Distributed task reconstruction method based on historical information backtracking

Abstract

The invention provides a distributed task reconstruction method based on historical information backtracking, and relates to the technical field of distributed decision making. The method comprises the following steps: establishing a distributed multi-unmanned aerial vehicle cooperative task allocation model according to constraint conditions of an information interaction architecture; generating a preliminary decision-making scheme, wherein the preliminary decision-making scheme comprises two links of task construction and task negotiation; in the task negotiation link, designing an information backtracking mechanism for the failed unmanned aerial vehicle; constructing a task priority chain type storage structure according to an information backtracking result; and carrying out quantitative analysis on the distribution model containing the random node failure, determining the failure position of the unmanned aerial vehicle, and executing task reconstruction. The invention provides a method for realizing dynamic redistribution after task distribution aiming at the problem of complex optimization of task distribution of multiple unmanned aerial vehicles. And distribution and reconstruction of tasks are jointly determined through communication and cooperation between the unmanned aerial vehicles, so that the flexibility and robustness are further improved.

Description

Technical Field

[0001] The present invention relates to the field of distributed decision-making technology, and in particular to a distributed task reconstruction method based on historical information backtracking. Background Art

[0002] The concept of drones can be traced back to the 19th century, when the prototype of unmanned aerial vehicles was first introduced. In the late 20th century, with the rapid development of electronic and aviation technologies, drones have gradually become an important tool in many fields.

[0003] Early UAV systems were mostly single-drone systems. With the advancement of technology, multi-drone systems have gradually been developed and applied. Since the 1990s, systems that use multiple drones to work together have appeared in many fields.

[0004] With the application of multi-UAV systems in various fields, multi-UAV task allocation and reconstruction technology has become more important. Traditional centralized task allocation and reconstruction methods have problems of single point failure and low efficiency. With the expansion of the scale and application scope of intelligent agent systems, the requirements for scalability and security will become more urgent. Summary of the invention

[0005] The present invention provides a distributed task reconstruction method based on historical information backtracking, which is used to solve the defects of single point failure and low efficiency in the prior art.

[0006] The present invention provides a distributed task reconstruction method based on historical information backtracking, which is used to realize dynamic reallocation after completing task allocation in a multi-UAV system, comprising: S1. According to the constraints of the information interaction architecture, a distributed multi-UAV collaborative task allocation model is established.

[0007] S2. Generate a preliminary decision-making plan, including task construction and task negotiation.

[0008] S3. During the task negotiation phase, an information backtracking mechanism is designed for failed drones.

[0009] S4. Build a task priority chain storage structure based on the information backtracking results.

[0010] S5. Quantitatively analyze the allocation model with random node failures, determine the failure locations of drones, and perform mission reconstruction.

[0011] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S1, it is assumed that there are n drones, and the upper limit of the target task that each drone can complete is , while given target points, the formula of the distributed multi-UAV collaborative task allocation model is expressed as:

[0012] The constraints are set as:

[0013]

[0014]

[0015] In the formula, Indicates drone The corresponding relationship mark with the task point f, when task f is assigned to the drone When the corresponding relationship is marked ,on the contrary ; Indicates a dimensional vector; Indicates all the tasks occupied by drone i The execution order of represents the profit calculation function; Represents the raw revenue value.

[0016] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S2, the task construction link constructs the task package of each drone by adopting a greedy strategy, and the construction process includes: Enter new task j and find the current benefit value after adding new task j. The formula is:

[0017] In the formula, represents the existing mission path of drone i, It means that the UAV i is in the mission path The total revenue value at Indicates that in the existing task path Add task j to the nth position of Indicates at location Find the position n with the maximum profit, and the current profit value is .

[0018] Compare the current benefit value with the original benefit value, and determine whether task j should be added to the task package based on the comparison result. , then use Alternative , indicating that task j can be added to the task package. , it means that task j cannot be added to the task package.

[0019] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S2, the task negotiation link resolves the task construction conflicts between drones by constructing a drone task negotiation algorithm based on a market auction mechanism. The task negotiation link includes agreed marking symbols and task negotiation.

