An Air Delivery Mission Planning Method Based on Spatiotemporal Divide and Conquer

Abstract

The present invention discloses an air delivery mission planning method based on spatio-temporal division and governance, comprising the following steps: obtaining warehouse information, resource information, unmanned aerial vehicle information and delivery point information; dividing the entire delivery mission area into multiple mission areas according to the number of warehouses based on the warehouse information and the delivery point information, each mission area including one warehouse and several delivery points; respectively establishing respective single-area delivery mission planning models within each mission area; solving the single-area delivery mission planning models within each mission area to obtain single-area delivery mission planning schemes for each mission area; synthesizing the single-area delivery mission planning schemes of multiple mission areas to obtain the uncompleted delivery missions, and establishing a supplementary delivery mission planning model across mission areas; solving the supplementary delivery mission planning scheme, and further obtaining the entire delivery mission planning scheme.

Description

Technical Field

[0001] The present invention belongs to the technical field of mission planning, and particularly relates to an air delivery mission planning method based on spatio-temporal division and governance. Background Art

[0002] Delivery is one of the important ways in modern warfare and material support. The problems that need to be solved in delivery mission planning include the selection of delivery methods, the allocation of delivery capacity, the planning of delivery routes, the determination of delivery sequences, and the planning of delivery support. Due to the rapid development of unmanned aerial vehicle (UAV) technology, using UAVs for delivery is also one of the key issues to be studied in modern delivery.

[0003] During the delivery mission process, a material warehouse can be used to store one or several types of materials, and can provide takeoff and landing for UAVs, and also provide maintenance support functions for UAVs. Assume that each UAV can carry one type of material and fly to a delivery point to perform a delivery mission. There are several delivery points in the mission area, and the delivery points can be classified according to the quantity of required delivery material types. In each mission cycle, the materials required at each delivery point need to be delivered under different time constraints. How to plan the scheduling strategy of UAVs in the mission cycle to achieve optimization goals such as minimizing the number of UAVs or minimizing the flight time of UAVs on the premise of being able to complete the delivery mission is an urgent problem to be solved in UAV delivery mission planning. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an air delivery mission planning method based on spatio-temporal division and governance to solve the problems existing in the prior art, so as to realize the efficient planning and scheduling of UAVs in the delivery mission.

[0005] Based on the above purpose, an air delivery mission planning method based on spatio-temporal division and governance is provided, including the following steps:

[0006] Step 1, obtain warehouse information, resource information, UAV information, and delivery point information; in the same mission cycle, if a certain delivery point needs to be delivered multiple times, then divide this delivery point into multiple delivery points according to the number of deliveries. The positions of the multiple delivery points are the same, but the flight distance and flight time between them are set to infinity;

[0007] Step 2, according to the warehouse information and delivery point information, divide the entire delivery mission area into multiple mission areas according to the number of warehouses. Each mission area includes one warehouse and several delivery points;

[0008] Step 3, establish respective single-area delivery mission planning models in each mission area;

[0009] Step 4: Solve the single-region delivery task planning model for each task region to obtain the single-region delivery task planning scheme for each task region;

[0010] Step 5: Integrate the single-region delivery task planning schemes of multiple task regions to obtain the outstanding delivery tasks, and establish a cross-task-region supplementary delivery task planning model;

[0011] Step 6: Solve the supplementary delivery task planning scheme, and further obtain the overall delivery task planning scheme.

[0012] Furthermore, the objective function of the single-region delivery task planning model includes two objective functions. The first objective function is to minimize the number of drones, expressed as Min(N UAV ), where N UAV represents the number of drones. The second objective function is to minimize the flight time of the drones, expressed as represents the direct flight of drone v from delivery point i to delivery point j, and t ij represents the time of drone v from delivery point i to delivery point j;

[0013] The constraint conditions of the single-region delivery task planning model include:

[0014] Drone flight ability constraint: represents that the actual flight track time of drone v does not exceed the endurance time T v ;

