Sanitation vehicle intelligent path planning method and system, scheduling method and storage medium

Abstract

The invention relates to the technical field of vehicle path planning, in particular to an intelligent path planning method and system for a sanitation vehicle, a scheduling method and a storage medium, and the path planning method comprises a data collection step, a data processing step and a path planning step. Calculating the shortest path from the current position to each garbage collection point and the shortest paths from each garbage collection point to other garbage collection points according to a shortest path algorithm; performing path planning according to a reinforcement learning algorithm to obtain a plurality of planned paths from the current position to the target garbage collection point; the influence coefficient of environmental data conversion and the vehicle information are input into a driving cost reward function, reward values of a plurality of planned paths are output, and the planned path corresponding to the minimum reward value is the optimal path of the environmental sanitation vehicle. The optimal path can be dynamically adjusted according to real-time data, and the cleaning efficiency and the resource utilization rate are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle path planning, and in particular to an intelligent path planning method, system, scheduling method and storage medium for sanitation vehicles. Background Art

[0002] Existing sanitation vehicle path planning methods mainly rely on preset fixed routes, which usually do not take into account real-time traffic conditions, special events (such as road construction, temporary road closures) or environmental changes (such as weather changes). This leads to the following problems:

[0003] 1) Inefficient cleaning: Fixed routes may not be optimal, resulting in longer cleaning times and increased energy consumption. For example, during peak traffic hours, fixed routes may cause sanitation vehicles to get stuck in traffic jams, delaying cleaning operations and increasing unnecessary waiting time. In addition, if the fixed route is longer, the distance traveled by the vehicle will also increase, leading to higher energy consumption.

[0004] 2) Uneven resource allocation: Fixed operating times and routes may result in over-cleaning of some areas and under-cleaning of others. For example, in densely populated areas, fixed routes may schedule too many cleanings, while in sparsely populated areas, the number of cleanings is insufficient. This not only wastes cleaning resources, but may also result in environmental sanitation problems in some areas not being resolved in a timely manner.

[0005] 3) Poor ability to respond to emergencies: When encountering emergencies, fixed routes cannot be adjusted quickly, affecting the continuity and efficiency of sanitation operations. For example, when a traffic accident occurs on a road or a road is closed due to road construction, the fixed route cannot be adjusted immediately, and sanitation vehicles may be forced to stop or take a long detour, thus affecting the overall progress of the operation.

[0006] 4) Insufficient data utilization: Existing methods fail to fully utilize urban big data, such as traffic flow, weather conditions, etc., to optimize cleaning routes. For example, although urban traffic management systems can provide real-time traffic flow information, this data has not been effectively integrated into route planning, resulting in the planned routes being unable to adapt to real-time changing traffic conditions.

[0007] Therefore, it is necessary to develop a new intelligent path planning method, system, scheduling method and storage medium for sanitation vehicles. Summary of the invention

[0008] The object of the present invention is to provide an intelligent path planning method, system, scheduling method and storage medium for sanitation vehicles, which can dynamically adjust the optimal path according to real-time data to improve cleaning efficiency and resource utilization.

[0009] In a first aspect, the present invention provides a method for intelligent path planning of a sanitation vehicle, comprising the following steps:

[0010] Data collection step: collecting vehicle information and environmental data in real time, wherein the environmental data includes traffic conditions and weather conditions, and the vehicle information includes vehicle location and status;

[0011] Data processing steps: clean and format the collected vehicle information and environmental data, and convert the environmental data into impact coefficients;

[0012] Path planning step: Calculate the shortest path from the current location to each garbage collection point, and the shortest path from each garbage collection point to other garbage collection points according to the shortest path algorithm;

[0013] Perform path planning based on the reinforcement learning algorithm to obtain several planned paths from the current location to the target garbage collection point;

[0014] The influence coefficient of environmental data conversion and vehicle information are input into the driving cost reward function, and the reward values ​​of several planned paths are output. The planned path corresponding to the minimum reward value is the optimal path for the sanitation vehicle.

[0015] Furthermore, the driving cost reward function is specifically:

[0016] R=d×R base ;

[0017] R base =ω1×D+ω2×E+ω3×T+ω2×W;

[0018] In the formula, R is the reward value of the planned path, d is the traffic condition influence coefficient, and R base is the basic reward value of the planned path, D is the length of the planned path, E is the theoretical energy consumption of the planned path for sanitation vehicles, T is the theoretical time of the planned path for sanitation vehicles, W is the influence coefficient of weather conditions, ω1, ω2, ω3, ω4 are weight coefficients and ω1+ω2+ω3+ω4=1.

