weights of each grid cell in the initial inspection path, and generate an optimized inspection path based on the greedy algorithm; Task assignment module: Split the optimized inspection path into multiple inspection subtasks, and assign the inspection subtasks according to the real-time location and load status of the inspection personnel or equipment; Monitoring and re-optimization module: Real-time collect the execution progress data of the inspection subtasks, and combine with the dynamic weights of the spatio-temporal grid coordinate system to re-optimize the paths of the unfinished inspection subtasks; Instruction issuing and feedback receiving module: Issue the re-optimized inspection path and task assignment instructions to the corresponding inspection terminals, and receive the inspection result data feedback by the terminals; Data update module: Update the facility feature data and the dynamic weights of the spatio-temporal grid coordinate system according to the inspection result data.

[0018] Compared with the prior art, the beneficial effects of the present invention are: In terms of data processing and management, by obtaining the basic information and real-time status data of municipal facilities, and performing structured processing according to the preset attribute dimensions to generate facility feature data, the effective integration and standardized management of data are realized. This enables the centralized storage and sharing of various facility information, facilitating the management department to query and call at any time. For example, when it is necessary to maintain the street lights in a certain area, information such as the geographical coordinates, maintenance cycle, and historical failure records of the street lights can be quickly obtained, providing comprehensive and accurate data support for formulating the maintenance plan and avoiding management difficulties caused by scattered data and inconsistent formats.

[0019] The construction of the spatio-temporal grid coordinate system and the mapping of the facility feature data provide an intuitive spatio-temporal perspective for facility management. By defining the X-axis as the time axis, the Y-axis as the space axis, the Z-axis as the facility type axis, and assigning dynamic weights to each grid cell, the distribution and importance of facilities in the urban spatio-temporal range can be clearly displayed. This helps managers quickly understand the real-time status of different types of facilities in different regions, providing a visual basis for inspection scheduling decisions. For example, by observing the spatio-temporal grid coordinate system, it can be quickly found that facilities in a certain area frequently fail, so as to adjust the inspection plan in time, increasing the inspection frequency and priority of this area.

[0020] In terms of inspection route planning, the present invention generates an initial inspection route based on preset inspection task requirements, and dynamically adjusts the weight of each grid unit in the route according to real-time traffic flow data and facility failure probability prediction model, and then uses a greedy algorithm to generate an optimized inspection route. This approach fully considers traffic conditions and facility failure risks, and effectively improves inspection efficiency. Taking urban road inspections as an example, during the morning rush hour, the system will avoid congested sections and select smooth routes for inspections based on the real-time traffic congestion index, reducing the time wasted by inspectors on the road and allowing them to have more time for facility inspections. At the same time, due to the priority inspection of facilities with a high probability of failure, potential problems can be discovered and dealt with in a timely manner, reducing the possibility of sudden facility failures and ensuring the normal operation of urban facilities.

[0021] In the task allocation stage, the optimized inspection path is split into multiple inspection subtasks, and they are reasonably allocated according to the real-time location, load status, and skill level of the inspection personnel or equipment. This ensures that each inspection task can be undertaken by the most suitable personnel or equipment, improving the quality and efficiency of the inspection. For example, for power facility inspection tasks with higher technical difficulty, priority will be given to inspection personnel with high skill levels and rich experience; for inspection personnel or equipment with closer distances and lower loads, nearby inspection subtasks will be assigned, reducing the time cost of task execution. Moreover, by constructing a priority scoring matrix and using the Hungarian algorithm for optimal matching, the intelligent and scientific nature of task allocation is achieved, avoiding the irrationality of traditional allocation methods.

[0022] When a sudden facility failure is detected, the system can interrupt the current inspection route, insert an emergency inspection task, and recalculate the optimized route. This mechanism ensures a quick response to sudden failures, and can handle the failures in a timely manner to prevent the scope of the failure from expanding. For example, in the urban drainage system, once a blockage is detected in a certain section of the pipeline, the system will immediately adjust the inspection plan, arrange the nearest inspection personnel to handle it, and replan the routes of other inspection tasks to ensure the normal operation of the entire drainage system and reduce the impact of waterlogging on urban traffic and residents' lives.

