Vehicle path planning method based on adaptive optimization algorithm

Abstract

The invention provides a vehicle path planning method based on an adaptive optimization algorithm, and relates to the technical field of adaptive algorithm application, and the method comprises the steps: 1, based on a current order demand, a vehicle resource and a preset constraint condition, obtaining a vehicle path planning result; an initial path scheme set meeting the requirements of a time window, the number of splitting times, vehicle type matching and a temperature zone is generated through a step-by-step progressive decision-making process; 2, starting a plurality of independent optimization links according to the initial path scheme set, and dynamically adjusting local path nodes by each link through a greedy strategy to generate an optimization candidate scheme set; and step 3, performing real-time evaluation on the optimized candidate scheme set, determining a final candidate scheme based on the cost saving rate, the demand satisfaction rate and the soft service quality index, and updating the final candidate scheme as the current reference scheme. According to the method, dynamic events are sensed in real time through an adaptive optimization algorithm, influences are evaluated, the path is intelligently adjusted, and the timeliness and accuracy of path planning are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of adaptive algorithm applications, and particularly to a vehicle path planning method based on an adaptive optimization algorithm. Background Art

[0002] In the shuttle bus pick-up and drop-off during fixed periods at school, although traditional vehicle path planning techniques can achieve basic route design, there is still room for optimization when dealing with dynamic variables. Traditional techniques rely on fixed reservation information for route planning and may have difficulty quickly adapting to real-time changes in riding demands. For example, on a Friday morning rush hour, the preset riding list included 120 junior high school students. However, 15 minutes before the actual departure, due to some parents choosing to drive their children to school temporarily, 23 students cancelled their rides. At the same time, 18 primary school students living far away needed to temporarily join the shuttle bus queue due to sudden work arrangements of their parents. Due to the lack of real-time data access interfaces, traditional planning methods often had to execute according to the original route, resulting in a large number of empty seats at the originally scheduled junior high school stop, while the pick-up points for the newly demanded primary school students were not included in the route, ultimately causing a double problem of waste of transport capacity resources and extended waiting times for some students.

[0003] In addition, in the face of the regular morning and evening rush hour congestion around the school and occasional temporary traffic control, the route optimization strategies of traditional techniques are relatively rigid. For example, on a Wednesday evening rush hour, when the shuttle bus was driving along the established route to a crossroads 3 kilometers away from the school, a one-way road closure occurred due to a sudden traffic accident ahead, and this crossroads was the main passage connecting the student residential areas. Traditional planning techniques rely on preset road lengths and average travel times, do not obtain real-time dynamic road closure information from the traffic management department, and cannot dynamically adjust parameters based on real-time congestion data. At this time, the shuttle bus may continue to drive towards the closed road intersection and be forced to detour until it is close. During the detour process, since the final stop order of surrounding communities is not recalculated, the original 40-minute commute time is extended to 75 minutes, and some students arrive home nearly half an hour later than usual. Moreover, there are multiple repeated detour sections in the detour route, increasing vehicle energy consumption and operating costs. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a vehicle path planning method based on an adaptive optimization algorithm, which can respond to dynamic riding demands and road condition changes in school shuttle buses in real time and achieve dynamic path adjustment.

[0005] To solve the above technical problem, the technical solution of the present invention is as follows:

[0006] In a first aspect, a vehicle path planning method based on an adaptive optimization algorithm, the method includes:

[0007] Step 1, based on the current order requirements, vehicle resources, and preset constraint conditions, generate an initial path plan set that meets the time window, splitting times, vehicle type matching, and temperature zone requirements through a step-by-step progressive decision-making process;

[0008] Step 2, according to the initial path plan set, start multiple independent optimization links. Each link dynamically adjusts local path nodes through a greedy strategy to generate an optimized candidate plan set;

[0009] Step 3, conduct real-time evaluation on the optimized candidate plan set, determine the final candidate plan based on the cost savings rate, demand satisfaction rate, and soft service quality indicators, and update the final candidate plan as the current benchmark plan;

[0010] Step 4, according to the current benchmark plan, when a dynamic event occurs within the closed triangular area composed of the order distribution center detection point, vehicle dispatching center detection point, and transportation hub detection point, trigger local link optimization, and generate dynamic correction parameters according to the real-time position relationship between the triangle vertices;

[0011] Step 5, according to the dynamic correction parameters, calculate the geometric centroid offset of the affected path segment, perform directional adjustment on the offset path segment, update the global plan library, and generate a path adjustment result covering the dynamic area;

[0012] Step 6, based on the updated global plan library, set a plan survival life threshold. When the plan optimization period ≥ threshold, combine the dynamic area topology association characteristics to eliminate old plans, and generate alternative optimization links based on the dynamic correction parameters.

[0013] Furthermore, based on the current order requirements, vehicle resources, and preset constraint conditions, generate an initial path plan set that meets the time window, splitting times, vehicle type matching, and temperature zone requirements through a step-by-step progressive decision-making process, including:

[0014] According to the temperature zone attribute and time window limit of the order, conduct multi-dimensional clustering on the current order requirements to generate multiple order subsets with the same temperature zone and a time window overlap degree ≥ preset threshold;

[0015] Based on the order subsets, traverse the vehicle types in the vehicle resource library that meet the temperature zone adaptation conditions, determine a candidate vehicle set whose vehicle capacity covers the total demand of the orders within the subset, and perform a splitting operation on the order subsets with > single loading capacity to ensure that the splitting times ≤ preset threshold, and generate a task set to be assigned including the split order units;

[0016] Match the task set to be assigned with the candidate vehicle set, and generate an initial path segment that meets the vehicle loading capacity constraint according to the distance priority between the vehicle's real-time position and the order delivery point;

[0017] Perform time window conflict detection on the initial path segment. If there is a time window conflict, dynamically adjust the node order of the path segment based on the vehicle driving speed to generate a set of conflict-free feasible path segments;

[0018] Combine and optimize the paths in the set of feasible path segments that are in the same temperature zone and geographically adjacent to generate a set of initial path plans including multi-vehicle collaborative scheduling and multi-order combined distribution.

[0019] Furthermore, according to the set of initial path plans, start multiple independent optimization links. Each link dynamically adjusts the local path nodes through a greedy strategy to generate a set of optimization candidate plans, including:

[0020] Extract the nodes with a path cost ratio > the preset ratio from the set of initial path plans as highly sensitive adjustment objects to generate a set of candidate nodes;

[0021] For different optimization links, based on the distance deviation between the set of candidate nodes and the current vehicle position, assign different node adjustment priorities to each link to generate a link-specific local adjustment range;

[0022] Within the local adjustment range, perform a greedy operation on the highly sensitive nodes. If the time window margin of the path where the node is located is sufficient, try to insert adjacent unassigned orders; if the time window is tight, exchange with adjacent path nodes to generate candidate sub-plans within the link;

[0023] Verify the conflicts of the candidate sub-plans generated by each link, eliminate the plans that conflict with the vehicle temperature zone adaptability, and merge the remaining sub-plans into a set of optimization candidate plans.

[0024] Furthermore, the different optimization links include that the first link dynamically calculates the deviation amplitude difference based on the real-time distance deviation between the candidate nodes and the current vehicle position to generate a distance-sensitive adjustment priority sequence; the second link quantifies the path interaction intensity based on the geographical overlap rate between the path where the candidate node is located and the adjacent path to generate an overlap-sensitive adjustment priority sequence.

[0025] Furthermore, perform real-time evaluation on the set of optimization candidate plans, determine the final candidate plan based on the cost savings rate, demand satisfaction rate, and soft service quality indicators, and update the final candidate plan as the current benchmark plan, including:

[0026] For each plan in the set of optimization candidate plans, calculate the difference between the total transportation cost and the historical benchmark plan, determine the plans with a cost savings rate ≥ the preset threshold, and form a set of primary cost reduction plans;

[0027] Verify the demand coverage of the primary fee reduction plan set, detect whether each plan covers all key order nodes. If there are urgent orders that are not covered, mark them as plans with missing requirements and eliminate them, and generate a secondary demand compliance plan set;

[0028] Conduct a service quality assessment on the secondary demand compliance plan set. Based on the order delivery on-time rate and the frequency of route adjustment, calculate the comprehensive service quality score, and generate a plan priority queue according to the score ranking;

[0029] Determine the candidate plan with the comprehensive score at the final rank from the plan priority queue, synchronously activate the compensation mechanism for unmet demands, and re-inject the unallocated orders in the eliminated plans with missing requirements into the initial route plan set to generate a pool of orders to be re-allocated;

[0030] Update the final candidate plan to the current benchmark plan and trigger the route re-planning link of the pool of orders to be re-allocated.