[0015] Delivery times constraint: represents that for any delivery point j, the actual number of deliveries is greater than or equal to the minimum number of deliveries Tmin j required for this delivery point and less than the maximum number of deliveries Tmax j for this delivery point; represents that for any delivery point i, the actual number of deliveries is greater than or equal to the minimum number of deliveries Tmin i required for this delivery point and less than the maximum number of deliveries Tmax i for this delivery point;

[0016] Delivery time constraint: Et i ≤Gt i ≤Lt i represents that for the actual delivery time Gt i of any delivery point i, it must be between the earliest execution time Et i and the latest execution time Lt i for the delivery task of i;

[0017] Delivery time sequence constraint: It means that for any drone v, if there exists a trajectory of drone v from delivery point i to delivery point j, then the actual time Gt of drone v at delivery point i i plus the flight time between the two delivery points should be less than or equal to the actual time Gt of drone v arriving at delivery point j j ;

[0018] Value constraint of the actual delivery time: Gt i ≥0, which means that for any delivery point i, its actual delivery time is greater than or equal to 0;

[0019] Constraint of the drone flight path: It means that for any drone v, there will be no sub - cycle of the flight path between delivery points;

[0020] Value constraint of the decision variable: It means that the decision variable can only take values of 0 or 1. When the value is 1, it means there exists a flight path of drone v directly flying from delivery point i to delivery point j. When the value is 0, it means there is no flight path of drone v directly flying from delivery point i to delivery point j;

[0021] Take - off and landing constraints of the drone: It means that the take - off point of any drone is the warehouse, and it must also land at the warehouse after completing the delivery task;

[0022] Among them, UAV = {1, 2, … m} represents the set of drones, Loc = {0, 1, 2, … n} represents the set of warehouse and delivery point targets, where {0} represents the only warehouse in the area.

[0023] Specifically, the single - area delivery task planning model described in step 4 uses a two - stage iterative solution method, including the following steps:

[0024] Step 401, initialize the number of drones to 1;

[0025] Step 402, input the single - area delivery task planning model into the planning solver for solution;

[0026] Step 403, if the single - area delivery task planning model has no solution, increase the number of drones by 1 and continue step 402;

[0027] Step 404, if the single - area delivery task planning model has a solution, save the number of drones and the flight time of the drones;

[0028] Step 405, according to the decision variables solved by the single - area delivery task planning model, perform decoding to obtain the single - area delivery task planning scheme.

[0029] Further, the supplementary delivery task planning model described in step 5 includes two stages. The first stage is the cross-regional division of the unfinished delivery tasks, and the second stage is to establish a supplementary delivery task planning model for the unfinished delivery tasks according to the division results of the first stage;

[0030] The first stage includes the following steps:

[0031] Step 501, initialize the set of unfinished delivery tasks Task = {t1, t2,..., t M}, the set of delivery points for unfinished delivery tasks Loc t = {loc1, loc2,..., loc M}, the set of resources R = {r1, r2,..., r M} required for the delivery tasks, the set of warehouses H = {h1, h2,..., h N} and the corresponding set of resources they possess

[0032] Step 502, take out a task t i from the set of unfinished delivery tasks, the corresponding delivery point loc i , and the types of resources r i required for this task;

[0033] Step 503, traverse the set of warehouses to obtain the set of warehouses H i that can meet the types of resources r i required for task t ti ;

[0034] Step 504, find the warehouse h ti in the set of warehouses H i that is the closest to the delivery point loc ti , divide the task t i and the delivery point loc i into the same region as the warehouse h ti , and remove t i from Task = {t1, t2,..., t M};

[0035] Step 505, repeat steps 502 to 504 until all unfinished delivery tasks are completed with regional division;

[0036] The second stage includes the following steps:

[0037] Step 506, according to the cross-regional division results of the first stage, partition the unfinished delivery tasks and delivery points according to the set of warehouses used;

[0038] Step 507: Based on the cross-region zoning result in Step 506, establish a single-zone delivery task planning model similar to that in Step 3, which is the supplementary delivery task planning model.

