A method and system for vehicle scheduling along specific routes in crop planting scenarios

Abstract

This invention discloses a vehicle scheduling method and system for specific passageways in crop planting scenarios, relating to the field of vehicle scheduling. The system includes an entrance allocation module, a global planning module, a monitoring and fine-tuning module, and a parameter optimization module. The entrance allocation module matches the optimal entrance based on the adaptation in-degree and calculates the departure interval. The global planning module uses an enhanced collaborative path planning algorithm to determine the optimal exit and path, verifies the safety range, and makes adjustments. The monitoring and fine-tuning module handles conflicts in real time and counts travel time. The parameter optimization module optimizes algorithm parameters based on time comparison to improve vehicle scheduling efficiency.

Description

Technical Field

[0001] This invention relates to the field of vehicle scheduling, and more specifically to a method and system for scheduling vehicles along specific routes in crop planting scenarios. Background Technology

[0002] In the industrialized cultivation of edible fungi, there are numerous process scheduling requirements, including intelligent harvesting, disinfection and inspection, and loading and unloading. Various vehicles, such as four-way vehicles and elevators, are used to complete these tasks. To meet the needs of mushroom growth and unified management, mushroom houses are designed with specific pathways, including U-shaped or H-shaped pathways. A key challenge is designing a vehicle scheduling system within these pathways to ensure safe collision avoidance when multiple vehicles are operating in parallel.

[0003] Existing vehicle dispatching systems mostly rely on static route planning, which makes it difficult to adapt to dynamic changes in the number of vehicles, departure intervals, and route topology in real time, thus failing to meet dynamic dispatching requirements. At the same time, the selection of vehicle entrances and exits does not effectively consider entrance and exit loads, making it difficult to ensure the safety and operational efficiency of multiple vehicles. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies by proposing a vehicle scheduling method and system for specific channel paths in crop planting scenarios. Through an entrance allocation module to match the optimal entrance, a global planning module to determine the optimal exit and path using an enhanced collaborative path planning algorithm, a monitoring and fine-tuning module to handle conflicts in real time and count the duration, and a parameter optimization module to iteratively optimize algorithm parameters, this invention achieves efficient scheduling of multiple vehicles and safe collision avoidance, adapting to the needs of crop planting scenarios and improving scheduling efficiency and reliability.

[0005] The technical solution to achieve the purpose of this invention is as follows:

[0006] A vehicle dispatching system for specific routes in crop planting scenarios, including an entrance allocation module, a global planning module, a monitoring and fine-tuning module, and a parameter optimization module;

[0007] The entrance allocation module obtains vehicle entry requests, matches the optimal entrance for the vehicle to be accessed based on the in-degree of each entrance, records the access time and calculates the departure interval, and generates and broadcasts the allocation results.

[0008] The global planning module generates a grid mapping table based on the allocation results, and uses an enhanced cooperative path planning algorithm to assign an agent to each vehicle while considering the priority of each vehicle. The optimal exit and optimal path for each vehicle are determined through simulation planning, the safe range of the driving vehicles is verified, and local adaptive adjustments are made according to priority and access order. The planning scheme is iteratively generated and broadcast.

[0009] The monitoring and fine-tuning module monitors each adjustment time step The system monitors all vehicles and updates the grid mapping table, determining in real time whether conflict groups exist. If a conflict group exists, it performs local adaptive adjustments based on the grid mapping table, replans the optimal path for the conflicting agents in the conflict group, and sends the corresponding information, including the travel time from leaving the vehicle. Placed in a temporary pool, a duration statistics package is generated and the temporary pool is cleared when the pool is full;

[0010] The parameter optimization module calculates the average driving time based on the time statistics package. and with expected duration The comparison determines whether to use an optimization algorithm to update the parameters of the enhanced cooperative path planning algorithm.

[0011] Furthermore, the entrance allocation module obtains vehicle entry requests, which include vehicle ID and priority. and vehicle location Priority Including high priority and low priority , obtain The system calculates the entrance load rate for each entrance by dividing the entrance queue number, the number of high-priority entrances, the maximum capacity of each entrance, the entrance location, the maximum number of vehicles departing per unit, and the departure interval. It then calculates the entrance priority rate by dividing the number of high-priority entrances by the number of queues. Based on the entrance location and vehicle location, it calculates the Euclidean distance between each entrance and the vehicle's location. These parameters are then used in a weighted adaptation in-degree model to calculate the adaptation in-degree between the vehicle to be connected and each entrance. The entrance with the highest adaptation in-degree is selected as the optimal entrance for the vehicle to be connected. The system then adaptively updates the entrance queue number and the number of high-priority entrances for the optimal entrance. .