[0019] Furthermore, after obtaining the reward values ​​of several planned paths, the optimal path of the sanitation vehicle is obtained based on the formula min{R(v), R(u)+R(u, v)}, where R(v) is the reward value of the sanitation vehicle from the current position to the target position, R(u) is the reward value of the sanitation vehicle from the current position to the intermediate node position, and R(u, v) is the reward value of the sanitation vehicle from the intermediate node position to the target position.

[0020] Further, the shortest path algorithm is Dijkstra algorithm or Bellman-Ford algorithm;

[0021] And / or, the reinforcement learning algorithm is a TD-learning algorithm.

[0022] Furthermore, after the optimal path of the sanitation vehicle is obtained, the sanitation vehicle is controlled to travel along the optimal path, vehicle information and environmental data are collected in real time or periodically, and the optimal path is dynamically adjusted based on the collected vehicle information and environmental data.

[0023] Furthermore, the driving cost reward function adjusts the weight coefficients ω1, ω2, ω3, and ω4 in real time or regularly according to the actual driving information data and weather forecast data fed back by the sanitation vehicles.

[0024] In a second aspect, the present invention provides an intelligent path planning system for sanitation vehicles, comprising a memory and a controller, wherein the memory stores a computer-readable program, and when the computer-readable program is called by the controller, it can execute the steps of the above-mentioned intelligent path planning method for sanitation vehicles.

[0025] In a third aspect, the present invention provides an intelligent scheduling method for sanitation vehicles, which uses the above-mentioned intelligent path planning method for sanitation vehicles to obtain the optimal path between several garbage collection points, and uses a greedy algorithm to generate an optimal scheduling task list;

[0026] Control the sanitation vehicles to drive along the generated optimal scheduling task list, collect vehicle information and environmental data in real time or periodically, and feed the vehicle information and environmental data back to the data collection step to form a closed-loop control.

[0027] Furthermore, a greedy algorithm is used to generate the optimal scheduling solution, specifically:

[0028] According to the optimal paths between several garbage collection points, an initial scheduling task list is generated, and several tasks in the initial scheduling task list are sorted in order of priority from high to low;

[0029] Traverse and check each task in the sorted initial scheduling task list;

[0030] When checking the i-th task, if t cur +t tra +t com ≤t end , and E cur -E the ≥0, then add the i-th task to the optimal scheduling plan; otherwise, skip the i-th task and continue to check the i+1-th task; where t cur is the current time before executing the i-th task, t tra is the estimated travel time of the optimal route planned for the sanitation vehicle when performing the i-th task, t com is the estimated operation time of the ith task, tend is the preset work deadline, E cur is the energy storage of the sanitation vehicle before the i-th mission, E the The theoretical energy consumption of the optimal path planned for the sanitation vehicle when performing the i-th task;

[0031] After the traversal check is completed, the optimal scheduling plan can be obtained.

[0032] In a fourth aspect, the present invention provides a storage medium storing a computer-readable program, which can execute the steps of the above-mentioned intelligent path planning method for sanitation vehicles when called by a controller.

[0033] The present invention has the following unexpected beneficial effects:

[0034] 1. The present invention collects vehicle information and environmental data in real time, cleans and formats the collected vehicle information and environmental data, and converts the environmental data into an influence coefficient. Using the shortest path algorithm in graph theory as the basis, combined with reinforcement learning technology, dynamic planning and adjustment are used to obtain the optimal path of sanitation vehicles, so that sanitation vehicles can complete cleaning tasks in the shortest time, reduce invalid driving, thereby reducing energy consumption, efficiently implementing cleaning operations, significantly improving cleaning efficiency and optimizing resource allocation. In practical applications, this means that sanitation vehicles can avoid additional energy consumption caused by detours or waiting, and can also adjust the path in time in situations such as traffic congestion or weather changes to maintain the continuity and efficiency of operations. In addition, through real-time data feedback and model parameter update mechanisms, the system can continuously optimize the path planning model to make path planning more accurate and more adaptable. For example, in the event of an emergency (such as road construction, temporary road closures), the system can quickly adjust the cleaning route to ensure that sanitation operations are not affected.