[0023] In addition, the established inspection effect feedback strategy dynamically corrects the parameters of the facility failure probability prediction model based on the correlation between historical inspection data and facility status changes. This continuously improves the prediction accuracy of the model and further optimizes the inspection scheduling plan. As time goes by, the system will be able to more accurately predict facility failures, schedule inspection tasks in advance, achieve preventive maintenance, reduce facility maintenance costs, and extend the service life of facilities, thereby comprehensively improving the management level and operating efficiency of urban municipal facilities and providing strong guarantees for the sustainable development of cities. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1This is a working principle diagram of the municipal facility inspection and dispatching method of the present invention; Figure 2 Flow chart of grid weight adjustment and path optimization based on traffic and failure probability; Figure 3 Flowchart for the allocation of subtasks for municipal facility inspection; Figure 4 Flowchart for collaborative planning of multiple inspection tasks in the same geographical area. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] See also Figure 1-Figure 4 The present invention provides a municipal facility inspection and dispatching method based on the city manager, and the specific implementation steps are as follows: By connecting with the municipal facility management system and collecting data in real time using sensors, we can obtain basic information and real-time status data of municipal facilities. Basic information covers the name, model, construction time and other aspects of the facility. These basic information are structured according to preset attribute dimensions, including facility type, geographic coordinates, maintenance cycle and historical fault records. For example, for a bridge, the facility type is a bridge, the geographic coordinates are accurate to longitude and latitude, the maintenance cycle may be once every six months, and the historical fault records include bridge deck cracks, bridge pier displacement and other situations that have occurred. After processing, the facility feature data is generated for subsequent analysis.

[0027] Based on the city's geographical area, the city is divided into multiple grid units. A spatiotemporal grid coordinate system is defined, where the X-axis is the time axis, which is measured in hours, days, etc., and is used to record information at different time points; the Y-axis is the space axis, which corresponds to the city's geographical spatial location; and the Z-axis is the facility type axis, which distinguishes different types of municipal facilities, such as road facilities, lighting facilities, drainage facilities, etc. Dynamic weights are assigned to each grid unit, and the initial value of the weight can be determined based on factors such as the importance of the facilities in the grid unit and the historical failure frequency.

[0028] Map the facility feature data generated previously to the corresponding space-time grid coordinate system. Determine the location of the facility on the Y axis based on its geographic coordinates, determine its location on the Z axis based on the facility type, and then determine its location on the X axis based on the current time or planned inspection time, thereby generating facility space-time coordinate information. This information can intuitively display the distribution of facilities within the city's space-time range.

[0029] Based on the preset inspection task requirements, the type, geographical location and priority of the inspection target facilities are extracted from the facility feature data. For example, power facilities in certain key areas have a higher priority and need to be inspected first. Based on this information, the existing path planning algorithm is used to generate the initial inspection path to ensure that all facilities that need to be inspected are covered.

[0030] Real-time traffic flow data can be obtained from the traffic management department or collected through sensors installed on the road. At the same time, the failure probability prediction model of each facility is used to predict the failure probability of each facility. The travel time cost of each grid unit is calculated based on the real-time traffic flow data, and the weight of each grid unit in the initial inspection path is dynamically adjusted in combination with the failure probability. Then, based on the greedy algorithm, each time the next inspection point is selected, the grid unit with the best current weight is selected, thereby generating an optimized inspection path and improving inspection efficiency.

[0031] The optimized inspection path is divided into multiple inspection subtasks. These inspection subtasks are assigned according to the real-time location and load status of the inspection personnel or equipment. For example, if an inspection personnel is currently near a grid unit and its load is low, the inspection subtasks near this area are assigned first to ensure the rationality and efficiency of task allocation.

[0032] The execution progress data of the inspection subtasks is collected in real time through the positioning devices and feedback systems carried by the inspection personnel or equipment. The paths of the unfinished inspection subtasks are re-optimized by combining the dynamic weights of the space-time grid coordinate system. If a certain area suddenly experiences traffic congestion, resulting in the obstruction of the originally planned path, the path is replanned to avoid the congested area and ensure that the inspection task is completed on time.

[0033] Instruction issuance and result feedback: The optimized inspection path and task allocation instructions are sent to the corresponding inspection terminal through wireless communication technology, such as the handheld device of the inspection personnel or the on-board system of the inspection vehicle. After receiving the instructions, the inspection terminal executes the inspection task and feeds back the inspection result data. The inspection result data includes the actual status of the facility, whether a fault is found, and other information.