[0039] The overall delivery task planning scheme described in Step 6 is obtained by integrating the single-zone delivery task planning scheme in Step 4 and the supplementary delivery task planning scheme in Step 6.

[0040] Specifically, the solution of the supplementary delivery task planning scheme described in Step 6 adopts a two-stage iterative solution method, including the following steps:

[0041] Step 401: Initialize the number of UAVs to 1.

[0042] Step 402: Input the supplementary delivery task planning model into the planning solver for solution.

[0043] Step 403: If the supplementary delivery task planning model has no solution, increase the number of UAVs by 1 and continue with Step 402.

[0044] Step 404: If the supplementary delivery task planning model has a solution, save the number of UAVs and the flight time of the UAVs.

[0045] Step 405: Decode according to the decision variables obtained by solving the supplementary delivery task planning model to obtain the supplementary delivery task planning scheme.

[0046] Preferably, the planning solver is selected from one or more of lingo, Matlab, iCplex, Gurobi, Ipsolve.

[0047] Preferably, the process of dividing the entire delivery task area into multiple task areas according to the number of warehouses in Step 2 is a classification process with the number of warehouses as the number of categories, dividing each warehouse into one category, and taking the distance between the delivery point and the warehouse location as an influencing factor.

[0048] Originally, the planning of the drone delivery mission is an extremely complex problem, including the number of warehouses, the number of drones, the types of supplies, the delivery intervals, the types of delivery points, etc. If the original problem is modeled and analyzed, the dimension of the decision variables is large and the modeling difficulty is high. Moreover, the method of the present invention adopts the idea of spatio-temporal divide and conquer. First, according to the warehouse information and the delivery point information, the entire mission area is divided into multiple sub-mission areas in space, and each warehouse guarantees a sub-mission area, thus turning the multi-warehouse planning problem into a single-warehouse planning problem. Second, according to the number of delivery times and the delivery time required by the delivery point, a delivery point that needs to be delivered multiple times is transformed into multiple delivery points at the same location to make the calculation simpler. Finally, considering that due to the limitations of the supplies in each warehouse and the requirements of the delivery points, some warehouses cannot meet the delivery tasks within the sub-region, so a cross-region planning model is established to make the solution results more accurate and comprehensive. The delivery mission planning method based on spatio-temporal divide and conquer not only greatly reduces the complexity in the modeling process, but also significantly reduces the complexity in the solution process, ensuring a relatively satisfactory planning result within an effective time. Brief Description of the Drawings

[0049] Figure 1 It is a schematic flow chart of the delivery unmanned planning method of the embodiment of the present invention. Detailed Embodiments

[0050] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0051] In this embodiment, the relevant information in the delivery task planning is clarified to make the model and the solution method more specific. There are 3 existing warehouses, which can realize the take-off and landing of unmanned aerial vehicles (UAVs). Each warehouse has a certain number of UAVs of the same type. Each UAV can carry a kind of material for delivery. It is known that Warehouse 1 has two kinds of materials, Warehouse 2 has two kinds of materials, and Warehouse 3 has three kinds of materials. There are 30 delivery points in the task area, including 6 type-I delivery points, 12 type-II delivery points and 12 type-III delivery points. The positions of the warehouses and the delivery points are both known. It is required to deliver at least three different materials to the type-I delivery points every day, at least two different materials to the type-II delivery points every day, and at least one different material to the type-III delivery points every day. The flight speed and the maximum flight time of the UAVs are known. And after flying for a certain period of time, the UAVs must land at the warehouse for maintenance, and the maintenance time is determined. It is required to determine at least how many UAVs need to be configured to ensure the completion of the delivery task, and to give the planning and scheduling scheme of all UAVs within one day, including the take-off warehouse and time, the carried material, the flight path, the delivery time, the landing warehouse and time, the maintenance time, etc. of each UAV.