[0012] Furthermore, the entrance allocation module determines the optimal entrance queue size. Calculate the entry priority of the optimal entry point. Add 1 and then multiply by the number of people in the queue at the entrance. The equivalent number of vehicles at the optimal entrance The equivalent number of vehicles at the optimal entrance Divide by the maximum number of trains departing per unit at the optimal entry point Then, round down to obtain the optimal departure interval for the entrance. And update.

[0013] Furthermore, the global planning module includes a global annotation unit, a collaborative decision-making unit, and a security verification unit;

[0014] The global annotation unit obtains the allocation results and simultaneously retrieves the vehicle ID and priority of all vehicles. and vehicle location Based on the location of each vehicle Mark the corresponding vehicle ID and priority in the corresponding grid on the virtual map. Establish a map mapping table, where the virtual map is a pre-built gridded map;

[0015] The collaborative decision-making unit assigns an intelligent agent to each vehicle and assigns it a vehicle ID and priority. Calculate the grid ID, combined with exit congestion rate and entrance priority rate. The algorithm determines the fit between each exit and each vehicle, identifies the optimal exit for each vehicle, and introduces grid load into the cost function of the Astar algorithm. and vehicle priority factor The initial optimal path for each vehicle is generated and integrated into an initial planning scheme.

[0016] The security verification unit is based on the initial planning scheme and adjusts the time steps. Perform step-by-step simulations on the virtual map. In each step of the simulation, determine whether a conflict group exists. If no conflict group exists, continue to the next step of the simulation. If a conflict group exists, consider the priorities of the two conflicting agents within the conflict group. and access time The system adaptively adjusts the optimal path of the conflicting agents to update the current planning scheme and re-executes the step-by-step simulation until all agents reach the corresponding optimal exit without conflict, thus generating a planning scheme.

[0017] Furthermore, collaborative decision-making units acquire... The exit queue number, high-priority exit number, and maximum capacity of each exit are calculated. The exit congestion rate for each exit is obtained by dividing the exit queue number by the high-priority exit number, and the exit priority rate is obtained by dividing the high-priority exit number by the exit queue number. These values ​​are then used in a weighted fit-out-degree model to calculate the fit-out-degree between each vehicle and each exit. The exit with the highest fit-out-degree is taken as the optimal exit for each vehicle, and the exit queue number of the optimal exit is adaptively updated. and export high priority numbers , This represents the total number of exports.

[0018] Furthermore, the collaborative decision-making unit improves the cost function of the existing Astar algorithm and defines the grid load. For the grid, two adjustment time steps in the future The probability of being occupied is used to define high priority. or low priority Vehicle priority factor The values ​​are either 1 or 0.5, and for each vehicle, the values ​​are determined by the vehicle's position. Export location relative to the optimal export These are the starting point and the end point, based on the cost improvement. The existing Astar algorithm is used for path planning to obtain the initial optimal path for each vehicle and integrate them into an initial planning scheme. The improvement cost is... Equal to path cost Adding grid load cost Reduce priority cost Path cost For vehicles to be located in a virtual map To the corresponding exit location Number of grids traversed, grid load cost The grid load equal to the grid load of the vehicle passing through. The sum multiplied by the fifth weight Priority cost Equal to vehicle priority factor Multiply by the sixth weight .

[0019] Furthermore, during the step-by-step simulation, the security verification unit determines whether there is a conflict group based on the security range. The security range of an agent includes the area enclosed by the grid where the agent is located and the surrounding 8 grids. If the security ranges of any two agents overlap in the same step of the simulation, they are considered to be a conflict group, and local adaptive adjustment is performed.