[0020] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, the agreed marking symbols include: Winner bid values ​​list vector ,vector Represents the bid value of each task in the storage unit of drone i. The bid value represents the task benefit value that drone i needs to achieve to win the task.

[0021] Winners List Vector ,vector Indicates the ownership of each task in the storage unit of drone i.

[0022] Timestamp vector ,vector Used to mark the information update time of drone i.

[0023] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, the process of performing task negotiation includes: when , , to update the information, that is, the information tag about task j in drone i is replaced by the information tag of the corresponding task j in drone k.

[0024] when , 0, reset the information, that is, all the information about task j in drone i is reset to the initial empty state.

[0025] when , , information no operation is performed, that is, the information about task j in UAV i is retained.

[0026] In the formula, represents the bid value of the jth task in the storage information of drone i. Indicates the serial number value of the drone to which the jth task belongs in the storage information of drone i.

[0027] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S3, the process of designing an information backtracking mechanism for a failed UAV includes: S31. Given a task T.

[0028] S32. Analyze the bidding process of multiple iterations between each drone, and let the drones following the greedy optimization principle bid for task T independently for the first iteration, compare the corresponding benefit values ​​of the drones following the greedy optimization principle to evaluate the bids, and determine the winner of the first round.

[0029] S33. Set a drone that does not follow the greedy optimization principle to join the auction, and compare the profit value of the current drone with the winner of the previous round, perform the current iteration, and determine the winner of the current iteration.

[0030] S34. Repeat step S33 until all drones complete the bidding and determine the execution right of task T.

[0031] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S4, the process of constructing a task priority chain storage structure includes: Build priority-based chain storage logic.

[0032] For each point in the chain node, record the node of the previous priority.

[0033] Get the execution priority sorting chain of task T.

[0034] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S5, the process of quantitatively analyzing the allocation model containing random node failures includes: Set the initial network of multi-UAV coordination task as , indicating a A statically connected initial network of nodes.

[0035] Get the set of sub-networks generated when one of the nodes, UAV R, fails:

[0036] If node R is directly connected to two or more other nodes, the number of subnetworks generated , task reconstruction is required.

[0037] If there is an adjacent node in node R, the number of subnetworks generated , no task reconstruction is required.

[0038] According to a distributed task reconstruction method based on historical information backtracking provided by the present invention, in step S5, the task reconstruction process includes: Get the location of the failed node.

[0039] Traverse all drones and obtain the next-level drone of the failed node according to the priority chain storage structure.

[0040] Reassign tasks to the next level of drones and insert tasks into the execution sequence.

[0041] The present invention provides a distributed task reconstruction method based on historical information backtracking. By adopting interactive threads between drones, the method can detect the failure of drone nodes in a timely and accurate manner, and has real-time and high accuracy. It can respond quickly to drone failures, thereby reducing the risk of interruption of task execution and improving reliability and robustness. The advantage of real-time is that it can respond immediately after a drone failure or malfunction occurs, without the need for manual intervention, thereby saving time and labor costs and being able to operate more efficiently.

[0042] By adopting advanced algorithms and technologies, the best task allocation plan can be automatically determined based on factors such as task requirements, drone performance and environmental conditions. Task allocation can be completed quickly and accurately without manual intervention, improving efficiency and responsiveness.

[0043] The system robustness is effectively improved through the automatic task reallocation process. When a drone fails, the task allocation can be adjusted in time to avoid task interruption or delay, thereby ensuring that it can continue to operate in the face of emergencies, which is crucial to ensuring the continuity and reliability of task execution. Through timely task reallocation, it can quickly adapt to different environments and situation changes, reducing the risk of being affected by external interference or internal failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 It is a flowchart of a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention; Figure 2 It is a flow chart of generating a preliminary decision-making solution in a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention; Figure 3 This is a schematic diagram of node failure in a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention; Figure 4 It is a schematic diagram of a UAV negotiation process in a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention; Figure 5 It is a flow chart of distributed collaborative search in a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention; Figure 6 It is a schematic diagram of search trajectory reconstruction in a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention; Figure 7 It is a flowchart of a synchronization mechanism in a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] Combine the following Figure 1-Figure 7 A distributed task reconstruction method based on historical information backtracking of the present invention is described.