[0031] Furthermore, conduct a service quality assessment on the secondary demand compliance plan set. Based on the order delivery on-time rate and the frequency of route adjustment, calculate the comprehensive service quality score, and generate a plan priority queue according to the score ranking, including:

[0032] Extract the matching deviation between the actual delivery time of the orders in each plan and the preset time window, and use the ratio of the total deviation duration to the total number of orders as the benchmark index to generate the on-time rate ranking sequence of each plan;

[0033] Based on the on-time rate ranking sequence, count the number of times the node order is adjusted during the route optimization process for each plan to generate an adjustment frequency sequence associated with the on-time rate;

[0034] According to the on-time rate ranking sequence and the adjustment frequency sequence, determine the plans with the total on-time rate deviation < preset upper limit and the adjustment frequency < preset limit value to form a preliminary preferred plan set;

[0035] For the plans within the preliminary preferred plan set, first sort them in ascending order of the total on-time rate deviation. If the total deviation is the same, perform a secondary sorting in ascending order according to the adjustment frequency, and extract the number of un-served orders within the route coverage area, and generate the final plan priority queue in ascending order of the number of un-served orders.

[0036] Furthermore, according to the current benchmark plan, when a dynamic event occurs within the closed triangular area composed of the detection points of the order distribution center, the vehicle dispatching center, and the transportation hub detection points, trigger local link optimization, and generate dynamic correction parameters according to the real-time position relationship between the triangle vertices, including:

[0037] Collect the position data of three detection points in the triangular area in real time. If the offset of the real-time coordinates of any detection point from the reference position > the preset tolerance threshold, it is determined as a dynamic disturbance event and the coordinate sequence is extracted;

[0038] Based on the coordinate sequence, calculate the real-time change rate of the side lengths of the triangle and the dynamic offset angle of the vertex angle, and generate a dynamic event influence intensity index in combination with the current path node distribution density;

[0039] According to the dynamic event influence intensity index and the path node distribution density in the area passed by the current reference scheme, dynamically calculate the path redirection priority and the detour compensation distance, and generate dynamic correction parameters.

[0040] Further, according to the dynamic correction parameters, calculate the geometric centroid offset of the affected path segment, perform directional adjustment on the offset path segment and then update the global scheme library to generate a path adjustment result covering the dynamic area, including:

[0041] Based on the path redirection priority in the dynamic correction parameters, extract the affected path segments in the triangular area and calculate the offset of the geometric centroid from the preset reference position;

[0042] According to the offset direction and the detour distance compensation value, perform a directional adjustment operation on the affected path segment to obtain the adjusted path segment;

[0043] Perform a constraint penetration verification on the adjusted path segment to detect whether the detour nodes > the vehicle temperature zone adaptation range and whether the path extension violates the split times limit;

[0044] Stitch the verified adjusted path segments with the unaffected paths, update the global scheme library and generate a path adjustment result covering the dynamic area.

[0045] Further, based on the updated global scheme library, set a scheme survival life threshold. When the scheme optimization period ≥ the threshold, combine the dynamic area topological association characteristics to eliminate the old scheme, and generate an alternative optimization link based on the dynamic correction parameters, including:

[0046] Dynamically calculate the scheme survival life threshold according to the frequency of the dynamic event and the stability of the path adjustment result, and compare it with the current scheme optimization period;

[0047] If the optimization period ≥ the threshold, extract the matching degree of the path node and the offset of the current area centroid, and eliminate the schemes with a matching degree < the preset critical value;

[0048] Based on the detour compensation distance and priority classification rules set in the dynamic correction parameters, alternative optimized paths are dynamically generated, including giving priority to reusing path segments where verified detour path segments exist and achieving smooth connection with the current real-time position of the vehicle; if there are no available detour segments, a set of alternative nodes is generated based on the detour compensation distance threshold to reconstruct the path.

[0049] In a second aspect, a computer-readable storage medium stores a program, which implements the method when executed by a processor.

[0050] The above solution of the present invention includes at least the following beneficial effects:

[0051] By dynamically adjusting parameters in real time through adaptive optimization algorithms, it can quickly adapt to complex and changing environments such as traffic conditions and order requirements. Compared with traditional algorithms, it can improve the efficiency of path planning by more than 30%, shorten the calculation time, improve the real-time and accuracy of path planning, and ensure that delivery tasks are completed on time. This method comprehensively considers multiple cost factors such as vehicle load, driving distance, and fuel consumption. By optimizing the path, it reduces invalid mileage, fuel consumption, and vehicle loss. The adaptive mechanism enables the algorithm to automatically adjust the search strategy according to the actual situation. Whether facing sudden traffic control, temporary new orders, or unexpected vehicle failures, it can quickly re-plan the path to ensure the stable operation of the distribution network, greatly enhancing the path planning system's ability to adapt to complex scenarios.

[0052] Make full use of the carrying capacity of each vehicle, avoid empty or overloaded vehicles, improve vehicle utilization, achieve reasonable allocation of resources, and reduce vehicle investment while ensuring the quality of distribution services. Efficient and accurate route planning ensures that goods can be delivered on time and accurately, reduces delivery delays and errors, and improves customer satisfaction with logistics services. The large amount of data accumulated during the route planning process provides rich operational information, and potential problems and optimization space can be discovered through data analysis to formulate long-term development strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flowchart of a vehicle path planning method based on an adaptive optimization algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0055] likeFigure 1 As shown in Figure 1 , an embodiment of the present invention provides a vehicle path planning method based on an adaptive optimization algorithm, and the method includes the following steps:

[0056] Step 1: Based on the current order requirements, vehicle resources, and preset constraint conditions, generate an initial path plan set that meets the time window, splitting times, vehicle type matching, and temperature zone requirements through a step-by-step progressive decision-making process;

[0057] Step 2: According to the initial path plan set, start multiple independent optimization links, and each link dynamically adjusts local path nodes through a greedy strategy to generate an optimized candidate plan set;

[0058] Step 3: Evaluate the optimized candidate plan set in real time, determine the final candidate plan based on the cost savings rate, demand satisfaction rate, and soft service quality indicators, and update the final candidate plan to the current benchmark plan;

[0059] Step 4: According to the current benchmark plan, when a dynamic event occurs within the closed triangular area composed of the order distribution center detection point, vehicle dispatching center detection point, and transportation hub detection point, trigger local link optimization, and generate dynamic correction parameters according to the real-time position relationship between the triangle vertices;

[0060] Step 5: According to the dynamic correction parameters, calculate the geometric centroid offset of the affected path segment, perform directional adjustment on the offset path segment, update the global plan library, and generate a path adjustment result covering the dynamic area;

[0061] Step 6: Based on the updated global plan library, set a plan survival life threshold. When the plan optimization period ≥ threshold, eliminate old plans in combination with the topological association characteristics of the dynamic area, and generate alternative optimization links based on the dynamic correction parameters.

[0062] In an embodiment of the present invention, based on order requirements, vehicle resources and preset constraints, a step-by-step progressive decision-making process is used to generate an initial path solution set, which can ensure that the solution accurately meets the time window, split times, vehicle model matching and temperature zone requirements. This process ensures the compliance and feasibility of path planning, avoids resource waste and delivery delays caused by the disconnection between the solution and actual operating conditions, and improves the initial quality and reliability of the solution. Start multiple independent optimization links, use the greedy strategy to dynamically adjust the local path nodes, and generate a set of optimized candidate solutions. Through parallel processing to improve optimization efficiency, the greedy strategy can quickly capture the local final solution and explore the potential optimization space in the initial solution. Different links optimize the path from multiple perspectives, increase the diversity of candidate solutions, and provide rich choices for obtaining better path solutions. The set of optimized candidate solutions is evaluated in real time based on the cost saving rate, demand satisfaction rate and soft service quality indicators, and the final candidate solution is determined and updated to the current benchmark solution. The evaluation system takes into account cost, demand and service quality, ensures that the solution is the final solution of comprehensive benefits, updates it to the benchmark solution, realizes effective control of operating costs, and improves economic benefits and market competitiveness.

[0063] When a dynamic event is detected in the closed triangle area formed by the order distribution center, vehicle dispatch center and transportation hub detection point, local link optimization is triggered, and dynamic correction parameters are generated according to the real-time position relationship between the triangle vertices. This dynamic event perception mechanism based on spatial topology can accurately locate the affected area and start targeted optimization in time. The generation of dynamic correction parameters ensures that the optimization strategy is closely related to the actual dynamic changes, improves the response speed and accuracy of the path planning system to emergencies, and ensures the stability of the delivery task. The geometric center of gravity offset of the affected path segment is calculated according to the dynamic correction parameters, and the global solution library is updated after the offset path segment is adjusted in a targeted manner to generate the path adjustment result covering the dynamic area. By quantifying the geometric center of gravity offset, the path adjustment can be accurately controlled to avoid the additional cost caused by excessive adjustment. While ensuring the effectiveness of the path, the targeted adjustment reduces the impact on other path segments, quickly generates new solutions that adapt to dynamic changes and updates the global solution library to ensure the consistency and timeliness of global path planning. The survival threshold of the solution is set based on the updated global solution library. When the solution optimization cycle ≥ the threshold, the old solution is eliminated in combination with the dynamic regional topology association characteristics, and an alternative optimization link is generated based on the dynamic correction parameters. This mechanism establishes a mechanism for the survival of the fittest among solutions, preventing outdated solutions from affecting planning results due to their inability to adapt to environmental changes. It generates alternative links based on the dynamic regional topology association characteristics to ensure that new solutions can seamlessly connect with dynamic changes and continuously maintain the advancement and efficiency of path planning solutions.