[0035] 2. The present invention can dynamically adjust the cleaning route to ensure that resources are reasonably allocated and avoid over-cleaning of some areas and under-cleaning of other areas. By comprehensively considering the working time and energy constraints of the vehicle and using a greedy algorithm for task scheduling, the system can ensure that each vehicle completes as many tasks as possible within its working time, thereby maximizing resource utilization. The intelligent path planning method of the present invention achieves high efficiency, flexibility and energy saving in sanitation operations through advanced algorithms and technical means, greatly improving the technical level of sanitation operations.

[0036] 3. The intelligent path planning method of the present invention not only improves the efficiency and adaptability of sanitation operations at the technical level, but also brings significant economic and social benefits to enterprises at the commercial level, and promotes the intelligent development of the sanitation industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flow chart of the intelligent path planning method for sanitation vehicles described in an embodiment of the present application;

[0038] Figure 2 It is a principle block diagram of the intelligent path planning system for sanitation vehicles described in the embodiments of the present application;

[0039] Figure 3 It is a functional module diagram of a computer-readable program in an embodiment of the present application. DETAILED DESCRIPTION

[0040] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0041] In one embodiment, see Figure 1 As shown, the present invention provides a method for intelligent path planning of sanitation vehicles, comprising the following steps:

[0042] S1, data collection step: collect vehicle information and environmental data in real time, the environmental data includes traffic conditions and weather conditions, and the vehicle information includes vehicle location and status.

[0043] Data collection is the basis of the entire path planning, where vehicle information includes the vehicle's current location (obtained through GPS and other positioning technologies), speed (obtained through speed sensors placed on sanitation vehicles), remaining power or fuel (obtained through power sensors or fuel sensors placed on sanitation vehicles) and other status information. Environmental data covers traffic conditions (such as congestion, road closure information, etc.) and weather conditions (such as temperature, humidity, wind speed, rain, snow, haze, etc.), which have a direct impact on vehicle speed and safety. For example, if it is raining in a certain area, the roads in that area may be more congested, which will affect the vehicle's driving speed and energy consumption. By collecting vehicle information and environmental data in real time, the real-time status of sanitation vehicles and their surroundings can be fully grasped, providing reliable data support for subsequent path planning.

[0044] Exemplarily, the current position coordinates, speed and remaining power data of three sanitation vehicles (V1, V2, V3) are obtained as shown in Table 1.

[0045] Table 1 Collected vehicle information

[0046] Vehicle number Position coordinates (x,y) Speed ​​(m / s) Remaining power (kWh) <![CDATA[V 1]]> (34.0522,－118.2437) 15 50 <![CDATA[V 2]]> (34.0515,－118.2432) 10 45 <![CDATA[V 3]]> (34.0528,－118.2445) 12 48

[0047] At the same time, the current traffic status data is obtained through the urban traffic management system, as shown in Table 2.

[0048] Table 2 Obtained traffic conditions

[0049] Road Number Traffic volume (vehicles / hour) Traffic conditions (congested / uncongested) 1 200 Smooth 2 150 Smooth 3 300 Congestion

[0050] In addition, the real-time weather condition data obtained through the weather station is shown in Table 3.

[0051] Table 3 Obtained weather conditions

[0052]

[0053]

[0054] S2, data processing step: cleaning and formatting the collected vehicle information and environmental data, and converting the environmental data into impact coefficients.

[0055] Since real-time data may contain noise or outliers, it needs to be cleaned to ensure the accuracy and reliability of the data. At the same time, the data is formatted into a form that the algorithm can recognize for subsequent processing. Specifically, first, the sensor data is cleaned to exclude invalid or erroneous data. For example, if the GPS signal is unstable and causes outliers in the vehicle position, the system will automatically correct these outliers through the data verification algorithm. Secondly, a smoothing algorithm (such as the moving average method) is used to remove noise from the sensor data to ensure the accuracy of the data.

[0056] Environmental data conversion into impact coefficients: complex traffic and weather condition data are converted into specific values ​​or coefficients, which will be directly used in subsequent path evaluation and cost calculation. For example, congested road sections can be assigned higher driving cost coefficients, while severe weather conditions may add additional safety cost coefficients.

[0057] For example, suppose there are some abnormal values ​​in the sensor data, for example, the position coordinates of vehicle V1 at a certain moment are (34.0522, -118.2437), but suddenly change to (34.0522, -118.2437) at the next moment, which is obviously inconsistent with the actual situation. The system will automatically correct these abnormal values. In addition, the system will also use the moving average method to remove noise in the sensor data to ensure the accuracy of the data.

[0058] Assume that a road is currently in a congested state with a traffic flow of 300 vehicles per hour. According to empirical data, the driving speed will be reduced by about 30% in a congested state, so the weight of road R3 will be increased to reflect its higher driving cost.