[0034] Update the dynamic weight of the facility feature data and the space-time grid coordinate system based on the inspection results. If a facility is found to be faulty, update the historical fault record of the facility; if the facilities in a grid unit frequently fail, increase the weight of the grid unit appropriately so that the area can be inspected more intensively in the future.

[0035] The present invention will be further described below in conjunction with Examples 1 to 6: Example 1: After obtaining the basic information and real-time status data of municipal facilities, structured processing is performed according to the preset attribute dimensions. Taking street lamp facilities as an example, the facility type is clearly a lighting facility; the geographic coordinates are obtained through the GPS positioning module installed on the street lamp, accurate to six decimal places to ensure the accuracy of positioning; the maintenance cycle is set according to the service life of the street lamp, the manufacturer's recommendations and local maintenance standards, generally a comprehensive maintenance is performed once a year; the historical fault record records the type of fault, fault time, maintenance status and other information that has occurred in the street lamp in detail. By combing and integrating these attributes, comprehensive and targeted facility feature data is generated.

[0036] When dynamically adjusting the weight of each grid unit in the initial inspection path, obtain the real-time traffic congestion index and weather data. The real-time traffic congestion index can be obtained from the traffic big data platform. The index reflects the degree of road congestion, with a value range of 0-10. The larger the value, the more serious the congestion. Weather data can be obtained from the interface of the meteorological department, including weather conditions (such as sunny, rainy, snowy, etc.), temperature, humidity and other information.

[0037] Calculate the travel time cost of each grid unit using the formula: ,in represents the travel time cost of the grid cell, It is the basic time required to pass through the grid cell under normal traffic conditions, calculated based on historical traffic data; It is the real-time traffic congestion index; is the weather factor, sunny , when it rains , snowy days and other bad weather For example, the Real-time traffic congestion index for 10 minutes , the current weather is rainy, then the travel time cost of this grid unit minute.

[0038] The failure probability of each facility is predicted based on the historical failure records of the facility and the machine learning model. This embodiment uses a time series-based LSTM network as a machine learning model. The input data includes facility operation parameters (such as the voltage and current of street lamps), environmental data (temperature, humidity, light intensity, etc.) and historical maintenance records. Among them, the facility operation parameters are collected in real time by sensors installed on street lamps; environmental data is obtained by nearby meteorological stations or environmental monitoring equipment; historical maintenance records are extracted from the municipal facility maintenance management database. The output is the failure probability within a preset time period in the future. The preset time period can be set according to actual conditions, such as the next week or month.

[0039] The weight of the grid unit is dynamically updated according to the product of the travel time cost and the failure probability. Assume that the travel time cost of a street light in a grid unit is Minutes, the LSTM network is used to predict the probability of failure within the next week. , then the weight update value of the grid unit is If the initial weight of the grid unit is 5, the updated weight is In this way, traffic factors and facility failure risks are comprehensively considered, and grid unit weights are reasonably adjusted to provide a more accurate basis for optimizing inspection routes.

[0040] Embodiment 2: In this embodiment, the specific implementation process of the inspection subtask allocation is emphasized.

[0041] After the optimized inspection path is generated, it is divided into multiple inspection subtasks. In order to reasonably allocate these inspection subtasks, it is necessary to build a priority scoring matrix based on the skill level, current location and task urgency of the inspection personnel or equipment.

[0042] The skill level of the inspection personnel or equipment is divided according to their training experience, work experience and professional qualifications, for example, into three levels: primary, intermediate and senior, corresponding to different skill scores, 3 points for primary, 5 points for intermediate and 8 points for senior. This quantitative method can clarify the professional ability level of different inspection personnel or equipment.

[0043] The current location is obtained through the positioning device installed on the inspection personnel or equipment, which is accurate to the specific grid unit. The urgency of the task is determined according to the priority and failure probability of the facility. For inspection tasks corresponding to facilities with high priority and high failure probability, the urgency is set to high and assigned a value of 8 points; tasks with high priority but relatively low failure probability have a medium urgency and are assigned a value of 5 points; tasks with low priority and low failure probability have a low urgency and are assigned a value of 3 points.