[0048] Figure 1 It is a schematic diagram of the steps of a distributed task reconstruction method based on historical information backtracking provided by an embodiment of the present invention.

[0049] like Figure 1 As shown, an embodiment of the present invention provides a distributed task reconstruction method based on historical information backtracking, including: S1. According to the constraints of the information interaction architecture, a distributed multi-UAV collaborative task allocation model is established.

[0050] In the collaborative decision-making of centralized multi-UAVs, the centralized terminal can obtain global information and solve and output unified decision-making control instructions under a given objective constraint function. However, the distributed multi-UAV system is constrained by the information interaction architecture and needs to solve the control instructions without task allocation conflicts through negotiation iteration under given objective constraints.

[0051] Suppose there are n drones, and the upper limit of the target mission that each drone can complete is , while given target points, at this time, the formula of the distributed multi-UAV collaborative task allocation model is expressed as:

[0052] The constraint formula is expressed as:

[0053]

[0054]

[0055] In the formula, Indicates drone The corresponding relationship mark with the task point f, when task f is assigned to the drone When the corresponding relationship is marked ,on the contrary ; is a dimensional vector, indicating that drone i is The marking status of each task; Indicates all the tasks occupied by drone i The execution order of represents the profit calculation function; Represents the raw revenue value.

[0056] The benefit value will be affected by all the tasks occupied by drone i and their execution order, and its calculation and solution meets the principle of diminishing marginal returns.

[0057] In the distributed task negotiation solution, the number of solution iterations will be affected by the minimum value of the number of drones, the total number of tasks, and the upper limit of the number of tasks that drones can execute.

[0058] Set the iteration to meet the conditions, the formula is expressed as:

[0059] In the formula, Indicates the minimum value of the number of drones, the total number of missions, and the upper limit of the number of missions that drones can execute.

[0060] S2. Generate a preliminary decision plan, including task construction and task negotiation.

[0061] S211, the task construction phase uses a greedy strategy to build the task package of each drone. The construction process includes: Enter new task j and find the current benefit value after adding new task j. The formula is:

[0062] In the formula, represents the existing mission path of drone i, It means that the UAV i is in the mission path The total revenue value at Indicates that in the existing task path Add task j to the nth position of Indicates at location Find the position n with the maximum profit, and the current profit value is .

[0063] Compare the current benefit value with the original benefit value, and based on the comparison result, determine whether task j should be added to the task package.

[0064] like , then use Alternative , indicating that task j can be added to the task package; if , it means that task j cannot be added to the task package.

[0065] S212. Construct a UAV task negotiation algorithm based on a market auction mechanism to resolve task construction conflicts between UAVs.

[0066] Since UAVs are connected through a distributed topology network, there are timing differences in information transmission between UAVs. Therefore, information synchronization conditions are specified in the UAV task negotiation process. At the same time, some basic marking symbols are agreed upon in the UAV task negotiation link to simplify the comparison conditions and operation descriptions of different negotiation situations.

[0067] Marking symbols include: Winner bid value list vector :vector is a The vector contains the bid value, which indicates the bid value of each task in the storage unit of drone i. The bid value is the calculated value of the task benefit that drone i needs to achieve to win the task. represents the bid value of the jth task in the storage information of drone i.

[0068] Winners List Vector :vector is a The vector contains the serial number of the drone that won the corresponding task, indicating which drone each task belongs to in the storage unit of drone i. The marking status will change with the bid value vector changes with the changes of. Indicates the serial number value of the drone to which the jth task belongs in the storage information of drone i.

[0069] Timestamp vector : The timestamp is used to mark the information update time of drone i, that is, the timestamp after each drone i completes the interaction with other drones Update once.

[0070] After the designated drone completes information exchange with its topologically adjacent drones, the drone can perform three actions based on the current information: information update, information reset, and information no operation. The specific actions are as follows: Information update: , , that is, the information tag about task j in drone i is replaced by the information tag of the corresponding task j in drone k.

[0071] Information reset: , , that is, all the information about task j in drone i is reset to the initial empty state.

[0072] Information not operated: , , that is, the information about task j in UAV i is retained.