[0059] The impact coefficient of a sunny day is 1.0, the impact coefficient of a rainy day is 1.2 (indicating a 20% increase in driving difficulty), and the impact coefficient of a snowy day is 1.5. These data processing steps ensure the quality of the data and provide a reliable basis for subsequent path planning.

[0060] S3, path planning step.

[0061] S31, calculating the shortest path from the current position to each garbage collection point, and the shortest path from each garbage collection point to other garbage collection points according to the shortest path algorithm.

[0062] S32, perform path planning according to the reinforcement learning algorithm to obtain several planned paths from the current location to the target garbage collection point; based on the basic path set, use the reinforcement learning algorithm to optimize the path. Reinforcement learning learns how to select the optimal driving path under different environmental conditions through continuous trial and error and reward mechanism.

[0063] S33, input the impact coefficient of environmental data conversion and vehicle information into the driving cost reward function, and output the reward values ​​of several planned paths. The planned path corresponding to the minimum reward value is the optimal path for the sanitation vehicle.

[0064] The impact coefficient of environmental data conversion and vehicle information are used as input to construct a driving cost reward function. This function comprehensively considers multiple factors such as driving distance, time, energy consumption, weather impact, etc., and calculates a comprehensive reward value for each planned path.

[0065] According to the reward value calculated by the reward function, the path with the smallest reward value is selected as the optimal driving path for the sanitation vehicle. After comprehensively considering various factors, this path is considered to be the most economical, efficient and safe driving solution under current conditions.

[0066] The intelligent path planning method for sanitation vehicles of the present invention realizes intelligent optimization of the driving path of sanitation vehicles through real-time data collection, intelligent data processing, path planning based on the shortest path algorithm and reinforcement learning algorithm, and a driving cost reward function that comprehensively considers multiple factors. This method not only improves the efficiency of sanitation operations, but also reduces operating costs, providing strong support for urban environmental sanitation management. In practical applications, this means that sanitation vehicles can avoid additional energy consumption caused by detours or waiting, and can also adjust the path in time in situations such as traffic congestion or weather changes to maintain the continuity and efficiency of operations. In addition, through real-time data feedback and model parameter update mechanisms, the system can continuously optimize the path planning model, making path planning more accurate and more adaptable. For example, when encountering emergencies (such as road construction, temporary road closures), the system can quickly adjust the cleaning route to ensure that sanitation operations are not affected.

[0067] In a preferred embodiment, the driving cost reward function is specifically:

[0068] R=d×R base ; R base =ω1×D+ω2×E+ω3×T+ω2×W.

[0069] Where R is the reward value of the planned path. d is the traffic condition influence coefficient. base is the basic reward value for planning the path.

[0070] D is the length of the planned path, which can be in meters, kilometers, etc.

[0071] E is the theoretical energy consumption of the planned route of the sanitation vehicle, and the unit can be joule, kilowatt-hour, liter per 100 kilometers, etc.

[0072] T is the theoretical time for the sanitation vehicle to plan the route, and the unit can be seconds, minutes, hours, etc.

[0073] W is the weather condition influence coefficient.

[0074] ω1, ω2, ω3, and ω4 are weight coefficients, ω1+ω2+ω3+ω4=1, and ω1, ω2, ω3, and ω4 are all ≥ 0, to ensure that each cost factor accounts for a reasonable proportion of the total cost. The selection of weight coefficients needs to be weighed and adjusted according to actual needs.

[0075] For example, suppose a city has three garbage collection points A, B, and C, and a sanitation vehicle is located at the starting point O. Through GPS and traffic management systems, it is known that the distance from O to A is 10km, and the theoretical energy consumption is 2kWh; the distance from O to B is 15km, and the theoretical energy consumption is 3kWh; the distance from O to C is 12km, and the theoretical energy consumption is 2.5kWh. Assuming that the weather is good on that day, the weight coefficients ω1, ω2, ω3, and ω4 are 0.4, 0.3, 0.2, and 0.1 respectively. The traffic conditions from sanitation vehicles to the three garbage collection points A, B, and C are all smooth, and the traffic condition influence coefficient d=1. The weather conditions are all sunny, and the weather influence coefficient W=1. In this example, the influence of driving time is ignored, so the driving time of the three routes is set to 0.