[0044] When constructing the priority scoring matrix, the inspection personnel or equipment are used as rows and the inspection subtasks are used as columns. The element values ​​in the matrix are calculated based on the skill level of the inspection personnel or equipment, the distance between the current location and the task location, and the urgency of the task. The calculation method is: ,in represents the element value in the scoring matrix, Score the skill level of inspection personnel or equipment. is the distance between the current location of the inspection personnel or equipment and the grid unit where the task is located (the distance is calculated by the coordinates of the grid unit and the unit is kilometers). Rate the urgency of the task.

[0045] For example, there is patrol inspector A, whose skill level is intermediate ( ), currently located in grid unit (1,2), there is an inspection subtask located in grid unit (3, 4), distance kilometers, the mission urgency is high , then the score of inspector A for this inspection subtask is point.

[0046] The Hungarian algorithm is used to optimally match the scoring matrix. The Hungarian algorithm is a classic algorithm for solving the assignment problem. Its core idea is to transform the original problem into an equivalent matching problem by continuously searching for augmenting paths, thereby finding the optimal allocation solution. In this embodiment, the Hungarian algorithm is used to process the constructed priority scoring matrix, which can quickly and accurately determine the allocation results of the inspection subtasks, ensure that each inspection subtask can be assigned to the most suitable inspection personnel or equipment, and improve the overall efficiency and quality of the inspection work.

[0047] Embodiment 3: During the inspection of urban municipal facilities, there are often multiple inspection tasks in the same geographical area. This embodiment introduces in detail a specific method for collaborative planning of multiple inspection tasks in the same geographical area.

[0048] First, collect information on multiple inspection tasks in the same geographic area, including the target facility type, geographic location, priority, and estimated inspection time of each task. For example, in a downtown area, there are three inspection tasks. Task A is to inspect the road facilities in the area, including multiple intersections and sections; Task B is to check the street lighting facilities; Task C is to check whether the drainage pipes are blocked.

[0049] The ant colony algorithm is introduced to optimize the intersection points of paths between multiple tasks and reduce repeated inspection areas. The ant colony algorithm is a heuristic algorithm that simulates the foraging behavior of ants. Ants leave pheromones on the path in the process of looking for food. The higher the pheromone concentration, the greater the probability of being selected by other ants.

[0050] In this embodiment, each inspection task is regarded as an ant, and the starting point and end point of the task are used as the starting point and target point of the ant. At the beginning of the algorithm, an initial path is randomly assigned to each inspection task. During the iteration process, the ant releases pheromones on the path, and the update formula of the pheromone is: ,in Indicated in Time Path The pheromone concentration on is the pheromone volatility coefficient, ranging from It is usually set to about 0.5 to simulate the natural volatilization of pheromones over time; is the number of ants, that is, the number of inspection tasks; Indicates Only ants in Always on the path The amount of pheromone released is calculated as: , It is a constant used to control the release intensity of pheromones, and its value is generally 100; It is The length of the path traveled by the ants.

[0051] When the ant chooses the next path point, it makes a decision based on the pheromone concentration and heuristic information on the path. The selection probability formula is: ,in Indicates Only ants in Time from node Select Node probability; and are two parameters, representing the relative importance of pheromone and heuristic information, respectively. , ; is the heuristic information, generally taken as the path The reciprocal of the length of , Is a node and nodes The distance between It is The set of nodes that the ant can currently select.

[0052] Through continuous iteration, ants will gradually find a better path, which reduces the intersection points of multiple inspection tasks and the repeated inspection areas. For example, after multiple iterations, Task A and Task B have repeated inspections on a certain road section. After optimization by the ant colony algorithm, the paths of the two tasks are reasonably planned, avoiding repeated inspections of the road section, effectively improving inspection efficiency and reducing manpower and time costs.

[0053] Example 4: This example describes in detail the process of constructing a multi-objective optimization model and solving it using the NSGA-II algorithm. The specific method is: A multi-objective optimization model is constructed with minimizing the total inspection time, balancing the load distribution and maximizing the fault coverage as the optimization goals.