[0020] Furthermore, when the security verification unit detects a conflict group, it obtains the priority of the two conflicting agents. and access time If all members in the conflict group are low priority When there is a conflict between intelligent agents, the access time is selected. Lower priority In conflicting agents, if there is only one low-priority agent... When there are conflicting agents, the optimal path of all other agents besides the selected conflicting agent is fixed, and the path planning is re-performed according to the Astar algorithm. Each adjustment time step in the path planning process... In the middle, the remaining agents will adjust the time step at the current time. When the entire safe zone is marked as occupied, the Astar algorithm becomes unplannable, necessitating a re-determination of the optimal path when all conflict groups are of high priority. When a conflicting intelligent agent is selected, the access time is chosen. High priority later Conflicting agents, fix all other high-priority agents. The optimal path for the agents is determined, and the selected conflicting agents are re-planned based on the existing Astar algorithm. Each adjustment time step in the path planning process... Only the remaining high-priority cases are considered. The system determines the safe range of the agent, adjusts the current planning scheme after completion, calls the map mapping table generated by the global annotation unit, redetermines the optimal path, updates the planning scheme, and re-executes the step-by-step simulation.

[0021] A vehicle scheduling method for specific route channels in crop planting scenarios includes the following steps:

[0022] Obtain vehicle entry request, based on The in-degree of each entry point is used to match the optimal entry point for the vehicle to be connected, and the connection time is recorded. It also calculates the corresponding departure intervals and generates allocation results;

[0023] A grid mapping table is obtained based on the allocation results. An enhanced cooperative path planning algorithm is then used, taking into account the priority of each vehicle. The optimal exit and optimal path for each vehicle are determined through simulation planning, the safe range of the driving vehicles is verified, and local adaptive adjustments are made according to priority and access order. The planning scheme is iteratively generated and executed.

[0024] At each adjustment time step Monitor vehicle information of all vehicles, update the grid mapping table and determine whether there is a conflict group. When a conflict group exists, perform local adaptive adjustment, replan the optimal path of the conflicting agents and send it accordingly.

[0025] When the monitoring system detects a vehicle leaving, the travel time will be recorded. Placed in a temporary pool, the average travel time is calculated when the temporary pool is full. And clear the temporary pool, matching the preset expected duration. The parameters of the enhanced cooperative path planning algorithm are compared to determine whether to use an optimization algorithm to iteratively optimize the algorithm.

[0026] Compared with existing technologies, this invention constructs a scheduling system with four collaborative modules: entry allocation, global planning, monitoring and fine-tuning, and parameter optimization. The global planning module adopts an enhanced collaborative path planning algorithm, improves the Astar cost function by introducing grid load and vehicle priority, and handles conflict groups according to priority and access order based on safety verification. The monitoring and fine-tuning module updates the grid mapping table in real time and handles sudden conflicts. The parameter optimization module iteratively optimizes algorithm parameters based on travel time comparison to adapt to the scheduling needs of crop planting scenarios. Attached Figure Description

[0027] Figure 1A flowchart of a vehicle scheduling method for specific channel paths used in crop planting scenarios;

[0028] Figure 2 Flowchart for allocating the optimal entry point;

[0029] Figure 3 Flowchart for conflict group determination;

[0030] Figure 4 This is an architecture diagram of a vehicle dispatching system for specific routes used in crop planting scenarios. Detailed Implementation

[0031] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0032] Example 1

[0033] like Figure 1 As shown, a specific embodiment of the present invention discloses a vehicle scheduling system for specific channel paths in crop planting scenarios, including an entrance allocation module, a global planning module, a monitoring and fine-tuning module, and a parameter optimization module;

[0034] The entry allocation module receives vehicle entry requests and calculates them based on load balancing, distance optimization, and priority balancing principles. The in-degree of each entry point is used to match the optimal entry point for the vehicle to be connected, and the connection time is recorded. It also calculates the corresponding departure intervals, generates and broadcasts the allocation results;

[0035] The global planning module receives the allocation results, marks them on a pre-built virtual map to obtain a grid mapping table, and uses an enhanced cooperative path planning algorithm to assign an agent to each vehicle while considering the priority of each vehicle. The system determines the optimal exit and optimal path for each vehicle through simulation planning, verifies the safe range of the driving vehicles, and makes local adaptive adjustments based on priority and access order. Iteratively generates planning schemes and broadcasts them to control all vehicles.