[0052] Based on the above problems, a delivery task planning method based on spatio-temporal division is proposed. The decision variables of the problem are simplified by using the spatio-temporal division method. First, according to the warehouse information and the delivery point information, the entire task area is divided into multiple sub-task areas in space. It is assumed that each warehouse guarantees a sub-task area, so that the multi-warehouse planning problem is transformed into a single-warehouse planning problem. Second, according to the delivery times and delivery times required by the delivery points, a delivery point that needs to be delivered multiple times is transformed into multiple delivery points at the same position to make the calculation simpler. Finally, considering that due to the limitations of the materials in each warehouse and the requirements of the delivery points, some warehouses cannot meet the delivery tasks within the sub-region range, so a cross-region planning model is established to uniformly consider the uncompleted tasks for supplementary planning, and finally make the solution results more accurate and comprehensive. The main steps are as follows:

[0053] Step 1: Obtain warehouse information, resource information, drone information, and delivery point information. During the same mission cycle, if a delivery point requires multiple deliveries, then divide this delivery point into multiple delivery points according to the number of deliveries. The positions of the multiple delivery points are the same, but the flight distance and flight time between them are set to infinity. For example, when delivering to a Type II delivery point, at least two different materials need to be delivered every day. At this time, this delivery point can be transformed into multiple independent but identically - located delivery points, so as to reduce the dimension of the decision variables. For example, delivery point II01 can be regarded as two delivery points II01 - 1 and II01 - 2. Although the coordinates of these two delivery points are the same, the flight distance or flight time between the two points is set to infinity, so that in the planning process, the same drone will not fly from delivery point II01 - 1 to delivery point II01 - 2;

[0054] Step 2: According to the warehouse information and delivery point information, divide the entire delivery mission area into multiple task areas according to the number of warehouses. Each task area includes one warehouse and several delivery points;

[0055] Step 3: Establish respective single - area delivery task planning models in each task area;

[0056] Step 4: Solve the single - area delivery task planning models in each task area to obtain the single - area delivery task planning solutions in each task area;

[0057] Step 5: Integrate the single - area delivery task planning solutions of multiple task areas to obtain the outstanding delivery tasks, and establish a supplementary delivery task planning model across task areas;

[0058] Step 6: Solve the supplementary delivery task planning solution, and then obtain the entire delivery task planning solution.

[0059] In fact, the so - called single - area delivery task should actually meet the following conditions:

[0060] (1) Minimize the number of drones used;

[0061] (2) Minimize the flight distance or flight time of all drones within the task area;

[0062] (3) Complete each delivery task within its respective time window.

[0063] The objective function of the single - area delivery task planning model includes two objective functions. The first objective function is to minimize the number of drones, denoted as Min(N UAV ), where N UAV represents the number of drones. The second objective function is to minimize the flight time of the drones, denoted as It means that the UAV v flies directly from the delivery point i to the delivery point j, t ij It represents the time for the UAV to fly from the delivery point i to the delivery point j;

[0064] The constraint conditions of the single - zone delivery task planning model include:

[0065] UAV flight ability constraint: It means that the actual flight track time of the UAV v does not exceed the endurance time of the UAV v;

[0066] Constraint on the number of deliveries: It means that for any delivery point j, the actual number of deliveries is greater than or equal to the minimum number of deliveries Tmin required for this delivery point j , and less than the maximum number of deliveries Tmax for this delivery point j ; It means that for any delivery point i, the actual number of deliveries is greater than or equal to the minimum number of deliveries Tmin required for this delivery point i , and less than the maximum number of deliveries Tmax for this delivery point i ;

[0067] Constraint on delivery time: Et i ≤Gt i ≤Lt i , which means that for the actual delivery time Gt of any delivery point i i must be between the earliest execution time Et i and the latest execution time Lt i for the delivery task of i;

[0068] Constraint on the delivery time sequence: It means that for any UAV v, if there is a trajectory of the UAV v from the delivery point i to the delivery point j, then the actual time of the UAV v at the delivery point i plus the flight time between the two delivery points should be less than or equal to the actual time of the UAV v arriving at the delivery point j;

[0069] Constraint on the value of the actual delivery time: Gt i ≥0, which means that for any delivery point i, its actual delivery time is greater than or equal to 0;