[0073] like Figure 2 As shown in Figure 1, the relationship between task package construction and conflict resolution. After providing task information, multi-UAV task decision-making will go through the two processes of task construction and task negotiation. When the output result of a specific iteration meets the termination condition, the task allocation is completed. At this point, the allocation results have reached consistency.

[0074] S3. During the task negotiation phase, an information backtracking mechanism is designed for failed drones.

[0075] In order to prevent task conflicts caused by communication interruption between drones and inability to exchange information after node failure, the bidding process needs to be traceable, and the remaining tasks need to be completed and reconstructed without conflict while ensuring the optimization of benefits.

[0076] like Figure 4 As shown in Figure 1, given a task T, the negotiation process between drones is analyzed in three iterations to illustrate the workflow of the information backtracking mechanism. In the first iteration, drones a, b, and c follow the greedy optimization principle and bid for task T independently. The bids are evaluated by comparing the corresponding benefit values, and the result is Therefore, drone b becomes the winner of this round and enters the next iteration, temporarily retaining the ownership of task T. In this round, the winner of task T will change, which can be determined by comparing the winner lists of the previous round and the current iteration. To determine:

[0077] If the above formula is true, the drone will record the priority as shown below:

[0078] In the second iteration, drone d joins the auction and submits a bid that exceeds that of drone b, successfully winning task T.

[0079] In the third iteration, drone d competed with drone e, but its bid was still superior and it finally won the right to execute task T.

[0080] To sum up, through the negotiation construction process of drones a, b, c, d and e for task T, we can analyze and conclude that the negotiation process of drones for tasks can be generally summarized into two situations: the first situation, for example, drone a and drone c are defeated by drone b in the same round. At this time, it is necessary to sort them according to the bids of the two defeated drones.

[0081] In the second case, for example, drone b and drone e, both drones are defeated by the same drone in their respective iteration rounds. At this time, further comparison is needed between the two to distinguish their relative priorities.

[0082] S4. Build a task priority chain storage structure based on the information backtracking results.

[0083] Based on the above Figure 4 The UAV task negotiation process analyzed in the previous section further builds the information storage logic based on the original negotiation structure of the UAV. If global information storage is used, each UAV stores the priority sorting sequence of each task, then the data matrix size is , its size will change rapidly with the number of tasks and the number of drones, causing unnecessary storage consumption. Therefore, this section proposes a priority-based chain storage logic.

[0084] In the priority-based chain storage logic, each node in the chain only needs to record the node and its corresponding node of the previous priority, which can greatly compress the storage space of information. For example, in the negotiation result shown in Figure 5, the execution priority of task T from high to low is: At this time, for UAV d, its mission For the drone with the highest priority, it only needs to record the corresponding task as task t. For drone e, since its priority for task T is lower than that of drone d, drone e needs to record both drone d with the next higher priority and the corresponding task T. The same goes for the rest. Thus, a task T execution priority sorting chain consisting of drones a, b, c, d and e can be obtained.

[0085] S5. Quantitatively analyze the allocation model with random node failures, determine the failure locations of drones, and perform mission reconstruction.

[0086] Through the information synchronization mechanism between drones, the node failure type is analyzed, the node failure location is accurately determined, and the task reconstruction is completed.

[0087] In a distributed communication network, the number of consistent interaction iterations of network information is positively correlated with the diameter D of the static connection network. At the same time, the smaller the diameter D of the communication network, the higher the network density is required, which will lead to a higher communication load on the communication network nodes. However, as the scale of the communication network continues to expand, the communication bandwidth of the communication network nodes with too high a density is far from being able to meet the demand. Therefore, under normal circumstances, multi-UAV networking will not adopt a high-density network such as a fully connected network. In a simplified statically connected initial network, although the communication load of each node is reduced, when a node in the connected network fails, it may affect the connectivity of the network, and thus affect the subsequent information flow process. Figure 3 It can be clearly seen that in the initial complete network, the information flow between node 3 and node 5 can be completed indirectly through node 2 and node 4. Similarly, node 1 and node 6 can complete the information flow indirectly through node 2. If random node failures occur in the allocation model, it can be divided into two categories: the first is that the network can still maintain connectivity, that is, any two valid nodes can still establish an information channel; the second is that the network loses connectivity, that is, there are at least two nodes that cannot establish an effective information flow channel. Figure 3 The failure of the node 2 shown in the figure belongs to the second case, in which case the network cannot continue to perform effective information negotiation interaction actions.