[0036] The monitoring and fine-tuning module monitors each adjustment time step The system monitors vehicle information for all vehicles and synchronously updates the grid mapping table in the virtual map. It identifies the agents corresponding to all moving vehicles and determines if conflict groups exist. If a conflict group exists, it combines the current grid mapping table with the local adaptive adjustment of the global planning module to replan the optimal path for the conflicting agents in the group and sends it to the adjusted vehicle according to the vehicle ID. This prevents sudden conflicts caused by command delays or unexpected accidents. When a vehicle is detected leaving, its travel time is recorded. The records are placed in a temporary pool. When the number of records in the temporary pool reaches the pool limit, the duration statistics package is generated and the temporary pool is cleared.

[0037] The parameter optimization module calculates the average driving time based on the time statistics package. And with the preset expected duration In comparison, if the average driving time Less than or equal to the expected duration If the average driving time is [not specified], then no action will be taken. longer than expected duration The parameters of the enhanced cooperative path planning algorithm are used as optimization variables. Multiple different vehicle sets are randomly initialized in the virtual map. The enhanced cooperative path planning algorithm is used to simulate and deduce different vehicle sets step by step. The deduction time of the last step-by-step simulation of the planning scheme is statistically determined. Existing optimization algorithms are used to iteratively optimize the optimization variables to continuously reduce the average deduction time of all vehicle sets until the average deduction time is minimized. The enhanced cooperative path planning algorithm is re-determined and broadcast to update. The optimization algorithms include particle swarm optimization, gray wolf optimization algorithm and whale foraging algorithm, which will not be elaborated in detail in the technical solution of this application.

[0038] like Figure 2 As shown, further, the entry allocation module obtains the vehicle entry request initiated by the vehicle to be accessed. The vehicle entry request includes the vehicle ID and priority. and vehicle location Priority Including high priority and low priority Based on the grid mapping table Get the annotations in the occupied grid of each entrance, and obtain Entry information for the first entry point, the first... The entry information for each entrance includes the number of people queuing at the entrance. Entry high priority number Maximum number of passengers allowed at the entrance Entrance location Maximum number of trains per unit Departure interval , will the Queue length at each entrance Divide by the maximum capacity of the entrance Get the first Ingress load rate of each entry point According to Article The number of high-priority entries for each entry. Divide by the number of people queuing at the entrance Get the first Entry priority of each entry point Based on the first The entrance location of each entrance With vehicle position Calculate the Euclidean distance to the entrance A weighted adaptation in-degree model is constructed based on the principles of load balancing, distance optimization, and priority balancing. The vehicle to be connected and the first... Adaptability of each entry point Specifically as follows:

[0039] ,

[0040] in, , and Based on pre-set first weight, second weight, and baseline distance, the entry point with the maximum fit in-degree is selected as the optimal entry point for vehicles to be connected, and the entry queue number of the optimal entry point is set. Add 1, based on the priority of the vehicle to be connected. The number of high-priority entries for the optimal decision entry point Whether to increment by 1, depending on priority. Low priority When the optimal entry point has a high priority number of entries, then the entry point with the highest priority is... Unchanged, when priority High priority When the optimal entry point has a high priority number of entries, then the optimal entry point has a high priority number of entries. Add 1, and achieve adaptive matching between vehicles and the optimal entrance based on the weighted adaptation in-degree model. .

[0041] Furthermore, the entry allocation module adjusts the time step. Determine the optimal entry point queue size per unit time. Recalculate the entry priority of the optimal entry point. Add 1 and then multiply by the corresponding number of queues at the entrance. The equivalent number of vehicles at the optimal entrance The equivalent number of vehicles at the optimal entrance Divide by the maximum number of trains departing per unit at the optimal entry point Then, round down to obtain the optimal departure interval for the entrance. And updated simultaneously, including the maximum number of departures per unit. The optimal entry point represents the maximum number of departures per unit time, with the departure interval being the departure interval. The next vehicle in the queue representing the optimal entry point needs to wait after the previous vehicle in the queue departs. Adjustment time step Only then can the train depart.