[0064] In a preferred embodiment of the present invention, in step 1, based on the current order requirements, vehicle resources, and preset constraint conditions, an initial path plan set that meets the time window, splitting times, vehicle type matching, and temperature zone requirements is generated through a step-by-step progressive decision-making process, which may include:

[0065] Step 110, according to the temperature zone attribute and time window limit of the order, perform multi-dimensional clustering on the current order requirements to generate multiple order subsets with the same temperature zone and a time window overlap degree ≥ preset threshold;

[0066] Step 111, based on the order subset, traverse the vehicle types in the vehicle resource library that meet the temperature zone adaptation conditions, determine a candidate vehicle set that can cover the total demand of the orders within the subset by vehicle capacity, and perform a splitting operation on the order subset with a demand > single loading capacity to ensure that the splitting times ≤ preset threshold, generating a task set to be assigned including the split order units;

[0067] Step 112, match the task set to be assigned with the candidate vehicle set, and generate an initial path segment that meets the vehicle type loading capacity constraint according to the distance priority between the real-time position of the vehicle and the order delivery point;

[0068] Step 113, perform a time window conflict detection on the initial path segment. If there is a time window conflict, dynamically adjust the node order of the path segment based on the vehicle driving speed to generate a set of conflict-free feasible path segments;

[0069] Step 114, combine and optimize the paths with the same temperature zone and geographical proximity in the set of feasible path segments to generate an initial path plan set including multi-vehicle type collaborative scheduling and multi-order combined distribution.

[0070] In the embodiment of the present invention, all current order data is extracted from the order management system, including order numbers, temperature zone attributes (such as normal temperature, refrigerated, frozen, etc.), delivery addresses, time window limit (earliest delivery time and latest delivery time) information. Using the temperature zone attribute as the primary clustering dimension, empty sets of different temperature zone categories (such as normal temperature zone, refrigerated zone, frozen zone) are created, and all orders are traversed and classified into the corresponding temperature zone category sets according to the order temperature zone attributes. Within each temperature zone category set, for any two orders and , calculate the overlapping duration of their time windows . Assume the time window of order is , the time window of order is , then the overlapping duration . At the same time, calculate the total duration , the overlap degree , where Represents the actual overlapping duration of the time windows of two orders. Represents the total coverage duration after merging the time windows of two orders. When the overlap degree ≥ the preset threshold (such as 70%), mark the orders and as clusterable. Through a clustering algorithm (such as the DBSCAN algorithm), based on the orders marked as clusterable, generate multiple order subsets within each temperature zone category set. The orders within each order subset have the same temperature zone and the time window overlap degree meets the requirements.

[0071] For each order subset, access the vehicle resource library to determine the vehicle models suitable for the temperature zone attributes of the order subset. For example, the refrigerated order subset corresponds to determining the refrigerated vehicle model to form a list of candidate vehicle models. Calculate the total quantity of goods of all orders within the order subset and compare it with the loading capacity of each vehicle model in the list of candidate vehicle models. Select the vehicle models from the list of candidate vehicle models whose loading capacity can fully cover the total quantity of goods of the order subset to form a set of candidate vehicles. For the case where the total quantity of goods of the order subset > the single - vehicle loading capacity, adopt a heuristic splitting strategy. First, split the orders according to the principle of similar goods type and weight, ensuring that the quantity of goods of each order unit after splitting < the single - vehicle loading capacity, and at the same time ensuring that the number of splits < the preset threshold (such as 3 times). After splitting, integrate the split order units with the original orders to generate a set of tasks to be assigned. Use the Geographic Information System (GIS) to obtain the real - time position coordinates (longitude, latitude) of each vehicle in the set of candidate vehicles, and the position coordinates of each order delivery point in the set of tasks to be assigned. Calculate the straight - line distance between the real - time position of the vehicle and the order delivery point through the Haversine formula , where is the radius of the earth, , are the latitude and longitude of the vehicle position, , are the latitude and longitude of the order delivery point. Sort the combinations of vehicles and orders in ascending order according to the calculated distance. The closer the distance, the higher the priority. According to the priority order, assign the order units in the set of tasks to be assigned to the candidate vehicles in sequence. Plan a path for each vehicle to start from the current position and go to the assigned order delivery points in sequence to generate an initial path segment that meets the vehicle model loading capacity constraint. During the assignment process, monitor the loading capacity of the vehicle in real time to avoid overloading.

[0072] For each initial path segment, based on the average driving speed of the vehicle, the distances between the delivery points of each order, and the time window constraints of the orders, calculate the estimated time for the vehicle to reach each order delivery point. If there is a situation where the estimated time for the vehicle to reach a certain order delivery point is earlier than the earliest delivery time or later than the latest delivery time, it is determined that there is a time window conflict in this path segment. When a time window conflict is detected, the simulated annealing algorithm is used to adjust the order of the delivery point nodes in the path segment. With minimizing the time window conflict as the objective function, by continuously trying new combinations of node orders, accepting or rejecting new combinations, gradually find the node order that minimizes the time window conflict, and generate a set of conflict-free feasible path segments. For the path segments in the set of feasible path segments, based on the position coordinates of the order delivery points, calculate the geographical distances between the path segments in the same temperature zone. By setting a distance threshold (such as 5 kilometers), determine the geographically adjacent path segments. For the determined geographically adjacent and same-temperature-zone path segments, considering the multi-vehicle collaborative scheduling and multi-order combined distribution strategies, use the genetic algorithm for combinatorial optimization. With the shortest total driving distance and the lowest distribution cost as the optimization objectives, merge and re-plan the path segments to generate a set of initial path plans including multi-vehicle collaborative scheduling and multi-order combined distribution.

[0073] Classify the orders with adapted temperature zones and time windows through multi-dimensional clustering to reduce the computational amount and improve the order processing efficiency. At the same time, ensure that the orders in the same order subset can better meet the temperature zone and time requirements during distribution, and improve the rationality of distribution. Determine the adapted vehicle types and reasonably split the orders to achieve an efficient match between vehicle resources and order requirements, avoid overloading or underloading of vehicles, improve vehicle utilization rate, and reduce transportation costs. Match according to the distance priority between the real-time position of the vehicle and the order delivery point to shorten the total driving mileage of the vehicle, reduce the distribution time, improve the distribution efficiency, and at the same time reduce the operating cost of fuel consumption. Effectively detect and solve the time window conflict problem to ensure that the orders can be delivered on time, improve customer satisfaction, enhance market competitiveness. Through the combinatorial optimization of the paths in the same temperature zone and geographically adjacent, realize multi-vehicle collaborative scheduling and multi-order combined distribution, further optimize the resource allocation, improve the vehicle loading rate, and reduce the distribution cost.

[0074] In a preferred embodiment of the present invention, for the above step 2, based on the set of initial path plans, start multiple independent optimization links. Each link dynamically adjusts the local path nodes through a greedy strategy to generate a set of optimization candidate plans, which may include:

[0075] Step 220, extract the nodes with a path cost ratio > a preset ratio from the set of initial path plans as high-sensitivity adjustment objects to generate a set of candidate nodes;

[0076] Step 221: For different optimization links, based on the distance deviation between the candidate node set and the current vehicle position, assign differential node adjustment priorities to each link to generate a locally exclusive adjustment range for the link, specifically including: The different optimization links include the first link that dynamically calculates the deviation amplitude difference based on the real-time distance deviation between the candidate node and the current vehicle position to generate a distance-sensitive adjustment priority sequence; the

[0077] second link that quantifies the path interaction intensity based on the geographical overlap rate between the path where the candidate node is located and the adjacent path to generate an overlap-sensitive adjustment priority sequence;

[0078] Step 222: Within the locally adjusted range, perform a greedy operation on the highly sensitive nodes. If the time window margin of the path where the node is located is sufficient, attempt to insert the adjacent unassigned order; if the time window is tight, exchange with the nodes on the adjacent path to generate a candidate sub-scheme within the link;

[0079] Step 223: Conduct conflict verification on the candidate sub-schemes generated by each link, eliminate the schemes that conflict with the vehicle temperature zone adaptability, and merge the remaining sub-schemes into an optimized candidate scheme set.

[0080] In the embodiment of the present invention, extract the detailed information of each path in the initial path scheme set from the database, specifically covering the specific route of vehicle travel, such as starting from the warehouse and passing through which streets and intersections in sequence to reach the order delivery point; the actual distance of each section, obtained through a map measurement tool; the fuel consumption coefficient of the vehicle in different sections, which can refer to the vehicle user manual; the toll standard corresponding to each section, querying the toll table published by the transportation department; the loss parameters corresponding to the vehicle model, such as the relationship between tire wear, component aging and the driving mileage; and the order information associated with each node (including order delivery points, transfer points, etc.), including the weight and volume of the goods, and the earliest and latest delivery times required by the customer. Conduct cost accounting for each path. The fuel consumption cost is calculated by multiplying the distance of each section by the fuel consumption coefficient of the vehicle in that section; the toll is directly determined according to the toll standard corresponding to the driving section; the vehicle loss cost is estimated based on the driving mileage and the vehicle model loss parameters. Add these three costs together to obtain the total path cost. Then, calculate the cost generated by the section corresponding to each node, and divide this cost by the total path cost to calculate the proportion of this node in the total path cost. For example, if the cost generated by the section corresponding to a certain node is 100 yuan and the total path cost is 500 yuan, then the cost proportion of this node is 20%. Set a preset proportion of the cost proportion, such as 35%. Check all the nodes of all paths one by one, select the nodes with a cost proportion > this preset proportion. These selected nodes have a greater impact on reducing the cost of the entire path when adjusted, and uniformly include them in the candidate node set.