[0070] Constraint on the UAV flight track: It means that for any UAV v, there will be no sub - cycle in the flight track between the delivery points;

[0071] Constraint on the value of the decision variable: It means that the decision variable can only take the value of 0 or 1. When the value is 1, it means that there is a track of the UAV v flying directly from the delivery point i to the delivery point j. When the value is 0, it means that there is no track of the UAV v flying directly from the delivery point i to the delivery point j;

[0072] Takeoff and landing constraints of the UAV: It means that for any UAV, the takeoff point is the warehouse, and after completing the delivery task, it must also land at the warehouse;

[0073] Among them, UAV = {1, 2, … m} represents the set of UAVs, Loc = {0, 1, 2, … n} represents the set of warehouse and delivery point targets, where {0} represents the only warehouse in the area.

[0074] The method for solving the single - area delivery task planning model described in step 4 adopts a two - stage iterative solution method, including the following steps:

[0075] Step 401, initialize the number of UAVs to 1;

[0076] Step 402, input the single - area delivery task planning model into the planning solver for solution;

[0077] Step 403, if the single - area delivery task planning model has no solution, increase the number of UAVs by 1, and continue with step 402;

[0078] Step 404, if the single - area delivery task planning model has a solution, save the number of UAVs and the flight time of the UAVs;

[0079] Step 405, according to the decision variables solved by the single - area delivery task planning model, perform decoding to obtain the single - area delivery task planning scheme.

[0080] The supplementary delivery task planning model described in step 5 includes two stages. The first stage is the cross - regional division of the uncompleted delivery tasks, and the second stage is to establish a supplementary delivery task planning model for the uncompleted delivery tasks according to the division results of the first stage;

[0081] It should be clear that the uncompleted delivery tasks are mainly because the warehouses in the task areas they belong to do not have corresponding resources to meet their delivery requirements. Therefore, when selecting a warehouse, the delivery point mainly considers the resource matching. The delivery point of the uncompleted delivery task and the warehouse it selects must be less than the endurance time of its UAV, and the delivery point of the uncompleted delivery task hopes to be as close as possible to the warehouse it selects.

[0082] The first stage includes the following steps:

[0083] Step 501, initialize the set of uncompleted delivery tasks Task = {t1, t2, …, t M}, and the set of delivery points of uncompleted delivery tasks Loc t = {loc1, loc2, …, loc M}, the resource set R required for the delivery task = {r1, r2, …, r M}, the warehouse set H = {h1, h2, …, h N} and the corresponding resource set it owns

[0084] Step 502: Take out a task t from the set of unfinished delivery tasks i , the corresponding delivery point loc i , and the type of resource r required for this task i ;

[0085] Step 503: Traverse the warehouse set to obtain the warehouse set H i that can meet the type of resource r i required for task t ti ;

[0086] Step 504: Find the warehouse h ti in the warehouse set H i that is the closest to the delivery point loc ti , divide the task t i and the delivery point loc i into the same area as the warehouse h ti , and remove t i from Task = {t1, t2, …, t M};

[0087] Step 505: Repeat Step 502 to Step 504 until all unfinished delivery tasks are completed for area division;

[0088] The second stage includes the following steps:

[0089] Step 506: According to the cross - area division result of the first stage, partition the unfinished delivery tasks and delivery points according to the used warehouse set;

[0090] Step 507: Based on the cross - area partition result in Step 506, establish the same single - area delivery task planning model as in Step 3, which is the supplementary delivery task planning model;

[0091] The entire delivery task planning scheme described in Step 6 is obtained by integrating the single - area delivery task planning scheme in Step 4 and the supplementary delivery task planning scheme in Step 6.

[0092] The solution of the supplementary delivery task planning scheme described in Step 6 adopts a two - stage iterative solution method, including the following steps:

[0093] Step 401: Initialize the number of drones to 1;

[0094] Step 402: Input the supplementary delivery task planning model into a planning solver for solution;

[0095] Step 403: If the calculation shows that the supplementary delivery task planning model has no solution, increase the number of UAVs by 1 and continue with Step 402;

[0096] Step 404: If the calculation shows that the supplementary delivery task planning model has a solution, save the number of UAVs and the flight time of the UAVs;

[0097] Step 405: Decode according to the decision variables obtained from the solution of the supplementary delivery task planning model to obtain a supplementary delivery task planning scheme.