[0076] Then it is calculated based on the driving cost reward function:

[0077] R(O→A)=d×R base =0.4×10+0.3×2.0+0.2×0+0.1×1=4.7;

[0078] R(O→B)=d×R base=0.4×15+0.3×3.0+0.2×0+0.1×1=7.0;

[0079] R(O→C)=d×R base =0.4×12+0.3×2.5+0.2×0+0.1×1=6.05.

[0080] By comparison, the reward value from the starting point O to the garbage collection point A is the smallest, indicating that the driving cost of this route is the lowest.

[0081] In a preferred embodiment, after obtaining the reward values ​​of several planned paths, the optimal path of the sanitation vehicle is obtained based on the formula min{R(v), R(u)+R(u, v)}, where R(v) is the reward value of the sanitation vehicle from the current position to the target position, R(u) is the reward value of the sanitation vehicle from the current position to the intermediate node position, and R(u, v) is the reward value of the sanitation vehicle from the intermediate node position to the target position.

[0082] By calculating the reward value of the sanitation vehicle going directly from the current position to the target position, and the reward value of the sanitation vehicle going from the current position to the target position through the intermediate node, and comparing the two, the path with the smaller reward value is selected as the optimal path to ensure that the sanitation vehicle's driving cost is lower.

[0083] In a preferred embodiment, the shortest path algorithm is the Dijkstra algorithm or the Bellman-Ford algorithm; and the reinforcement learning algorithm is the TD-learning algorithm.

[0084] The present invention makes special restrictions on the shortest path algorithm and the reinforcement learning algorithm. The Dijkstra algorithm is applicable to graphs with non-negative weights. It can find the shortest path from a single source point to all other vertices. The Dijkstra algorithm maintains a distance array (or priority queue), continuously selects the node closest to the source point among the currently unprocessed nodes, and updates the distance of its adjacent nodes until all nodes have been processed.

[0085] The Bellman-Ford algorithm works on graphs containing negative weight edges, but cannot handle negative weight cycles. It is able to find the shortest path from a single source vertex to all other vertices, even if there are negative weight edges in the graph. The Bellman-Ford algorithm progressively relaxes (i.e. tries to shorten) the shortest path estimates for all vertices by traversing all edges multiple times until no more relaxation operations can be performed.

[0086] If the road weights (such as travel time or distance) are all non-negative, then Dijkstra's algorithm may be a more efficient choice. However, if some special cases (such as negative weight edges caused by temporary traffic control) are taken into account, then Bellman-Ford's algorithm may be more applicable.

[0087] TD-learning (Temporal Difference Learning) is a method in reinforcement learning that combines the Monte Carlo method with the idea of ​​dynamic programming. The TD-learning algorithm does not need to wait until the entire sequence (or "round") is over before updating, but can be updated immediately after each step based on the current estimated future reward value. The key concept in the TD-learning algorithm is "temporal difference", which is the difference between the current estimated future reward value and the actual observed reward value.

[0088] In the application of reinforcement learning in sanitation vehicle path planning, the TD-learning algorithm can help vehicles dynamically adjust their driving paths to minimize driving costs based on real-time traffic conditions, weather changes and other factors. Through continuous learning and optimization, vehicles can gradually find more efficient and economical driving routes.

[0089] In a preferred embodiment, after the optimal path of the sanitation vehicle is obtained, the sanitation vehicle is controlled to travel along the optimal path, vehicle information and environmental data are collected in real time or periodically, and the optimal path is dynamically adjusted based on the collected vehicle information and environmental data.

[0090] The detailed analysis and explanation are given below.

[0091] 1. Control sanitation vehicles to travel along the optimal path

[0092] Instruction sending: The calculated optimal path is converted into instructions that the sanitation vehicle can understand (such as direction, speed, etc.), and sent to the sanitation vehicle through the on-board system.

[0093] Navigation and monitoring: Use the on-board GPS and navigation system to guide sanitation vehicles along the optimal route, while monitoring the driving status and location of sanitation vehicles in real time.

[0094] Communication guarantee: Ensure unimpeded communication between sanitation vehicles and the control center so that new instructions or updated information can be received quickly when needed.

[0095] 2. Real-time or periodic collection of vehicle information and environmental data

[0096] Vehicle information: including the current location, speed, direction, remaining battery / fuel, fault status, etc. of the sanitation vehicle. This information can be obtained in real time through on-board sensors and control systems.

[0097] Environmental data: including traffic conditions (congestion, accidents, road construction, etc.), weather conditions (rain, snow, haze, temperature, humidity, wind speed, etc.), and possible emergencies (such as road closures, temporary traffic control, etc.). These data can be obtained through external equipment such as on-board cameras, radars, and weather stations, or through data sharing with traffic management systems, meteorological departments, etc.