[0042] Furthermore, the global planning module includes a global annotation unit, a collaborative decision-making unit, and a security verification unit;

[0043] The global annotation unit obtains the allocation results, executes the initial annotation in the enhanced cooperative path planning algorithm, and obtains the vehicle ID and priority of all vehicles. and vehicle location Based on the location of each vehicle Mark the corresponding vehicle ID and priority in the corresponding grid on the virtual map. Establish a map mapping table of grid ID-vehicle ID-priority, where all vehicles are included. The virtual map, which includes vehicles queuing at each entrance and vehicles already moving in the passageway, is pre-built. It involves constructing a scene model of the specific passageway path in modeling software, dividing the scene model into equal intervals based on the maximum vehicle length as the grid edge length, and then labeling the data. One entry and The grid occupied by each exit is used to generate a virtual map. Total exports;

[0044] The collaborative decision-making unit executes the simulation planning in the enhanced collaborative path planning algorithm, assigning an agent to each vehicle and assigning it a vehicle ID and the priority corresponding to that vehicle ID. Combined with grid ID, exit congestion rate and exit priority are used to calculate... The algorithm determines the optimal exit for each vehicle and the fit between each exit, introducing grid load into the cost function of the existing Astar algorithm. and vehicle priority factor Adaptive planning generates the initial optimal path for each vehicle, and the paths are integrated to obtain the initial planning scheme.

[0045] The security verification unit performs verification adjustments in the enhanced cooperative path planning algorithm, based on the initial planning scheme and according to the adjustment time steps. Perform step-by-step simulations on the virtual map. In each step of the simulation, identify the agents corresponding to all moving vehicles and determine if any conflict groups exist. If no conflict groups exist, continue simulating the next step for all agents based on the current planning scheme. If conflict groups exist, consider the priorities of the two conflicting agents within the group. and access time The system adaptively adjusts the optimal paths of the two conflicting agents to update the current planning scheme and re-executes the step-by-step simulation until all agents reach the corresponding optimal exit without conflict. It then generates a planning scheme and broadcasts it to all vehicles. When a new vehicle joins, it achieves optimal global replanning and controls all vehicles to execute it sequentially.

[0046] Furthermore, the collaborative decision-making unit is based on the grid mapping table. The occupancy grid of each exit is marked to obtain... Export information for each export, of which the first export... Export information for each export includes the number of export queues. Export high priority number Maximum export capacity , will the Number of exit queues for each exit Divide by the export priority number Get the first Exit congestion rate at each exit , will the Export priority number of exports Divide by the number of exit queues Get the first Export priority of each export A weighted adaptive exit degree model is constructed based on the congestion equilibrium principle and the priority equilibrium principle. Each vehicle is matched with the first... The adaptability of each export Specifically as follows:

[0047] ,

[0048] in, and The third and fourth weights are pre-set, and the exit with the maximum fit degree is taken as the optimal exit for each vehicle. The number of exits at the optimal exit is then set as the queue number. Add 1, based on vehicle priority. The number of high-priority exports for determining the optimal export. Whether to increment by 1, depending on priority. Low priority When the optimal export priority number is reached, the export priority number is determined. Unchanged, when priority High priority When the optimal export priority number is reached, the export priority number is determined. Add 1, and achieve adaptive matching between each vehicle and the optimal exit based on the weighted adaptive exit degree model. .

[0049] Furthermore, the collaborative decision-making unit improves the cost function of the existing Astar algorithm and defines the grid load. For the grid, two adjustment time steps in the future The probability of being occupied is used to define the vehicle priority factor. Based on the priority of each vehicle High priority or low priority The corresponding vehicle priority factor Set to 1 or 0.5 for each vehicle, based on vehicle position. Starting from the point of origin, the corresponding optimal exit position is determined. As the endpoint, based on improved costs The existing Astar algorithm is used for path planning to obtain the improvement cost for each vehicle. The initial optimal path is minimized, and an initial planning scheme is generated by integrating these paths. The execution process of the existing Astar algorithm is not the focus of this application's technical solution and will not be elaborated upon further. The improvement cost... Including the original path cost of the existing Astar algorithm and the cost of newly introduced grid load and priority cost The details are as follows:

[0050] ,

[0051] Among them, path cost For the vehicle from the vehicle position The exit position from the grid to the corresponding optimal exit Number of grids traversed by the current grid, grid load cost The grid load equal to the grid load of the vehicle passing through. The sum multiplied by the fifth weight Grid load cost The higher the priority cost, the greater the probability of a conflict occurring on the vehicle's travel path. Equal to vehicle priority factor Multiply by the sixth weight High priority The priority cost of the vehicle The lower value makes the existing Astar algorithm focus more on this aspect.

[0052] like Figure 3 As shown, further, in the step-by-step simulation, the security verification unit determines whether there is a conflict group based on the security range. The security range of an agent is defined as the area enclosed by eight grids around the grid where the agent is located. If the security ranges of any two agents overlap in the same step of the simulation, it is considered a conflict group, and the local adaptive adjustment of the conflict group is performed.