[0098] The planning solver selected is one or more of lingo, Matlab, iCplex, Gurobi, and Ipsolve.

[0099] In the process of dividing the entire delivery task area into multiple task areas according to the number of warehouses in Step 2, taking the number of warehouses as the number of categories, each warehouse is divided into one category, and at the same time, the distance between the delivery point and the warehouse location is used as a classification factor for the classification process. In fact, there are various methods for multi-task area partitioning, as long as they are reasonable and practical.

[0100] As can be seen from the invention content and embodiments, a delivery task planning method based on spatio-temporal divide and conquer of the present invention simplifies the problem in the time and space dimensions, establishes a single-area delivery task planning model, and at the same time, to ensure a comprehensive solution to the problem, a supplementary task planning model across task areas is designed, so that the method of the present invention can solve the task planning problem in an effective time and the obtained task planning scheme can achieve satisfactory results.

Claims

1. A method for planning air delivery tasks based on spatio-temporal division and control, characterized in that, It includes the following steps: Step 1: Obtain warehouse information, resource information, UAV information, and delivery point information. During the same task cycle, if a delivery point needs to be delivered multiple times, the delivery point is divided into multiple delivery points according to the number of deliveries. The positions of the multiple delivery points are the same, but the flight distance and flight time between them are set to infinity. Step 2: According to the warehouse information and delivery point information, divide the entire delivery task area into multiple task areas according to the number of warehouses. Each task area includes one warehouse and several delivery points. Step 3: Establish respective single - area delivery task planning models in each task area. Step 4: Solve the single - area delivery task planning models in each task area to obtain the single - area delivery task planning solutions for each task area. Step 5: Integrate the single - area delivery task planning solutions of multiple task areas to obtain the uncompleted delivery tasks, and establish a supplementary delivery task planning model across task areas. Step 6: Solve the supplementary delivery task planning solution to obtain the entire delivery task planning solution. The objective function of the single - zone delivery mission planning model includes two objective functions. The first objective function is to minimize the number of drones, expressed as Min(N UAV ), where N UAV represents the number of drones. The second objective function is to minimize the flight time of the drones, expressed as represents the drone v flying directly from delivery point i to delivery point j, and t ij represents the time for the drone to fly from delivery point i to delivery point j; The constraint conditions of the single - area delivery task planning model include: Constraints on the flight ability of the drone: Indicates that the actual flight track time of drone v does not exceed the endurance time T of drone v v ; Constraint on the number of deliveries: It means that for any delivery point j, the actual number of deliveries is greater than or equal to the minimum number of deliveries Tmin required for this delivery point j , and less than the maximum number of deliveries Tmax of this delivery point j ; It means that for any delivery point i, the actual number of deliveries is greater than or equal to the minimum number of deliveries Tmin required for this delivery point i , and less than the maximum number of deliveries Tmax of this delivery point i ; Constraint on delivery time: Et i ≤ Gt i ≤ Lt i , indicating that the actual delivery time Gt for any delivery point i i must be between the earliest execution time Et i and the latest execution time Lt i for the delivery task to i; Constraints on the delivery time series: It means that for any drone v, if there is a trajectory of drone v from delivery point i to delivery point j, then the actual time of drone v at delivery point i plus the flight time between the two delivery points should be less than or equal to the actual time of drone v arriving at delivery point j; Value constraint of actual delivery time: Gt i ≥0, which means that for any delivery point i, its actual delivery time is greater than or equal to 0; Constraints on the UAV flight path: It means that for any UAV v, there will be no sub-loops in the flight path between the delivery points; Value constraints of decision variables: It means that the decision variable can only take the value of 0 or 1. When the value is 1, it indicates that there is a flight path for the UAV v directly flying from the delivery point i to the delivery point j. When the value is 0, it indicates that there is no flight path for the UAV v directly flying from the delivery point i to the delivery point j; Takeoff and landing constraints of the drone: It means that for any drone, the takeoff point is the warehouse, and it must also land at the warehouse after completing the delivery task; Among them, UAV = {1, 2, … m} represents the set of UAVs, and Loc = {0, 1, 2, … n} represents the set of warehouse and delivery point targets, where {0} represents the only warehouse in the area.