[0081] With the help of Geographic Information System (GIS), obtain the accurate geographical location coordinates of each node in the candidate node set, as well as the real-time location coordinates of each vehicle participating in the path planning. Based on these coordinates, calculate the actual straight-line distance from the current location of the vehicle to each candidate node, and at the same time obtain the theoretical driving distance of the vehicle from its current location to each candidate node according to the planned route from the initial path planning scheme. For example, if the vehicle is currently at and the candidate node is at , through GIS calculation, the actual straight-line distance from to is 15 kilometers, while the theoretical distance traveled according to the planned route is 20 kilometers. For each candidate node, subtract the theoretical driving distance from the actual distance to obtain the distance deviation value. If the actual distance > the theoretical distance, it indicates that there may be unreasonable situations such as detours in the currently planned path when going to this node, which needs to be focused on; even if the actual distance < the theoretical distance, this deviation value is also recorded. For different optimization links, sort the candidate nodes according to the calculated distance deviation values, set different screening criteria for each link. For example, Link 1 specifically selects nodes with a distance deviation value > 10 kilometers, and Link 2 selects nodes with a distance deviation value between 5 - 10 kilometers. According to the priority order, delimit a dedicated local adjustment range for each link, and clarify the set of nodes that each link focuses on adjusting, so that each link can perform its own duties and work simultaneously during optimization, improving the overall optimization efficiency.

[0082] Within the locally adjustable range defined for each link, for each highly sensitive node, based on the average driving speed of the vehicle on each section, the distance between this node and adjacent nodes, as well as the earliest and latest delivery times of the orders corresponding to this node, calculate the time window margin when the vehicle arrives at this node. Specifically, subtract the estimated arrival time of the vehicle at this node from the latest delivery time of the order at this node, and set a judgment criterion. For example, when the time window margin > 1.5 times the estimated travel time to the next node, it is determined that the time window margin is sufficient; if this criterion is not met, it is determined that the time window is tight. For instance, if the vehicle is estimated to arrive at a certain node at 10 o'clock, the latest delivery time of the order at this node is 11 o'clock, and it is estimated to take 30 minutes to travel to the next node, and 1 hour > 1.5 times of 30 minutes (i.e., 45 minutes), then the time window margin is sufficient. When the time window margin of a certain highly sensitive node is sufficient, centered on this node, search for unassigned orders within a certain geographical range (such as a radius of 5 kilometers). For each unassigned order found, respectively simulate inserting it in front of or behind this node in the current path, and calculate the changes in the total cost of the path after insertion (including the increase or decrease in fuel consumption caused by the change in driving distance, the change in tolls, etc.) and the time. Prioritize the unassigned order that can reduce the total cost of the path after insertion and the arrival time of the vehicle at the subsequent nodes still meets the time window requirements of each node's order, and insert it into the current path to generate a new path plan, that is, the candidate sub-plan within the link.

[0083] When the time window of a certain highly sensitive node is tight, find the node adjacent to this node in the current path. Successively try to swap the positions of the highly sensitive node and the adjacent node, and calculate the change in the total cost of the path after the swap (involving aspects such as driving distance, fuel consumption, tolls, etc.), and whether the arrival time of the vehicle at each node can still meet the order time window requirements. If the total cost of the path decreases after the swap and the time window requirements of all nodes can be met, retain the path plan after the swap as the candidate sub-plan within the link. For all candidate sub-plans generated for each link, conduct a vehicle temperature zone adaptability check one by one. Extract the temperature zone requirements of the orders involved from the candidate sub-plans, such as normal temperature, refrigeration, and freezing, and then obtain the vehicle temperature zone type for implementing this plan. Compare the order temperature zone requirements with the vehicle temperature zone type. If there is a mismatch, such as there is a refrigerated order in the plan, but the assigned vehicle is a normal temperature vehicle, it is determined that this candidate sub-plan has a temperature zone adaptability conflict and remove it from the candidate plans. Collect the candidate sub-plans that pass the temperature zone adaptability conflict verification and put them into the optimized candidate plan set uniformly. The plans in this set have all undergone local optimization and meet the vehicle temperature zone adaptability requirements.

[0084] Precisely calculating the proportion of node costs and identifying highly sensitive nodes is like pointing out a precise direction for the optimization work, enabling the optimization resources to be concentrated on the key links that have the greatest impact on the path cost, avoiding meaningless adjustments to all nodes. At the same time, it clarifies the optimization focus and provides a clear and operable goal for reducing the path cost. Allocating different priorities and adjustment ranges to different optimization links based on the distance deviation realizes an efficient mode of simultaneous parallel optimization of multiple links. Each link focuses on optimizing nodes with different characteristics, making full use of computing resources, and eliminating duplicate work and resource waste in the optimization process. This method can comprehensively explore the optimization potential of path planning in all aspects, making the overall optimization effect better, and ultimately making the path planning result more reasonable, efficient, and more in line with the actual operation requirements. Flexibly executing the greedy operation according to the time window margin, inserting unassigned orders when the time window permits, making full use of the transportation capacity of the vehicle, improving the loading rate of the vehicle and the order delivery efficiency, reducing the situation of the vehicle running empty, and reducing the operation cost. Performing node exchange operations when the time window is tight ensures that the order can be delivered on time, avoiding delivery delays caused by time conflicts. While ensuring the timeliness of distribution, the cost is reduced by reasonably adjusting the path, improving the comprehensive benefits and practicality of the path planning scheme in actual applications. A strict temperature zone adaptability conflict verification mechanism eliminates the risk of cargo damage caused by vehicle temperature zone mismatch from the root, ensuring the quality and safety of the cargo and safeguarding the interests of customers. Merging the verified schemes into a set ensures that the finally determined scheme has high feasibility and effectiveness in actual applications.

[0085] In a preferred embodiment of the present invention, in step 3 above, for real-time evaluation of the set of optimization candidate solutions, determining the final candidate solution based on the cost savings rate, demand satisfaction rate, and soft service quality index, and updating the final candidate solution to the current baseline solution may include:

[0086] Step 330, for each solution in the set of optimization candidate solutions, calculate the difference between the total transportation cost and the historical baseline solution, determine the solutions with a cost savings rate ≥ the preset threshold, and form a primary cost reduction solution set;

[0087] Step 331, conduct demand coverage verification on the primary cost reduction solution set, detect whether each solution covers all key order nodes. If there are uncovered urgent orders, mark them as demand missing solutions and eliminate them to generate a secondary demand compliance solution set;

[0088] Step 332: Conduct a service quality assessment on the set of secondary demand compliance solutions. Based on the order delivery on-time rate and the frequency of route adjustments, calculate the comprehensive service quality score, and generate a solution priority queue according to the score ranking, which specifically includes: extracting the matching deviation between the actual delivery time of the orders in each solution and the preset time window, taking the ratio of the total deviation duration to the total number of orders as the benchmark index, and generating an on-time rate ranking sequence for each solution;

[0089] Based on the on-time rate ranking sequence, count the number of node order adjustments triggered during the route optimization process for each solution, and generate an adjustment frequency sequence associated with the on-time rate;

[0090] According to the on-time rate ranking sequence and the adjustment frequency sequence, determine the solutions with the total on-time rate deviation < preset upper limit and the adjustment frequency < preset limit value, and form a preliminary preferred solution set;

[0091] For the solutions within the preliminary preferred solution set, first sort them in ascending order of the total on-time rate deviation. If the total deviations are the same, perform a secondary sorting in ascending order according to the adjustment frequency, and extract the number of un-served orders within the route coverage area, and generate the final solution priority queue in ascending order of the number of un-served orders;

[0092] Step 333: Determine the candidate solution with the comprehensive score at the final rank from the solution priority queue, synchronously activate the compensation mechanism for unmet demands, and re-inject the unallocated orders in the eliminated demand-lacking solutions into the initial route solution set to generate a pool of orders to be re-allocated;

[0093] Step 334: Update the final candidate solution as the current benchmark solution, and trigger the route re-planning link for the pool of orders to be re-allocated.