[0054] Minimize the total inspection time, that is, the total time required to complete all inspection tasks is the shortest. Suppose the inspection task set is , each task The inspection time is , then the total inspection time In actual calculation, It can be determined based on factors such as the complexity of the task, the number of target facilities, and the expected inspection speed. For example, for an inspection task involving 10 street lights, the expected inspection time for each street light is 5 minutes, and the intermediate movement time is 10 minutes, then the inspection time of the task is minute.

[0055] Balanced load distribution means ensuring that the workload of each inspector or device is relatively balanced, avoiding the situation where some inspectors have too much work to do while others have too little work to do. , each inspector or equipment The set of tasks undertaken is , the workload it undertakes , then the load balance can be measured by calculating the standard deviation of the workload. The standard deviation formula is: ,in It is the average workload of all inspection personnel or equipment. The smaller the standard deviation, the more balanced the load distribution.

[0056] Maximizing fault coverage requires discovering as many potential facility faults as possible. Suppose the set of faulty facilities is , the set of faulty facilities found by inspection is , then the fault coverage ,in and Respectively represent sets and The number of elements.

[0057] The NSGA-II algorithm is used to solve the Pareto optimal solution set. The NSGA-II algorithm is an efficient multi-objective optimization algorithm, and its main steps include initializing the population, calculating individual fitness, selection, crossover and mutation, etc.

[0058] When the population is initialized, a certain number of inspection plans are randomly generated, each of which contains information such as inspection path planning and task allocation. The individual fitness is calculated, that is, each inspection plan is evaluated according to the above three optimization goals to obtain the fitness value corresponding to each goal.

[0059] The selection operation uses the tournament selection method to randomly select a certain number of individuals from the population for comparison, and select individuals with better fitness to enter the next generation. The crossover operation exchanges genes of the selected individuals to generate new individuals. The mutation operation randomly changes the genes of individuals to increase the diversity of the population.

[0060] Through continuous iteration, the algorithm will gradually converge to the Pareto optimal solution set. The Pareto optimal solution set is a set of non-dominated solutions, in which there is no solution that is superior to other solutions in all objectives. For example, if solution A has a shorter total inspection time than solution B, but a lower fault coverage rate than solution B, and there is no other solution that can be superior to A and B in both of these two objectives at the same time, then both A and B belong to the Pareto optimal solution set.

[0061] Select the final inspection plan based on preset rules. The preset rules can be set according to actual needs. For example, when paying more attention to the fault coverage rate, a plan with a high fault coverage rate, relatively good total inspection time and load balancing degree can be preferentially selected; if the current manpower is limited and more attention is paid to load balancing, a plan with a good load balancing degree can be preferentially selected. In this way, the most practical final inspection plan is selected from the Pareto optimal solution set to improve the comprehensive effect of municipal facility inspection scheduling.

[0062] Example 5: During the inspection of municipal facilities, it is inevitable to encounter sudden facility failures. This example details the processing flow for dealing with sudden facility failures.

[0063] Through a real-time monitoring system, such as sensors installed on municipal facilities, citizen feedback platforms, etc., sudden facility failures are detected in a timely manner. When the system receives the fault information, it will immediately obtain the spatio-temporal coordinates and priority of the faulty facility.

[0064] Taking the drainage pipe system of a certain city as an example, assume that during the inspection process, the pressure sensor of a drainage pipe located in a certain section suddenly detects that the pressure has risen abnormally, far exceeding the normal range, and the system determines that the drainage pipe is blocked. At this time, the system quickly locates the spatio-temporal coordinates of the drainage pipe. For example, in the spatio-temporal grid coordinate system, its Y-axis coordinate represents the specific geographical location of the pipe (such as a specific section of a certain street), the Z-axis coordinate indicates its drainage facility type, and the X-axis coordinate is the time point when the fault occurs. At the same time, its priority is determined according to factors such as the importance and influence range of the area where the drainage pipe is located. If the pipe is located in the central business district of the city and a blockage may cause serious waterlogging, affecting traffic and commercial activities, then its priority is set to high.

[0065] After detecting a sudden facility failure, the system immediately interrupts the current inspection path. Taking the inspection personnel who are performing the comprehensive inspection task in this area as an example, the inspection personnel were originally inspecting road facilities, lighting facilities, etc. according to the established optimized inspection path. After receiving the fault alarm, the inspection terminal in their hands will receive an interruption instruction and stop the current inspection task being carried out.