[0053] Furthermore, when the security verification unit detects a conflict group, it obtains the priority of the two conflicting agents. and access time If all members in the conflict group are low priority When there is conflict between agents, two lower priority agents are selected. Access time in conflicting intelligent agents Later conflicting agents, when there is only one low-priority agent. In the event of conflicting agents, a direct selection is made, fixing the selected lower-priority agent. The optimal path for all agents other than the conflicting agent is determined based on the existing Astar algorithm, prioritizing lower-priority agents. The conflicting agents backtrack to the starting point of the simulation in the map mapping table and replan the path. In the process of determining the optimal path in the existing Astar algorithm, each adjustment time step is determined. The remaining agents will adjust their time steps at the current time step. The entire safe zone is marked as occupied, and the remaining agents are placed in the previous adjustment time step. All safe zones are unmarked, and the marked grid cells cannot be selected by the existing Astar algorithm, thus redetermining low priorities without affecting other agents. The optimal path for conflicting agents, when all conflict groups are high priority. When a conflicting intelligent agent is selected, the access time is chosen. High priority later Conflicting agents, fix all other high-priority agents. The agent's optimal path, ignoring all other low-priority paths. The intelligent agent, based on the existing Astar algorithm, selects high-priority... The conflicting agent backtracks to the starting point of the simulation in the map mapping table and re-plans the path, adjusting the path at each time step in the existing Astar algorithm to determine the optimal path. Within this scope, only all other high-priority cases need to be considered. The intelligent agent's safe range is sufficient. After adjustment, the current planning scheme is updated and the map mapping table generated by the global annotation unit is called. The step-by-step simulation is re-executed based on the updated planning scheme. During the step-by-step simulation, the map mapping table remains unchanged.

[0054] Example 2

[0055] like Figure 4 As shown, a specific embodiment of the present invention discloses a vehicle scheduling method for specific passage routes in a crop planting scenario, comprising the following steps:

[0056] Vehicle entry requests are obtained, and calculations are performed based on load balancing, distance optimization, and priority balancing principles. The in-degree of each entry point is used to match the optimal entry point for the vehicle to be connected, and the connection time is recorded. It also calculates the corresponding departure intervals and generates allocation results;

[0057] Based on the allocation results, a grid mapping table is obtained by marking the virtual map. An enhanced cooperative path planning algorithm is then used to collaboratively consider the priority of each vehicle. The optimal exit and optimal path for each vehicle are determined through simulation planning, the safe range of the driving vehicles is verified, and local adaptive adjustments are made according to priority and access order. The planning scheme is iteratively generated and executed.

[0058] At each adjustment time step Monitor vehicle information of all vehicles and update the grid mapping table. Determine if there is a conflict group. When a conflict group exists, combine the current grid mapping table and perform local adaptive adjustment to replan the optimal path of the conflicting agents in the conflict group and send it to the adjusted vehicle according to the vehicle ID.

[0059] When the monitoring system detects a vehicle leaving, the vehicle's travel time is recorded. The records are placed in a temporary pool. When the number of records in the temporary pool reaches the pool's upper limit, the average travel time is calculated. And clear the temporary pool, matching the preset expected duration. The parameters of the enhanced cooperative path planning algorithm are compared to determine whether to use an optimization algorithm to iteratively optimize the algorithm.

[0060] This invention discloses a vehicle scheduling method and system for specific channel paths in crop planting scenarios. The entry allocation module acquires vehicle entry requests, matches the optimal entry point for each vehicle based on its in-degree of fit, records the entry time, and calculates the departure interval, achieving entry load balancing and efficient access while avoiding congestion at a single entry point. The global planning module annotates the grid mapping table based on the allocation results, uses an enhanced collaborative path planning algorithm to determine the optimal exit and path for vehicles, verifies the safety range, and adjusts according to priority and access order, ensuring optimal paths and driving safety while reducing vehicle path conflicts. The monitoring and fine-tuning module monitors vehicles, updates the grid mapping table, handles conflict groups, and counts the travel time of departing vehicles at each adjustment time step, enabling real-time response to sudden conflicts, ensuring stable scheduling, and providing data support for parameter optimization. The parameter optimization module calculates the average travel time based on the time statistics package and compares it with the expected time, optimizing the parameters of the enhanced collaborative path planning algorithm to continuously improve scheduling performance and better adapt to the scheduling needs of crop planting scenarios.