[0094] In the embodiment of the present invention, automatically connect to the transportation cost database, extract the detailed cost data of each solution in the set of optimization candidate solutions. For fuel costs, retrieve the actual fuel consumption records of the corresponding vehicles in different sections of each solution from the vehicle fuel consumption monitoring system, and calculate the total fuel cost in combination with the current oil price; obtain the tolls by querying the toll details of the driving sections of each solution in the electronic toll collection system; calculate the vehicle depreciation cost based on the vehicle driving mileage, service life, and depreciation standard table. At the same time, retrieve the total transportation cost data of the past benchmark solutions, which is stored in the past solution database and contains the total cost information of the solutions recognized as relatively optimal within a past period of time. For each candidate solution, calculate according to the formula "cost savings rate = × 100%". For example, if the total cost of the past benchmark solution is 8000 yuan, and the fuel cost of a certain candidate solution is 2500 yuan, the toll is 1200 yuan, and the vehicle depreciation cost is 800 yuan, and its total cost is 4500 yuan, then the cost savings rate of this solution is × 100% = 43.75%. Automatically calculate the cost savings rate for all candidate solutions. The enterprise presets a cost savings rate threshold, such as 25%, and compares the cost savings rate of each candidate solution with this threshold one by one. Determine the solutions with a cost savings rate ≥ 25% to form a primary cost reduction solution set. For the solutions that do not reach the threshold, automatically mark and remove them from the scope to be evaluated, and at the same time generate an analysis report on the non-compliant solutions to explain the reasons for non-compliance. The order management department comprehensively evaluates all orders based on multi-dimensional information such as the urgency of the order, the value of the goods, and the importance of the customer. For urgent orders, such as urgent document orders where the customer requires delivery within 2 hours; high-value goods orders, such as electronic product orders with a value > 100,000 yuan; and important customer orders, such as orders from long-term cooperative VIP customers, determine them as key order nodes. Organize the information of these key order nodes into a detailed list, including order number, customer name, goods information, and urgency. Analyze each solution in the primary cost reduction solution set one by one. Through the Geographic Information System (GIS) and order matching algorithm, compare the order delivery points in the solution with the key order node list. The specific operation is to extract the coordinates of all order delivery points in the path for each solution and match them with the coordinates of the key order nodes. If there are coordinates of key order nodes that do not appear in the solution path and the order is an urgent order, then determine that the solution is a solution lacking requirements. For example, if a delivery point of an urgent medicine delivery order is not included in a certain solution path, then this solution is marked as a solution lacking requirements. Automatically remove the detected solutions lacking requirements from the primary cost reduction solution set and generate a list of removed solutions, detailing the names, numbers, and missing key order information of the removed solutions. The remaining solutions form a secondary requirement compliance solution set to ensure that the solutions in this set cover at least all key order nodes.

[0095] Obtain the order delivery on-time rate data corresponding to each solution in the secondary requirement compliance solution set from the order delivery record database. The specific statistical method is to calculate the ratio of the number of orders delivered on time within a past period (such as one month) to the total number of orders for each solution. For example, if a certain solution delivered a total of 150 orders in the past month, and 135 of them were delivered on time, then the on-time rate is × 100% = 90%. At the same time, obtain the path adjustment frequency data of each plan during the actual execution process from the path adjustment monitoring system. This data records the time, reason, and adjusted path information of each path adjustment in real time. Set the service quality score calculation rule, such as "Comprehensive service quality score = on-time delivery rate of orders × 0.7 - path adjustment frequency × 0.3". According to this rule, calculate the comprehensive score for each plan. Taking a certain plan as an example, if its on-time rate is 85% and the path adjustment frequency is 3 times, then the comprehensive score = 85% × 0.7 - 3 × 0.3 = 0.595 - 0.9 = -0.305. Automatically complete the comprehensive score calculation for all plans. According to the calculated comprehensive service quality scores, sort the plans in the secondary demand compliance plan set from high to low. The higher the score of a plan, the more forward its position in the queue. Generate a plan priority queue list to clearly show the ranking of each plan. At the same time, provide a detailed score comparison analysis chart to facilitate managers to view the differences in service quality among each plan. Directly select the candidate plan with the highest comprehensive score from the plan priority queue as the final candidate plan. This plan achieves a relatively final balance in terms of cost savings, demand coverage, and service quality. Automatically generate a final candidate plan report, which details the various indicators, advantages of the plan, and its comparison with other plans. For the demand missing plans eliminated in step 331, summarize and organize the unassigned order information among them. Through the order management system and the initial path plan set database, re-inject these unassigned orders into the initial path plan set to form a pool of orders to be reallocated. At the same time, automatically notify the path planners and dispatchers, indicating that there are new orders that need to be reallocated and providing the detailed information of the orders to be reallocated.

[0096] Update all the information of the determined final candidate plan, including path planning, vehicle allocation, estimated cost, and order delivery sequence, to the current baseline plan database, replacing the original baseline plan information, and generate an update record log, recording the update time, operator, and update content. At the same time, automatically send a baseline plan update notice to relevant departments and personnel to ensure that all parties are promptly informed of the latest path planning baseline. For the orders in the pool of orders to be reallocated, automatically initiate a dedicated path re-planning link. First, obtain the current available vehicle information from the vehicle resource management system, including vehicle location, load capacity, remaining driving mileage, etc.; then, combined with the road congestion situation and traffic restriction information data provided by the real-time traffic information system, use the path planning algorithm to re-plan the paths of these orders. During the planning process, refer to the relevant parameters of the current baseline plan to ensure that the newly planned paths are coordinated with the overall distribution strategy. Finally, generate an optimized re-allocation plan, which is the current baseline plan.

[0097] By precisely calculating the cost savings rate and strictly determining the plan according to the threshold, it is possible to quickly and accurately identify the plans with advantages in cost control. This enables enterprises to concentrate resources on the plans that can truly reduce transportation costs during the path planning decision-making process, avoiding wasting time and effort on plans with poor cost-effectiveness, and directly saving a large amount of operating costs for enterprises. At the same time, the detailed calculation and analysis process of cost data provides a reliable basis for enterprise cost management and budget formulation. The strict demand coverage verification mechanism fundamentally ensures that key order nodes can be included in the distribution plan, avoiding business losses caused by omitting urgent orders or important customer orders, such as customer complaints and order cancellations, effectively maintaining the enterprise's service reputation and customer relationship. By eliminating plans with missing requirements, it ensures the integrity and reliability of the plans in the secondary demand compliance plan set in terms of order delivery, improving the stability and customer satisfaction of the enterprise during the order processing process. Conducting a comprehensive service quality assessment based on the order delivery on-time rate and the frequency of path adjustments fully considers the two key factors of the timeliness and stability of the distribution service. Through scientific comprehensive scoring calculation and ranking, it is possible to clearly and accurately distinguish the service quality advantages and disadvantages of each plan. Enterprises can give priority to choosing plans with high service quality, thereby effectively improving customer satisfaction and enhancing the competitiveness of the enterprise in the market.

[0098] In a preferred embodiment of the present invention, step 4, according to the current reference plan, when a dynamic event occurs within the closed triangular region composed of the order distribution center detection point, the vehicle scheduling center detection point, and the transportation hub detection point, triggering local link optimization and generating dynamic correction parameters according to the real-time position relationship between the triangle vertices, may include:

[0099] Step 440, collecting the position data of the three detection points within the triangular region in real time. If the offset of the real-time coordinate of any detection point from the reference position > the preset tolerance threshold, it is determined as a dynamic disturbance event and the coordinate sequence is extracted;

[0100] Step 441, based on the coordinate sequence, calculating the real-time change rate of each side length of the triangle and the dynamic offset angle of the vertex angle, and generating a dynamic event influence intensity index in combination with the current path node distribution density;

[0101] Step 442, according to the dynamic event influence intensity index and the path node distribution density of the area passed by in the current reference plan, dynamically calculating the path redirection priority and the detour compensation distance, and generating dynamic correction parameters.

[0102] In the embodiments of the present invention, high-precision GPS positioning devices and Internet of Things sensors are deployed at order collection and distribution centers, vehicle dispatching centers, and transportation hubs, and the position data of the detection points are collected once per second, including longitude, latitude information accurate to the meter level, and altitude data. These data are transmitted to the dedicated database for path planning in real time through the network. The database is equipped with a data verification mechanism to automatically eliminate abnormal data (such as data with obvious jumps or exceeding reasonable ranges) to ensure the accuracy and effectiveness of the collected data. The reference position coordinates of three detection points are retrieved from the current reference scheme database, and the offset calculation is performed for each detection point. Taking the detection point as an example, assuming its reference longitude is , the real-time longitude is , the reference latitude is , the real-time latitude is , and the offset distance on the plane is calculated using the Pythagorean theorem . The tolerance threshold is preset to 50 meters. When the offset distance of any detection point > this threshold, it is determined that a dynamic disturbance event has occurred in this area. For example, if the calculated result of the offset distance of the detection point is 60 meters, the event response mechanism is immediately triggered. Once it is determined that a dynamic disturbance event has occurred, all coordinate data of the three detection points from 5 minutes before the event to the current moment are automatically extracted, and a continuous coordinate sequence is formed in chronological order. These coordinate sequences are stored in a dedicated event data storage area and marked with the time stamp and unique identifier of the event occurrence. Based on the extracted coordinate sequences, using the professional distance calculation algorithm of the Geographic Information System (GIS), the lengths of the three sides of the triangle at different moments are calculated respectively. For example, at the moment , the lengths of the three sides of the triangle are , , ; at the moment , the lengths of the three sides become , , . The real-time change rate of the side length is calculated according to the formula . Similarly, the change rates of the other two sides and are calculated. The rationality verification will be performed on the calculation results. If there are abnormal change rates (such as exceeding 50%), the calculation will be automatically recalculated or marked as suspicious data. Using the cosine theorem and arctangent function in trigonometric functions, according to the coordinates of the triangle vertices at different moments, the degrees of the angles at each vertex are accurately calculated. For example, at the moment , the angle at the vertex is . At the moment , this angle becomes . Through the formula Calculating vertices The dynamic offset angle of the included angle, and similarly calculate the vertices and the vertices The offset angle of the included angle and , to improve the calculation accuracy, interpolation processing will be performed on the coordinate data to ensure the accuracy of the angle calculation.