[0098] 3. Dynamically adjust the optimal route based on the collected vehicle information and environmental data

[0099] Data analysis: Real-time analysis of collected vehicle information and environmental data to assess their impact on the current optimal path. For example, if there is severe congestion on the road ahead, the route may need to be replanned to avoid delays.

[0100] Path replanning: If the analysis results show that the current optimal path is no longer applicable, the path replanning process is triggered. This can be based on the same method as the initial path planning (such as combining the shortest path algorithm and reinforcement learning algorithm), but this time taking into account new vehicle information and environmental data.

[0101] Update instructions: The re-planned optimal path is converted into a new driving instruction and sent to the sanitation vehicle through the vehicle system. At the same time, the monitoring interface of the control center is updated so that the operator can understand the latest driving route of the vehicle in real time.

[0102] 4. Continuous optimization and feedback

[0103] Performance evaluation: Regularly evaluate the driving efficiency and cost of sanitation vehicles, as well as the effect of dynamic route adjustment. This can be achieved by analyzing vehicle driving data, energy consumption data, task completion, etc.

[0104] Algorithm tuning: Based on the performance evaluation results, the shortest path algorithm, reinforcement learning algorithm, and dynamic path adjustment strategy are continuously optimized and improved. For example, the algorithm parameters are adjusted, new optimization strategies are introduced, or other advanced technologies are integrated. For example, the weight coefficients ω1, ω2, ω3, and ω4 are adjusted in real time or regularly based on the actual driving information data and weather forecast data fed back by sanitation vehicles.

[0105] User feedback: Collect feedback from operators and sanitation workers to understand their needs and problems during use, so as to further optimize system functions and user experience.

[0106] Through the implementation of the above process, it can be ensured that sanitation vehicles can complete sanitation tasks efficiently and safely in complex and changing environments, while reducing operating costs and improving service quality.

[0107] In one embodiment, participating Figure 2 As shown, the present invention provides an intelligent path planning system for sanitation vehicles, including a memory and a controller, wherein the memory stores a computer-readable program, and when the computer-readable program is called by the controller, it can execute the steps of the above-mentioned intelligent path planning method for sanitation vehicles.

[0108] In one embodiment, the present invention provides an intelligent scheduling method for sanitation vehicles, which uses the above-mentioned intelligent path planning method for sanitation vehicles to obtain the optimal path between several garbage collection points, and uses a greedy algorithm to generate an optimal scheduling task list. The basic idea of ​​the greedy algorithm is to take the best or optimal choice (i.e., the most favorable choice) in the current state in each step of selection, so as to hope that the result is the best or optimal algorithm in the world.

[0109] Control the sanitation vehicles to drive along the generated optimal scheduling task list, collect vehicle information and environmental data in real time or periodically, and feed the vehicle information and environmental data back to the data collection step to form a closed-loop control.

[0110] In a preferred implementation, a greedy algorithm is used to generate an optimal scheduling solution, specifically:

[0111] According to the optimal path between several garbage collection points, an initial scheduling task list is generated, which contains all the garbage collection tasks that need to be executed. Several tasks in the initial scheduling task list are sorted in order of priority from high to low. The priority can be determined based on a variety of factors, such as the amount of garbage at the garbage collection point, urgency, geographical location (whether it is close to other tasks to reduce travel time), etc.

[0112] Traverse and check each task in the sorted initial scheduling task list, starting from the first task in the sorted task list, and check whether each task can be added to the optimal scheduling plan one by one.

[0113] When checking the i-th task, if t cur +t tra +t com ≤t end , that is, the current time t before executing the i-th task cur Add the estimated travel time t tra and the estimated operation time t com If this sum is less than or equal to the preset work deadline t end , indicating that it is feasible in terms of time. And E cur-E the ≥0, that is, the energy storage E of the sanitation vehicle before the i-th task cur Subtract the expected theoretical energy consumption E when performing the task the If the difference is greater than or equal to 0, it means that it is feasible in terms of energy consumption. Then add the i-th task to the optimal scheduling plan; otherwise, skip the i-th task and continue to check the i+1-th task.

[0114] Where, t cur is the current time before executing the i-th task, t tra is the estimated travel time of the optimal route planned for the sanitation vehicle when performing the i-th task, t com is the estimated operation time of the ith task, t end is the preset work deadline, E cur is the energy storage of the sanitation vehicle before the i-th mission, E the The theoretical energy consumption of the optimal path planned for the sanitation vehicle when performing the i-th task.