[0061] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A specific lane path vehicle dispatch system for crop planting scenarios, characterized in that, The system comprises an entrance distribution module, a global planning module, a monitoring fine-tuning module and a parameter optimization module. The entrance distribution module obtains a vehicle entry request, matches an optimal entrance for a vehicle to be accessed according to the adaptive access degree of each entrance, records the access time and calculates the departure interval, generates and broadcasts the distribution result, wherein the entrance load rate of each entrance is obtained by dividing the entrance queue number of each entrance by the maximum carrying capacity, the entrance priority rate of each entrance is obtained by dividing the entrance high priority number of each entrance by the entrance queue number, the entrance Euclidean distance is calculated based on the entrance position of each entrance and the vehicle position, and the adaptive access degree of the vehicle to be accessed and each entrance is calculated by substituting into the weighted adaptive access degree model; The global planning module obtains the grid mapping table according to the distribution result, assigns an intelligent agent to each vehicle by considering the priority of the vehicle using an enhanced cooperative path planning algorithm, determines the optimal exit and optimal path of each vehicle through simulation planning, checks the safety range of the running vehicle and performs local adaptive adjustment according to the priority and access order, iteratively generates a planning scheme and broadcasts it; The monitoring fine-tuning module monitors all vehicles and updates the grid mapping table at each adjustment time step, determines in real time whether there is a conflict group, and if so, performs local adaptive adjustment combined with the grid mapping table, re-plans the optimal path of the intelligent agent in the conflict group and sends it correspondingly, and places the running time of the leaving vehicle in a temporary pool, and when the temporary pool is full, generates a time length statistical package and empties the temporary pool; The parameter optimization module calculates the average running time based on the time length statistical package and compares it with the expected time length to decide whether to use an optimization algorithm to optimize and update the parameters of the enhanced cooperative path planning algorithm; The global planning module comprises a global labeling unit, a cooperative decision unit and a safety checking unit; the global labeling unit obtains the distribution result and the vehicle ID, priority and vehicle position of each vehicle, and labels the information in the grid of the pre-constructed virtual map to establish a map mapping table; The cooperative decision unit assigns an intelligent agent to each vehicle and assigns a vehicle ID, priority and grid ID to the vehicle, calculates the adaptive out-degree of each vehicle and determines the optimal exit, introduces the grid load and vehicle priority factor in the cost function of the Astar algorithm, and plans to generate an initial planning scheme, wherein the exit congestion rate of each exit is obtained by dividing the exit queue number of each exit by the exit high priority number, the exit priority rate of each exit is obtained by dividing the exit high priority number of each exit by the exit queue number, and the adaptive out-degree of each vehicle and each exit is calculated by substituting into the weighted adaptive out-degree model; The safety checking unit gradually simulates and deduces based on the initial planning scheme, determines the conflict group in each simulation and executes local adaptive adjustment to generate a planning scheme.

2. The specific lane path vehicle dispatch system for crop planting scenarios of claim 1, wherein, The cooperative decision unit takes the exit with the maximum adaptive out-degree as the optimal exit of the vehicle, and updates the queue number and high priority number of the optimal exit.

3. The specific lane path vehicle dispatch system for crop planting scenarios of claim 1, wherein, The cooperative decision unit defines the grid load as the occupancy probability of the grid in the next 2 adjustment time steps, and sets the priority factor of vehicles with different priorities respectively; the vehicle position and the optimal exit are taken as the starting point and the end point respectively, and the path is planned based on the improved cost by using the Astar algorithm; the improved cost is equal to the sum of the path cost and the grid load cost minus the priority cost, the path cost is equal to the number of grids passed by the path, the grid load cost is positively correlated with the sum of the grid load of the grids passed by the path, and the priority cost is positively correlated with the priority factor.

4. The specific lane path vehicle dispatch system for crop planting scenarios of claim 1, wherein, The safety checking unit sets the area surrounded by the grid where the agent is located and the surrounding 8 grids as the safety range in the step-by-step simulation and deduction; in the same step deduction, when the safety ranges of any two agents overlap, they are determined as a conflict group, and local adaptive adjustment is performed.