[0103] From the path data of the current baseline scheme, determine all path nodes located within the triangular region, and calculate the area of the triangular region through the spatial analysis function of GIS , and count the number of nodes , so as to obtain the node distribution density , calculate the side length change rate, included angle offset angle and node distribution density, set the weight of the side length change rate to 0.4, the weight of the included angle offset angle to 0.3, and the weight of the node distribution density to 0.3, through the formula , obtain the dynamic event impact intensity index . According to the dynamic event impact intensity index and the path node distribution density , set detailed priority calculation rules. When > 0.6 and > 10 nodes per square kilometer, the path redirection priority is set to high; when 0.3 < ≤ 0.6 and 5 nodes / square kilometer < ≤ 10 nodes per square kilometer, the priority is set to medium; when ≤ 0.3 and ≤ 5 nodes per square kilometer, the priority is set to low. At the same time, use the formula for quantitative calculation, and accurately divide the priority into three levels: high, medium, and low according to the calculation results. For example, if = 0.7, = 12 nodes per square kilometer, then = = 16.8, and the priority is determined to be high. Combining the dynamic event impact intensity index , comprehensively consider the vehicle driving speed (obtained in real time through in-vehicle sensors) and road conditions (obtaining real-time congestion data from the traffic information platform) factors, and calculate the detour compensation distance. For example, set the formula = , where is a coefficient dynamically adjusted according to the actual situation. When the traffic is smooth, = 100 meters / unit intensity; when the traffic is congested, = 150 meters / unit intensity. If the calculated = 0.5, under smooth traffic conditions, the detour compensation distance = 100 0.5 = 50 meters. Integrate the calculated path redirection priority and detour compensation distance to form a complete dynamic correction parameter. This parameter is stored in a structured data format, including a priority identifier and the numerical value of the detour compensation distance.

[0104] By collecting real-time position data of the detection points and comparing them with the reference position, dynamic disturbance events within the triangular area can be quickly and accurately captured. The second-level data collection frequency and precise offset calculation method ensure the timeliness and sensitivity of event detection, avoiding the failure of path planning caused by missed event detection. Extracting the coordinate sequence can comprehensively understand the potential impact of dynamic events on the path. Consider the triangle side length change rate, vertex angle offset angle, and path node distribution density to calculate the dynamic event impact intensity index , which quantifies the impact degree of dynamic events on path planning from multiple dimensions. This multi-factor comprehensive evaluation method is more scientific and comprehensive than single-index judgment, and can accurately reflect the actual impact range and severity of dynamic events. Generate dynamic correction parameters according to the dynamic event impact intensity and path node distribution, realizing the precision and dynamicization of the path adjustment strategy. The setting of the path redirection priority can reasonably arrange the order of path adjustment according to the severity of the event, and give priority to processing paths with greater impact; the calculation of the detour compensation distance provides specific reference values for vehicle detours, avoiding resource waste and time delays caused by blind detours. The generation of dynamic correction parameters improves the adaptive ability and optimization effect of the path planning system in the face of dynamic events, ensuring the smooth progress of the distribution task.

[0105] In a preferred embodiment of the present invention, in step 5 above, according to the dynamic correction parameter, calculate the geometric centroid offset of the affected path segment, perform directional adjustment on the offset path segment, and then update the global solution library to generate a path adjustment result covering the dynamic area, which may include:

[0106] Step 550, based on the path redirection priority in the dynamic correction parameter, extract the affected path segments within the triangular area, and calculate the offset of the geometric centroid from the preset reference position;

[0107] Step 551, according to the offset direction and detour distance compensation value, perform a directional adjustment operation on the affected path segment to obtain the adjusted path segment;

[0108] Step 552, perform a constraint penetration verification on the adjusted path segment to detect whether the detour node is > the vehicle temperature zone adaptation range, and whether the path expansion violates the split times limit;

[0109] Step 553: splice the verified adjusted path segments with the unaffected paths, update the global solution library, and generate the path adjustment result for dynamic area coverage.

[0110] In the embodiment of the present invention, according to the path redirection priority in the dynamic correction parameters, through the spatial query function of the Geographic Information System (GIS), the path segments within the triangular area are screened from the current baseline solution. If the priority is high, using the buffer analysis tool of GIS, the triangular area is expanded outward by 500 meters (which can be adjusted according to actual needs), and all path segments that completely or partially fall into the expanded area are extracted; if the priority is medium or low, combined with the coordinates of the dynamic event occurrence point, the Euclidean distance from each path segment node to the event occurrence point is calculated, and only the critical path segments with a distance < 2 kilometers (which can be adjusted according to actual needs) are extracted. During the screening process, the path segments that have been marked as invalid or abandoned are automatically excluded. For each extracted path segment, it is regarded as a polyline composed of a series of ordered nodes. Assume the path segment contains nodes, and the node has coordinates ([[]] , ). The weighted average algorithm is used to calculate the geometric centroid coordinates. If the node represents an order delivery point, different weights are assigned according to the order weight, and the greater the weight, the higher the priority; if it is a transfer point, the same weight is assigned. Through the formula , (where and represent the planar coordinates of the -th node in the path segment, is the abscissa, is the ordinate, represents the importance weight of the -th node, and the weight is proportional to the order weight, represents the total number of nodes in the path segment, and are the weighted geometric centroids of the path segment, that is, the balance points considering the node weights), after the calculation is completed, the accuracy verification is automatically performed. If the deviation between the centroid coordinates and the path segment node coordinates is too large (exceeding 10% of the path segment length), the node data is rechecked, and if necessary, manual intervention is carried out for correction. The preset reference position coordinates ([[]] , ) of the geometric centroid of the path segment are retrieved from the baseline solution database, and the actual offset is calculated based on WGS84 using the distance measurement function of GIS, and the result is accurate to meters. At the same time, the offset angle is calculated through the arctangent function , and the angle is accurate to 0.1 degrees. The data of the offset and the offset angle are recorded for analyzing the trend of the impact of dynamic events on the path. Combining the offset direction calculated in Step 550 , Dynamically correct the detour compensation distance in the parameters, and use real-time traffic flow data (obtained from the traffic management department API) to determine the adjustment direction. If the offset direction points outside the dynamic event area and the real-time traffic flow in this direction of the road < threshold (such as the traffic volume per hour < 500 vehicles), then adjust along this direction; if there are multiple feasible directions, construct a multi-objective optimization with the goals of avoiding dynamic events, minimizing the increase in the total driving distance, and minimizing the expected driving time, and solve the final adjustment direction through a genetic algorithm. During the calculation process, consider factors such as road speed limits and turning restrictions to ensure the feasibility of the adjustment direction.

[0111] When it is necessary to increase the detour distance, according to the adjustment direction and the detour compensation distance, insert a new node at an appropriate position in the path segment. First, calculate the time required for the vehicle to travel the detour compensation distance at the current average speed (obtained in real time from the on-vehicle sensor). , and then insert a new node at the position corresponding to the time from the end point of the original path segment. The coordinates of the new node are obtained by taking the end point of the original path segment as the starting point, along the adjustment direction, and calculating the displacement according to the driving time and speed, so as to obtain the coordinates of the new node. After inserting the new node, recalculate the geometric centroid and total length of the path segment to ensure that the adjustment requirements are met. For the case where only the direction needs to be adjusted, use the simulated annealing algorithm to move some nodes in the path segment, set the maximum step size of node movement (such as not exceeding 50 meters) and the temperature decay parameter to avoid over-adjustment. After each node movement, calculate the distance between the adjusted path segment and the dynamic event area, and the change in the total length of the path segment. If the adjusted path not only avoids the dynamic event but also minimizes the increase in the total length, then retain this adjustment; otherwise, continue iterative adjustment until the conditions are met. Obtain the temperature zone requirements (such as normal temperature 0 -30 , refrigerated -18 -0 , frozen < -18 of the orders associated with the detour nodes involved from the order information database. At the same time, obtain the temperature zone type and temperature zone adjustment range of the vehicle executing this path from the vehicle information database. Compare the order temperature zone requirements and the vehicle temperature zone type one by one. If there is a mismatch, such as the order requires frozen transportation but the vehicle only supports normal temperature transportation, then it is determined that the adjusted path segment fails the temperature zone adaptability verification. For the case where the temperature zone requirements are similar but not exactly matched (such as the order requires refrigerated and the vehicle temperature zone is -15 ), evaluate the damage risk probability of the goods in this temperature zone according to the data. If the risk probability > threshold (such as 5%), it is also determined that the verification fails. By analyzing the connection relationship of the path segment nodes, count the number of path splits caused by the adjusted path segment in the entire global solution. The preset limit on the number of path splits is 3 times. If the number of splits > this limit after the path is extended, analyze the impact of the split on the distribution efficiency, calculate the expected driving time and vehicle waiting time indicators of each sub-path after the split. If the sum of these indicators > 120% of the original path's expected total time (which can be adjusted according to actual needs), then it is determined that the adjusted path segment fails the verification of the split number limit.