2. The method for planning air delivery tasks based on spatio-temporal division according to claim 1, characterized in that The method for solving the single - area delivery task planning model in Step 4 adopts a two - stage iterative solution method, including the following steps: Step 401: Initialize the number of UAVs to 1. Step 402: Input the single - area delivery task planning model into the planning solver for solution. Step 403: If the single - area delivery task planning model has no solution, increase the number of UAVs by 1 and continue with Step 402. Step 404: If the single - area delivery task planning model has a solution, save the number of UAVs and the flight time of the UAVs. Step 405: Decode according to the decision variables obtained by solving the single - area delivery task planning model to obtain the single - area delivery task planning solution.

3. The method for planning air delivery missions based on spatio-temporal division of claims 2, characterized in that, The supplementary delivery task planning model in Step 5 includes two stages. The first stage is the cross - regional division of the uncompleted delivery tasks, and the second stage is to establish a supplementary delivery task planning model for the uncompleted delivery tasks according to the division results of the first stage. The first stage includes the following steps: Step 501, initialize the set of undelivered tasks Task = {t1, t2, …, t M}, the set of delivery points for undelivered tasks Loc t = {loc1, loc2, …, loc M}, the set of resources R = {r1, r2, …, r M} required for the delivery tasks, the set of warehouses H = {h1, h2, …, h N} and the corresponding set of resources they own Step 502, take out a task t from the set of undelivered tasks i , the corresponding delivery point loc i , and the resource type r required for this task i ; Step 503, traverse the warehouse set to obtain the set of warehouses H that can meet the resource type r required by task t i i ti ;​​ Step 504, find the warehouse set H ti The warehouse h closest to the delivery point loc i in ti ti Divide the task t i and the delivery point loc i into the same area as the warehouse h ti and remove t i from Task = {t1, t2, …, t M}; Step 505: Repeat Step 502 to Step 504 until all uncompleted delivery tasks have completed regional division. The second stage includes the following steps: Step 506: According to the cross - regional division results of the first stage, partition the uncompleted delivery tasks and delivery points according to the set of warehouses used. Step 507: On the basis of the cross - regional partition results in Step 506, establish the same single - area delivery task planning model as in Step 3, which is the supplementary delivery task planning model. The entire delivery task planning solution described in Step 6 is obtained by integrating the single - area delivery task planning solution in Step 4 and the supplementary delivery task planning solution in Step 6.

4. The method for planning an air delivery mission based on spatio-temporal division according to claim 3, wherein The method for solving the supplementary delivery task planning solution in Step 6 adopts a two - stage iterative solution method, including the following steps: Step 401, initialize the number of drones to 1; Step 402, input the supplementary delivery task planning model into a planning solver for solution; Step 403, if it is calculated that the supplementary delivery task planning model has no solution, increase the number of drones by 1 and continue with Step 402; Step 404, if it is calculated that the supplementary delivery task planning model has a solution, save the number of drones and the flight time of the drones; Step 405, according to the decision variables obtained by solving the supplementary delivery task planning model, perform decoding to obtain a supplementary delivery task planning scheme.

5. The method for planning an air delivery mission based on spatio-temporal division according to claim 4, wherein The planning solver selected is one or more of lingo, Matlab, iCplex, Gurobi, Ipsolve.

6. The method for planning air delivery missions based on spatio-temporal division according to claim 5, wherein The process of dividing the entire delivery task area into multiple task areas according to the number of warehouses in Step 2 is a classification process with the number of warehouses as the number of categories, dividing each warehouse into one category, and at the same time taking the distance between the delivery point and the warehouse location as an influencing factor.

Images