5. The specific lane path vehicle dispatch system for crop planting scenarios of claim 4, wherein, When the safety checking unit processes the conflict group, the priority and access time of the conflict agent are obtained; when both are low-priority, the one with the later access time is selected, and the paths of the remaining agents are fixed for re-planning; when both are high-priority, the one with the later access time is selected, the paths of the remaining high-priority agents are fixed for re-planning, the scheme is updated, and the deduction is performed.

6. The specific lane path vehicle dispatch system for crop planting scenarios of claim 5, wherein, When the safety checking unit re-plans the path of the low-priority conflict agent, the agent to be planned is backtracked to the starting point of the simulation and deduction in the map mapping table; in each adjustment time step, the safety range of the remaining agents in the current adjustment time step is marked as occupied, and the safety range marking in the previous adjustment time step is cancelled, and the new path is planned by using the Astar algorithm, and the grid marked as occupied cannot be selected.

7. The specific lane path vehicle dispatch system for crop planting scenarios of claim 5, wherein, When the path of the high-priority conflict agent is re-planned, the agent to be planned is backtracked to the starting point of the simulation and deduction in the map mapping table; all low-priority agents are ignored; in each adjustment time step, only the safety range of the remaining high-priority agents is marked as occupied, and the new path is planned by using the Astar algorithm, and the planning scheme is updated and re-deduced after the planning scheme is updated.

8. The specific lane path vehicle dispatch system for crop planting scenarios of claim 1, wherein, The entrance allocation module determines the queuing number of the optimal entrance, calculates the entrance priority rate, and obtains the equivalent vehicle number by multiplying the queuing number by 1 after adding 1; the equivalent vehicle number is divided by the unit maximum dispatching capacity of the optimal entrance and rounded to obtain the dispatching interval and update it.

9. A specific lane path vehicle dispatching method for crop planting scenarios, characterized in that, The method comprises the following steps: obtaining a vehicle entry request, matching the optimal entrance for the vehicle to be accessed according to the adaptive entry degree of each entrance, recording the access time and calculating the corresponding dispatching interval, and generating an allocation result, wherein the entrance load rate of each entrance is obtained by dividing the entrance queuing number of each entrance by the maximum carrying capacity, the entrance priority rate of each entrance is obtained by dividing the number of high-priority entrances of each entrance by the entrance queuing number, the entrance Euclidean distance is calculated based on the entrance position of each entrance and the vehicle position, and the adaptive entry degree of the vehicle to be accessed and each entrance is calculated by substituting the weighted adaptive entry degree model; the grid mapping table is marked according to the allocation result, each vehicle is allocated an agent by using the enhanced cooperative path planning algorithm and considering the priority, the optimal exit and the optimal path of each vehicle are determined by simulation and planning, the safety range of the running vehicle is checked, and local adaptive adjustment is performed according to the priority and the access order, and a planning scheme is iteratively generated and executed; Vehicle information of all vehicles is monitored at each adjustment time step, the grid mapping table is updated, and it is determined whether there is a conflict group, when there is a conflict group, local adaptive adjustment is performed, the optimal path of the conflict agent is replanned, and corresponding sending is performed; When it is monitored that a vehicle leaves, the driving time length is placed in a temporary pool, when the temporary pool is full, the average driving time length is calculated and the temporary pool is emptied, and whether to adopt an optimization algorithm to iteratively optimize the parameters of the enhanced cooperative path planning algorithm is decided by comparing with a preset expected time length; The generating the planning scheme comprises: obtaining the allocation result and the vehicle ID, priority and vehicle position of each vehicle, corresponding marking information in a grid of a pre-constructed virtual map, and establishing a map mapping table; allocating an agent to each vehicle and giving the vehicle ID, priority and grid ID, calculating the adaptive out-degree of each vehicle and determining the optimal exit, introducing the grid load and vehicle priority factor in the cost function of the Astar algorithm, planning to generate an initial planning scheme, gradually simulating and deducing based on the initial planning scheme, determining the conflict group and performing local adaptive adjustment in each step of the simulation and deduction, and generating the planning scheme, wherein the number of high-priority exits of each exit is divided by the number of exit queues to obtain the exit congestion rate of each exit, the number of high-priority exits of each exit is divided by the number of exit queues to obtain the exit priority rate of each exit, and the adaptive out-degree of each vehicle and each exit is calculated by substituting into the weighted adaptive out-degree model.

Images