[0112] According to the unique identifier and connection order of the path nodes, splice the verified adjusted path segment with the unaffected paths in the global solution library, and use the shortest path algorithm in graph theory (such as Dijkstra algorithm) to ensure that the spliced path is final in terms of node connection and avoid unreasonable connections. During the splicing process, automatically update the access order and adjacent relationship of the path nodes, and recalculate the key parameters such as the total length and expected driving time of the spliced path. Overwrite the corresponding original path plan in the global solution library with the new spliced path, and at the same time update all information related to the path, including the total path length, expected driving time, access order of each node, and vehicle allocation, to generate a detailed path adjustment result report.

[0113] By extracting the affected path segments based on path redirection priority and spatial analysis technology and accurately calculating the geometric centroid offset, it is possible to quickly and accurately locate the scope and degree of the impact of dynamic events on the path from a large amount of path data, avoiding the waste of resources and low efficiency caused by blind adjustment. The importance of different nodes is fully considered, making the analysis results more in line with the actual business needs, improving the pertinence and accuracy of path adjustment, and ensuring that enterprises concentrate the optimized resources on the path segments that really need improvement. Determine the adjustment direction according to the offset direction, detour compensation distance, and real-time traffic data, and use scientific node insertion and movement algorithms to execute the adjustment operation, making the path adjustment have a clear goal and rigorous planning. While avoiding dynamic events, reasonably control the path extension distance through algorithms such as multi-objective optimization and simulated annealing, effectively reducing the increase in transportation costs and delivery time delays caused by unnecessary detours. In addition, the algorithm fully considers the actual road conditions and vehicle driving characteristics, ensuring the geometric and logical rationality of the adjusted path, improving the feasibility and practicality of the path, and making the adjusted path more suitable for the actual logistics distribution scenario. A strict constraint penetration verification mechanism checks the adjusted path segments from two key dimensions: vehicle temperature zone adaptability and path splitting times limit, providing double guarantees for the smooth development of the logistics distribution business. The temperature zone adaptability verification effectively avoids the risk of cargo damage caused by the mismatch between the vehicle and the order temperature zone, ensuring the quality and safety of the cargo. The splitting times limit verification ensures that the path adjustment will not be overly complex, avoiding problems such as scheduling chaos and decreased distribution efficiency caused by excessive path splitting, ensuring the executability and scheduling convenience of the path, and reducing the operation management difficulty and risk.

[0114] In a preferred embodiment of the present invention, for step 6 above, setting a scheme survival year threshold based on the updated global scheme library, when the scheme optimization period ≥ the threshold, combining the dynamic region topology association characteristics to eliminate the old scheme, and generating an alternative optimized link based on the dynamic correction parameter may include:

[0115] Step 660, dynamically calculate the scheme survival year threshold according to the frequency of dynamic events and the stability of the path adjustment result, and compare it with the current scheme optimization period;

[0116] Step 661, if the optimization period ≥ the threshold, extract the matching degree of the path node and the current regional centroid offset, and eliminate the scheme with a matching degree < the preset critical value;

[0117] Step 662, based on the detour compensation distance and priority classification rules set in the dynamic correction parameter, dynamically generate an alternative optimized path, including preferentially reusing the path segment for the existing verified detour path segment and realizing smooth connection with the current vehicle real-time position; if there is no available detour segment, generate an alternative node set according to the detour compensation distance threshold for path reconstruction.

[0118] In an embodiment of the present invention, dynamic event records within a triangular region composed of an order distribution center, a vehicle dispatching center, and a transportation hub in the past 12 months are retrieved from the event database. Each event record in the database includes event type (traffic accident, road construction, weather disaster), occurrence time, duration, and affected area information. The records are classified and statistically analyzed according to the event type. For example:

[0119] Traffic accidents: By screening the records with the event type field being "traffic accident", it is statistically found that there are a total of 30 occurrences, among which 12 occur during the morning rush hour and 18 occur during the evening rush hour;

[0120] Road construction: 20 "road construction" events are identified, involving 5 main transportation roads, and the average construction period is 15 days;

[0121] Weather disasters: 10 "weather disaster" events are statistically found, including 6 heavy rain events and 4 heavy snow events.

[0122] Calculate the total dynamic event occurrence frequency by adding the occurrence times of various events and dividing by 12 months. At the same time, the occurrence patterns of events in different time periods (such as weekdays, weekends, holidays) and different seasons are also analyzed. Obtain the actual execution data after each route adjustment from the route execution record database, including actual driving time, actual transportation cost, goods delivery time, and vehicle fuel consumption information, and compare them with the expected driving time and expected transportation cost data in the route adjustment plan. Taking a certain route adjustment as an example, after this route adjustment, a total of 20 transportation tasks have been executed. In terms of actual driving time, there are 5 times when the deviation rate of the actual driving time from the expected time > 15%. In terms of actual transportation cost, there are 4 times when the deviation rate of the actual transportation cost > 10%. According to the dynamic event frequency and the stability index of the route adjustment result , use the formula to calculate the threshold of the scheme survival years When = 5, = 0.8, = ≈ 0.27 years. According to the transportation characteristics and business requirements of different regions, the numbers in the formula are dynamically adjusted. For example, for regions with complex traffic conditions and frequent dynamic events, appropriately reduce it to make the scheme update more frequently; for relatively stable traffic regions, increase it to reduce unnecessary scheme updates. Obtain the optimization period of the current scheme from the scheme management database, that is, the duration from the timestamp when the scheme is formulated to the current moment, and compare this duration with the calculated threshold of the scheme survival years as follows:

[0123] If the current scheme optimization period ≥ threshold , the scenario elimination mechanism is triggered, and step 661 is entered;

[0124] If the current scenario optimization period < threshold , it is considered that the scenario is still within the valid usage period, and the scenario is continued to be used for path planning.

[0125] According to the path adjustment record database, which stores the detailed information of all path adjustments in the past period, including the reasons for adjustment, the adjusted path segments, and the verification results (whether the execution is successful and the evaluation of the execution effect). When retrieving, according to the detour compensation distance and priority classification rules set in the current dynamic correction parameters, the verified detour path segments that match the current situation are determined. Assume that the detour compensation distance in the current dynamic correction parameters is 800 meters. Search in the database for path segments with a detour compensation distance in the range of 700 - 900 meters and a successful verification result. If a matching path segment is found, according to the real-time position of the current vehicle (obtained through in-vehicle GPS with a precision up to the meter level), use the Dijkstra algorithm to calculate the shortest path from the current vehicle position to the starting point of the path segment; use the path planning function of GIS to generate a transition path from the current vehicle position to the starting point of the path segment without violating traffic rules (such as traffic restrictions and bans) and road restrictions (such as bridge load-bearing and tunnel height restrictions); connect the transition path with the verified path segment to form a complete alternative optimized path. During the connection process, the path will be smoothed to avoid unreasonable situations such as sharp turns and sudden lane changes, ensuring the safety and comfort of vehicle driving. If there are no available detour path segments, according to the detour compensation distance threshold set in the dynamic correction parameters, search for eligible alternative nodes in the current area. Taking the current vehicle position as the center, determine the search radius according to the detour compensation distance threshold (for example, if the detour compensation distance threshold is 800 meters, the search radius is set to 1000 meters, reserving a certain buffer range). Through the spatial analysis function of GIS, determine all positions that can be used as path nodes within the search radius, including road intersections, parking lots, gas stations, etc., to form a set of alternative nodes.

[0126] Dynamically calculate the survival life threshold of the solution by comprehensively analyzing the dynamic event frequency and the stability of the path adjustment result, making the update cycle judgment of the path planning solution more in line with the changes in the actual transportation environment, avoiding the waste of resources that may be caused by fixed-cycle updates, such as frequently updating the solution when the traffic condition is stable, increasing unnecessary calculation and management costs; at the same time, it also avoids the problem of solution lag, preventing the long-term use of invalid solutions in areas with frequent dynamic events, resulting in low transportation efficiency and increased costs. Flexibly adjust the threshold according to the actual situation to ensure that the latest and effective path planning solution can be used when facing dynamic events in different regions and different time periods, improving the timeliness and adaptability of path planning. Eliminate the old solution based on the matching degree of the path node and the regional center of gravity offset, and strictly screen the path solution from the perspective of spatial relevance, ensuring that the solutions retained in the global solution library are closely related to the current transportation area, eliminating the solutions with weak relevance to the area, reducing the occupation of system resources by invalid solutions, such as database storage space, computing resources, etc., and improving the quality and retrieval efficiency of the solution library. Prioritize the reuse of the verified detour path segments and achieve smooth connection, making full use of the successful experience. When facing dynamic events, it can quickly reuse the mature path segments and generate alternative optimized paths in the shortest time.

[0127] An embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when run on a computer, cause the computer to execute the method as described above. All implementation manners in the above method embodiments are applicable to this embodiment and can also achieve the same technical effects.