[0115] After the traversal check is completed, an optimal scheduling plan that meets the time and energy consumption constraints can be obtained.

[0116] Assume that the maximum working time of the vehicle is 8 hours, the maximum energy consumption is 60kWh, and the current remaining energy storage is 50kWh. The initial scheduling task list includes three tasks, which are sorted in descending order of priority, namely task1, task2, and task3.

[0117] According to the greedy algorithm: task 1 is the first task to be considered. After evaluation, the vehicle can complete task 1, and the remaining energy storage of the sanitation vehicle is updated to 35kWh. Next, task 2 is considered. At this time, the vehicle can still complete task 2, and the remaining energy storage of the sanitation vehicle is updated to 15kWh. Finally, task 3 is considered, but due to E cur -E the <0, so the vehicle cannot complete task3. Therefore, the final task scheduling result is to execute task1 and task2.

[0118] In the actual scheduling process, it is also necessary to consider the dynamically changing traffic conditions and weather conditions to ensure the rationality and feasibility of task allocation. To this end, a dynamic adjustment mechanism can be added to the greedy algorithm. If new traffic conditions are updated during the task scheduling process, such as road closures or traffic congestion information, the estimated travel time t of the optimal path planned for the sanitation vehicle route when executing the i-th task of the affected task is recalculated. tra The theoretical energy consumption E of the optimal path planned by the sanitation vehicle when performing the i-th task the, and adjust the order of task execution. If weather conditions change (such as rain or snow), the energy consumption of the task is adjusted according to the weather impact coefficient, and whether the current task can be completed is re-evaluated.

[0119] In one embodiment, the present invention provides a storage medium storing a computer-readable program, which can execute the steps of the above-mentioned intelligent path planning method for sanitation vehicles when called by a controller.

[0120] The computer-readable program is divided into functional modules, including a data acquisition module, a data processing module, an optimization scheduling module and an execution monitoring module. Among them:

[0121] Data acquisition module: used to collect vehicle information and environmental data in real time, the environmental data includes traffic conditions and weather conditions, and the vehicle information includes vehicle location and status.

[0122] Data processing module: used to clean and format the collected vehicle information and environmental data, and convert the environmental data into impact coefficients.

[0123] Path planning module: used to calculate the shortest path from the current position to each garbage collection point and the shortest path from each garbage collection point to other garbage collection points according to the shortest path algorithm; perform path planning according to the reinforcement learning algorithm to obtain several planned paths from the current position to the target garbage collection point; input the influence coefficient of environmental data conversion and vehicle information into the driving cost reward function, and output the reward values ​​of several planned paths. The planned path corresponding to the minimum reward value is the optimal path for the sanitation vehicle.

[0124] Optimization scheduling module: Use quantum-inspired evolutionary algorithms to process the data generated by the data processing module to find the optimal scheduling plan, including the optimal path and cleaning order of the vehicle.

[0125] Execution monitoring module: Executes cleaning tasks according to the optimal scheduling plan generated by the optimization scheduling module, monitors and collects vehicle information and environmental data in real time, and feeds the information back to the data acquisition module to form a closed-loop control.

[0126] Through the collaborative work of the above five modules, the system can achieve effective management and scheduling of sanitation vehicles. The data acquisition module ensures that the system obtains accurate real-time data; the data processing module ensures the availability and consistency of data; the path planning module realizes the optimal path planning acquisition of sanitation vehicles; the task scheduling module uses the greedy algorithm to generate the optimal scheduling plan; the execution monitoring module is responsible for supervising the execution of the plan and collecting feedback to ensure the efficient operation of the entire system. This closed-loop control mechanism not only improves the efficiency of cleaning work, but also achieves cost savings and resource optimization.

[0127] In an embodiment of the present application, the storage medium may be a tangible storage medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. The storage medium may be a machine-readable signal medium or a machine-readable storage medium. The storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0128] The above-mentioned embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above-mentioned embodiments. Any other changes, modifications, substitutions, combinations and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement modes and shall be included in the protection scope of the present invention.