[0066] Based on the time and space coordinates and priority of the faulty facility, an emergency inspection task is inserted. The system will re-plan the route, giving priority to the location and priority of the faulty facility. When planning the route, the shortest path from the inspector's current location to the location of the faulty facility is calculated, while avoiding traffic congestion. For example, if real-time traffic data shows that a regular road leading to a faulty drainage pipe is congested, the system will automatically select another road that is slightly longer but has smooth traffic as a temporary inspection route.

[0067] When recalculating the optimized path, the system will take emergency inspection tasks into overall consideration. It not only takes into account the path to the faulty facility, but also plans the subsequent inspection path after the fault is resolved. Assuming that the inspector needs to continue to complete the remaining inspection tasks after the fault is resolved, the system will recalculate an optimized inspection path based on the time, traffic conditions, location and priority of the remaining facilities at this time to ensure that subsequent inspection tasks can be completed efficiently. In the process of replanning the path, the weights of the relevant grid cells in the space-time grid coordinate system will also be updated in real time. For example, the weight of the grid cell in the fault area will increase due to this fault, so that more attention can be paid to this area in the future.

[0068] Example 6: Establishing an inspection effect feedback strategy can continuously improve the accuracy of the facility failure probability prediction model, thereby optimizing the municipal facility inspection scheduling. The specific method is: Collect historical inspection data, including the time, location, type of facilities involved, and status of facilities (normal, faulty, and specific fault types) of each inspection. For example, in the street light inspection records of the past year, the specific time of each inspection was recorded in detail, accurate to the minute; the specific location of the street light, accurate to the street house number; the type of street light, such as high-pressure sodium lamp, LED lamp, etc.; the status of the street light at the time, whether it was normally lit, partially damaged, or completely extinguished, and if there was any damage, the specific cause of the damage was recorded, such as burnt bulb, short circuit, etc.

[0069] At the same time, data on facility status changes within the corresponding time period are collected. Taking street lights as an example, in addition to the status recorded during inspections, data is also collected on whether street lights have experienced any malfunctions reported by citizens between two inspections, as well as the time and results of handling these malfunctions. By analyzing these data, the correlation between historical inspection data and facility status changes is established.

[0070] When analyzing the correlation, data mining and statistical analysis methods are used. For example, through statistics, it is found that in a certain area, when the ambient temperature exceeds 35°C continuously and the operating time reaches 5,000 hours, the probability of the bulb burning out increases significantly. This shows that there is a certain correlation between ambient temperature and operating time and street light bulb failure.

[0071] According to the above correlation, the parameters of the facility failure probability prediction model are dynamically corrected. Assume that a time series-based LSTM network is used as the facility failure probability prediction model, and its input data includes facility operation parameters, environmental data, and historical maintenance records. According to the correlation obtained by analysis, the model's weight, bias and other parameters are adjusted. If it is found that the ambient temperature has a greater impact on street lamp failure, then the weight of the input feature of ambient temperature is increased accordingly in the model, so that the model can more accurately reflect the impact of environmental factors when predicting the probability of street lamp failure.

[0072] In actual operation, the model parameters are regularly revised. For example, the correlation is re-analyzed and the model parameters are adjusted every month based on the newly collected historical inspection data and facility status change data. Through continuous feedback and revision, the accuracy of the facility failure probability prediction model is gradually improved, thereby providing a more reliable basis for the inspection and scheduling of municipal facilities. This helps to plan inspection tasks more reasonably, detect potential failures in advance, reduce the adverse effects of facility failures, and improve the operating efficiency and reliability of urban municipal facilities.