[0128] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A vehicle route planning method based on an adaptive optimization algorithm, characterized in that The method includes: Step 1, based on the current order requirements, vehicle resources, and preset constraint conditions, generate an initial path plan set that meets the time window, splitting times, vehicle type matching, and temperature zone requirements through a step-by-step progressive decision-making process; Step 2, according to the initial path plan set, start multiple independent optimization links, and each link dynamically adjusts the local path nodes through a greedy strategy to generate an optimized candidate plan set; Step 3, conduct real-time evaluation on the optimized candidate plan set, determine the final candidate plan based on the cost savings rate, demand satisfaction rate, and soft service quality indicators, and update the final candidate plan as the current benchmark plan; Step 4, according to the current benchmark plan, when a dynamic event occurs within the closed triangular area composed of the order distribution center detection point, vehicle dispatching center detection point, and transportation hub detection point, trigger local link optimization, and generate dynamic correction parameters according to the real-time position relationship between the triangle vertices; Step 5, according to the dynamic correction parameters, calculate the geometric centroid offset of the affected path segment, perform directional adjustment on the offset path segment, and then update the global plan library to generate a path adjustment result covering the dynamic area; Step 6, based on the updated global plan library, set a plan survival life threshold. When the plan optimization period ≥ threshold, combine the dynamic area topology association characteristics to eliminate the old plan, and generate an alternative optimization link based on the dynamic correction parameters.

2. The vehicle path planning method based on the adaptive optimization algorithm according to claim 1, wherein Based on the current order requirements, vehicle resources, and preset constraint conditions, generate an initial path plan set that meets the time window, splitting times, vehicle type matching, and temperature zone requirements through a step-by-step progressive decision-making process, including: According to the temperature zone attribute and time window limit of the order, perform multi-dimensional clustering on the current order requirements to generate multiple order subsets with the same temperature zone and a time window overlap degree ≥ preset threshold; Based on the order subsets, traverse the vehicle types in the vehicle resource library that meet the temperature zone adaptation conditions, determine a candidate vehicle set that can cover the total demand of the orders within the vehicle capacity subset, and perform splitting operations on the order subsets with > single loading capacity to ensure that the splitting times ≤ preset threshold, and generate a task set to be assigned including the split order units; Match the task set to be assigned with the candidate vehicle set, and generate an initial path segment that meets the vehicle loading capacity constraint according to the distance priority between the vehicle's real-time position and the order delivery point; Conduct time window conflict detection on the initial path segment. If there is a time window conflict, dynamically adjust the node order of the path segment based on the vehicle driving speed to generate a conflict-free set of feasible path segments; Combine and optimize the paths with the same temperature zone and geographical proximity in the set of feasible path segments to generate an initial path plan set including multi-vehicle type collaborative scheduling and multi-order combined distribution.

3. The vehicle path planning method based on the adaptive optimization algorithm according to claim 2, characterized in that, According to the initial path plan set, start multiple independent optimization links, and each link dynamically adjusts the local path nodes through a greedy strategy to generate an optimized candidate plan set, including: Extract the nodes with a path cost ratio > preset ratio from the initial path plan set as high-sensitivity adjustment objects to generate a candidate node set; For different optimization links, based on the distance deviation between the candidate node set and the current vehicle position, assign differential node adjustment priorities to each link to generate a locally exclusive adjustment range for the link; Within the locally adjusted range, perform a greedy operation on highly sensitive nodes. If there is sufficient margin in the time window of the path where the node is located, attempt to insert adjacent unassigned orders; if the time window is tight, exchange with adjacent path nodes to generate candidate sub-solutions within the link; Perform conflict verification on the candidate sub-solutions generated by each link, eliminate the solutions that conflict with the vehicle temperature zone adaptability, and merge the remaining sub-solutions into an optimized candidate solution set.

4. The vehicle path planning method based on the adaptive optimization algorithm according to claim 3, wherein The different optimization links include that the first link dynamically calculates the deviation amplitude difference based on the real-time distance deviation between the candidate node and the current vehicle position to generate a distance-sensitive adjustment priority sequence; The The second link quantifies the path interaction intensity based on the geographical overlap rate between the path where the candidate node is located and the adjacent path to generate an overlap-sensitive adjustment priority sequence.

5. The vehicle path planning method based on an adaptive optimization algorithm according to claim 4, wherein, Perform real-time evaluation on the optimized candidate solution set, determine the final candidate solution based on the cost savings rate, demand satisfaction rate, and soft service quality indicators, and update the final candidate solution to the current benchmark solution, including: For each solution in the optimized candidate solution set, calculate the difference between the total transportation cost and the historical benchmark solution, determine the solutions with a cost savings rate ≥ the preset threshold, and form a primary cost reduction solution set; Perform demand coverage verification on the primary cost reduction solution set, detect whether each solution covers all key order nodes. If there are uncovered urgent orders, mark them as demand missing solutions and eliminate them to generate a secondary demand compliance solution set; Perform service quality evaluation on the secondary demand compliance solution set. Based on the order delivery on-time rate and the path adjustment frequency, calculate the comprehensive service quality score, and generate a solution priority queue according to the score ranking; Determine the candidate solution with the final comprehensive score in the solution priority queue, synchronously activate the compensation mechanism for unmet demands, and re-inject the unassigned orders in the eliminated demand missing solutions into the initial path solution set to generate a pool of orders to be reallocated; Update the final candidate solution to the current benchmark solution and trigger the path replanning link for the pool of orders to be reallocated.

6. The vehicle path planning method based on the adaptive optimization algorithm according to claim 5, wherein Perform service quality evaluation on the secondary demand compliance solution set. Based on the order delivery on-time rate and the path adjustment frequency, calculate the comprehensive service quality score, and generate a solution priority queue according to the score ranking, including: Extract the matching deviation between the actual delivery time of the orders in each solution and the preset time window, and use the ratio of the total deviation duration to the total number of orders as the benchmark indicator to generate an on-time rate ranking sequence for each solution; Based on the on-time rate ranking sequence, count the number of node order adjustment times triggered by each solution during the path optimization process to generate an adjustment frequency sequence associated with the on-time rate; According to the on-time rate ranking sequence and the adjustment frequency sequence, determine the solutions with a total on-time rate deviation < the preset upper limit and an adjustment frequency < the preset limit value to form a preliminary preferred solution set; For the solutions within the set of initially preferred solutions, they are preferentially sorted in ascending order according to the total deviation of on-time rate. If the total deviations are the same, a secondary sorting is performed in ascending order according to the adjustment frequency, and the number of unserved orders within the path coverage area is extracted and sorted in ascending order of the number of unserved orders to generate the final solution priority queue.

7. The vehicle path planning method based on an adaptive optimization algorithm according to claim 6, wherein According to the current baseline solution, when a dynamic event occurs within the closed triangular area formed by the detection points of the order distribution center, the vehicle dispatching center, and the transportation hub, local link optimization is triggered, and dynamic correction parameters are generated based on the real-time position relationship between the vertices of the triangle, including: Real-time collect the position data of the three detection points within the triangular area. If the offset of the real-time coordinate of any detection point from the reference position > the preset tolerance threshold, it is determined as a dynamic disturbance event and the coordinate sequence is extracted; Based on the coordinate sequence, calculate the real-time change rate of the side lengths of each triangle and the dynamic offset angle of the vertex angles, and generate a dynamic event impact intensity index in combination with the current path node distribution density; According to the dynamic event impact intensity index and the path node distribution density of the area passed through in the current baseline solution, dynamically calculate the path redirection priority and the detour compensation distance, and generate dynamic correction parameters.

8. The vehicle path planning method based on an adaptive optimization algorithm according to claim 7, wherein According to the dynamic correction parameters, calculate the geometric centroid offset of the affected path segment, perform directional adjustment on the offset path segment, and then update the global solution library to generate a path adjustment result covering the dynamic area, including: Based on the path redirection priority in the dynamic correction parameters, extract the affected path segments within the triangular area and calculate the offset of the geometric centroid from the preset reference position; According to the offset direction and the detour distance compensation value, perform a directional adjustment operation on the affected path segment to obtain the adjusted path segment; Perform a constraint penetration verification on the adjusted path segment to detect whether the detour nodes > the vehicle temperature zone adaptation range and whether the split count limit is violated after the path is extended; Splice the verified adjusted path segments with the unaffected paths, update the global solution library, and generate a path adjustment result covering the dynamic area.

9. The vehicle path planning method based on an adaptive optimization algorithm according to claim 8, characterized in that Based on the updated global solution library, set a solution survival year threshold. When the solution optimization period ≥ the threshold, combine the dynamic area topological association characteristics to eliminate the old solutions, and generate an alternative optimization link based on the dynamic correction parameters, including: Dynamically calculate the solution survival year threshold according to the frequency of the dynamic event and the stability of the path adjustment result, and compare it with the current solution optimization period; If the optimization period ≥ the threshold, extract the matching degree of the path nodes and the offset of the current area centroid, and eliminate the solutions with a matching degree < the preset critical value; Based on the detour compensation distance and the priority classification rules set in the dynamic correction parameters, dynamically generate an alternative optimization path, including preferentially reusing the path segment for the existing verified detour path segment and realizing a smooth connection with the current vehicle real-time position; if there is no available detour segment, generate an alternative node set according to the detour compensation distance threshold for path reconstruction.

10. A computer-readable storage medium, characterized in that, The program stored in the computer-readable storage medium, when executed by the processor, implements the method described in any one of claims 1 to 9.

Images