Claims

1. A method for intelligent path planning of sanitation vehicles, characterized in that: The following steps are involved: Data collection step: collecting vehicle information and environmental data in real time, wherein the environmental data includes traffic conditions and weather conditions, and the vehicle information includes vehicle location and status; Data processing steps: clean and format the collected vehicle information and environmental data, and convert the environmental data into impact coefficients; Path planning step: Calculate the shortest path from the current location to each garbage collection point, and the shortest path from each garbage collection point to other garbage collection points according to the shortest path algorithm; Perform path planning based on the reinforcement learning algorithm to obtain several planned paths from the current location to the target garbage collection point; The influence coefficient of environmental data conversion and vehicle information are input into the driving cost reward function, and the reward values ​​of several planned paths are output. The planned path corresponding to the minimum reward value is the optimal path for the sanitation vehicle.

2. The intelligent path planning method for sanitation vehicles according to claim 1, characterized in that: The driving cost reward function is specifically: R=d×R base ; R base =ω1×D+ω2×E+ω3×T+ω2×W; In the formula, R is the reward value of the planned path, d is the traffic condition influence coefficient, and R base is the basic reward value of the planned path, D is the length of the planned path, E is the theoretical energy consumption of the planned path for sanitation vehicles, T is the theoretical time of the planned path for sanitation vehicles, W is the influence coefficient of weather conditions, ω1, ω2, ω3, ω4 are weight coefficients and ω1+ω2+ω3+ω4=1.

3. The intelligent path planning method for sanitation vehicles according to claim 2 is characterized in that: After obtaining the reward values ​​of several planned paths, the optimal path of the sanitation vehicle is obtained based on the formula min{R(v), R(u)+R(u, v)}, where R(v) is the reward value of the sanitation vehicle from the current position to the target position, R(u) is the reward value of the sanitation vehicle from the current position to the intermediate node position, and R(u, v) is the reward value of the sanitation vehicle from the intermediate node position to the target position.

4. The intelligent path planning method for sanitation vehicles according to claim 1 or 2, characterized in that: The shortest path algorithm is Dijkstra algorithm or Bellman-Ford algorithm; And / or, the reinforcement learning algorithm is a TD-learning algorithm.

5. The intelligent path planning method for sanitation vehicles according to claim 1 or 2, characterized in that: After obtaining the optimal path for the sanitation vehicle, the sanitation vehicle is controlled to travel along the optimal path, vehicle information and environmental data are collected in real time or periodically, and the optimal path is dynamically adjusted based on the collected vehicle information and environmental data.

6. The intelligent path planning method for sanitation vehicles according to claim 1 or 2, characterized in that: The driving cost reward function adjusts the weight coefficients ω1, ω2, ω3, and ω4 in real time or regularly according to the actual driving information data and weather forecast data fed back by the sanitation vehicles.

7. An intelligent path planning system for sanitation vehicles, characterized in that: It includes a memory and a controller, wherein the memory stores a computer-readable program, and when the computer-readable program is called by the controller, it can execute the steps of the intelligent path planning method for sanitation vehicles as described in any one of claims 1 to 6.

8. A method for intelligent dispatching of sanitation vehicles, characterized in that: The sanitation vehicle intelligent path planning method as described in any one of claims 1 to 6 is used to obtain the optimal path between several garbage collection points, and the greedy algorithm is used to generate the optimal scheduling task list; Control the sanitation vehicles to drive along the generated optimal scheduling task list, collect vehicle information and environmental data in real time or periodically, and feed the vehicle information and environmental data back to the data collection step to form a closed-loop control.

9. The intelligent dispatching system for sanitation vehicles according to claim 8, characterized in that: The greedy algorithm is used to generate the optimal scheduling solution, specifically: According to the optimal paths between several garbage collection points, an initial scheduling task list is generated, and several tasks in the initial scheduling task list are sorted in order of priority from high to low; Traverse and check each task in the sorted initial scheduling task list; When checking the i-th task, if t cur +t tra +t com ≤t end , and E cur -E the ≥0, then add the i-th task to the optimal scheduling plan; otherwise, skip the i-th task and continue to check the i+1-th task; where t cur is the current time before executing the i-th task, t tra is the estimated travel time of the optimal route planned for the sanitation vehicle when performing the i-th task, t com is the estimated operation time of the ith task, t end is the preset work deadline, E cur is the energy storage of the sanitation vehicle before the i-th mission, E the The theoretical energy consumption of the optimal path planned for the sanitation vehicle when performing the i-th task; After the traversal check is completed, the optimal scheduling plan can be obtained.

10. A storage medium, characterized in that: A computer-readable program is stored therein, and when the computer-readable program is called by the controller, the steps of the intelligent path planning method for a sanitation vehicle as described in any one of claims 1 to 6 can be executed.

Images