[0073] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0074] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A municipal facility inspection and dispatching method based on the city's big housekeeper, characterized in that: include: Obtain basic information and real-time status data of municipal facilities, structure the basic information according to preset attribute dimensions, and generate facility feature data; Constructing a space-time grid coordinate system, dividing grid units based on the urban geographical area, defining the X-axis of the space-time grid coordinate system as the time axis, the Y-axis as the space axis, and the Z-axis as the facility type axis, and assigning a dynamic weight to each grid unit; Mapping the facility characteristic data to a corresponding space-time grid coordinate system to generate facility space-time coordinate information; Based on the preset inspection task requirements, the type, geographical location and priority of the inspection target facilities are extracted to generate the initial inspection route; According to the real-time traffic flow data and the facility failure probability prediction model, the weight of each grid unit in the initial inspection path is dynamically adjusted, and the optimized inspection path is generated based on the greedy algorithm; Splitting the optimized inspection path into multiple inspection subtasks, and allocating the inspection subtasks according to the real-time location and load status of the inspection personnel or equipment; Collect the execution progress data of the inspection subtasks in real time, and re-optimize the paths of the unfinished inspection subtasks in combination with the dynamic weights of the space-time grid coordinate system; Send the optimized inspection path and task allocation instructions to the corresponding inspection terminal, and receive the inspection result data fed back by the terminal; The facility characteristic data and the dynamic weight of the space-time grid coordinate system are updated according to the inspection result data.

2. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The preset attribute dimensions include facility type, geographic coordinates, maintenance cycle and historical fault records.

3. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The dynamically adjusting the weight of each grid unit in the initial inspection path includes: Obtain real-time traffic congestion index and weather data, and calculate the travel time cost of each grid unit; Predict the failure probability of each facility based on the facility's historical failure records and machine learning models; The weight of the grid unit is dynamically updated according to the product of the travel time cost and the failure probability.

4. The municipal facility inspection and dispatching method based on the city steward according to claim 3 is characterized in that: The machine learning model is a time series-based LSTM network. The input data includes facility operating parameters, environmental data, and historical maintenance records. The output is the failure probability within a preset time period in the future.

5. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The allocating the inspection subtask includes: Construct a priority scoring matrix based on the skill level, current location and urgency of the inspection personnel or equipment; The Hungarian algorithm is used to optimally match the scoring matrix to determine the allocation results of the inspection subtasks.

6. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The method further comprises: Collaborative planning is performed for multiple inspection tasks in the same geographical area. The path intersections between multiple tasks are optimized by introducing the ant colony algorithm to reduce repeated inspection areas.

7. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The method further comprises: A multi-objective optimization model is constructed with minimizing the total inspection time, balancing the load distribution and maximizing the fault coverage as the optimization goals. The NSGA-II algorithm is used to solve the Pareto optimal solution set, and the final inspection plan is selected based on the preset rules.

8. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The method further comprises: When a sudden facility failure is detected, the current inspection route is interrupted, an emergency inspection task is inserted based on the time-space coordinates and priority of the faulty facility, and the optimized route is recalculated.

9. The municipal facility inspection and dispatching method based on the city steward according to claim 1 is characterized in that: The method further comprises: An inspection effect feedback strategy is established to dynamically correct the parameters of the facility failure probability prediction model based on the correlation between historical inspection data and facility status changes.

10. A municipal facility inspection and dispatching system based on the city manager, used to implement any of the methods described in claims 1 to 9, characterized in that: include: Data collection and processing module: used to obtain basic information and real-time status data of municipal facilities, structure the basic information according to preset attribute dimensions, and generate facility feature data; Space-time grid construction module: construct a space-time grid coordinate system, divide the grid units based on the urban geographical area, define the X-axis of the space-time grid coordinate system as the time axis, the Y-axis as the space axis, and the Z-axis as the facility type axis, and assign dynamic weights to each grid unit; Mapping module: maps the facility feature data to the corresponding space-time grid coordinate system to generate facility space-time coordinate information; Initial path generation module: Based on the preset inspection task requirements, the type, geographical location and priority of the inspection target facilities are extracted to generate the initial inspection path; Path optimization module: dynamically adjusts the weight of each grid unit in the initial inspection path according to real-time traffic flow data and facility failure probability prediction model, and generates an optimized inspection path based on a greedy algorithm; Task allocation module: splits the optimized inspection path into multiple inspection subtasks, and allocates the inspection subtasks according to the real-time location and load status of the inspection personnel or equipment; Monitoring and re-optimization module: collects the execution progress data of the inspection subtasks in real time, and re-optimizes the paths of the unfinished inspection subtasks in combination with the dynamic weights of the spatiotemporal grid coordinate system; Instruction sending and feedback receiving module: sends the optimized inspection path and task allocation instructions to the corresponding inspection terminal, and receives the inspection result data fed back by the terminal; Data update module: updates the facility feature data and the dynamic weight of the space-time grid coordinate system according to the inspection result data.

Images