[0088] In the distributed task allocation algorithm designed in step 1, if the second node failure occurs, the search trajectory planned by the drone corresponding to the node will not be able to complete effective traversal, which will lead to the direct failure of the search task. If the unfinished search trajectory of the failed drone cannot be decomposed consistently, it will lead to task conflicts between the remaining drones. Here, we quantitatively analyze the second situation and assume that the allocation model in the above is recorded as Indicates a The statically connected initial network of nodes. When one of the nodes, UAV R, fails, multiple sub-networks may be generated. The set of sub-networks is recorded as:

[0089] The above Figure 3 From the analysis of the situation in , we can see that in the initial connected network, if node R is directly connected to two or more other nodes, then the failure of node R will cause the network to be disconnected. At this time, the number of subnetworks generated is , if there is an adjacent node in node R, then the failure of node R will not affect the connectivity of the network. At this time, the number of subnetworks When there is a node failure, but the number of sub-networks generated When the multi-UAV system only needs to re-plan the remaining search paths. Therefore, the present invention focuses on solving the problem of When multiple sub-networks are parameterized and the sub-networks cannot effectively exchange information, task replanning occurs.

[0090] like Figure 7 As shown in the figure, in the response thread between drones, drones use time-staggered response rhythms, and two adjacent drones use different time rhythms to interact, which can effectively reduce information interaction resources while ensuring the reliability of responses between drones. If no signal is sent, it is considered that the drone R has failed, and then according to the initial drone network A collection of subnetworks can be inferred simultaneously , the remaining UAVs synchronize the sub-network set , and enter task reconstruction mode.

[0091] Using the priority chain storage structure, the remaining drones responsible for reconfiguration will traverse whether the failed drone is their last priority drone. If so, the task will be reallocated to the corresponding drone, and by inserting the task into the execution sequence at the best position, it is ensured that the task reconfiguration will not affect the drone's task execution efficiency. This approach can achieve conflict-free task reconfiguration of failed nodes even when multiple unconnected sub-networks are formed.

[0092] In order to verify the effectiveness of the distributed task reconstruction method using historical information backtracking proposed in this invention, a search trajectory reconstruction simulation verification is carried out in a multi-UAV collaborative search scenario. First, the map is rasterized and segmented, and probability elements are integrated and designated as task allocation blocks. Then, a boundary gain function is established, and the path is used as an influencing factor. The reconstruction algorithm is integrated into the multi-UAV system to guide the collaborative search work, and finally efficient task reconstruction and search optimization are achieved. The overall process is as follows: Figure 4 As shown. Figure 5 is the reconstruction result of the search path. As can be seen from the figure, when one of the UAVs fails, the distributed task reconstruction method based on historical information backtracking proposed in this paper successfully reallocates the tasks originally assigned to the failed UAV. In addition, no conflicts are observed in the trajectory planning of the remaining UAVs. In addition, the new search path undergoes significant adjustments compared to the initially planned search trajectory, which is due to the replanning of the entire task package list by the UAV when it undertakes the new search grid.

[0093] In summary, this embodiment provides a distributed task reconstruction method based on historical information backtracking, which uses interactive threads between drones and can detect the failure of drone nodes in a timely and accurate manner, with real-time and high accuracy. It can respond quickly to drone failures, thereby reducing the risk of interruption of task execution and improving reliability and robustness. The advantage of real-time is that it can respond immediately after a drone failure or malfunction occurs, without the need for manual intervention, thereby saving time and labor costs and being able to operate more efficiently.

[0094] By adopting advanced algorithms and technologies, the best task allocation plan can be automatically determined based on factors such as task requirements, drone performance and environmental conditions. Task allocation can be completed quickly and accurately without manual intervention, improving efficiency and responsiveness.

[0095] The system robustness is effectively improved through the automatic task reallocation process. When a drone fails, the task allocation can be adjusted in time to avoid task interruption or delay, thereby ensuring that it can continue to operate in the face of emergencies, which is crucial to ensuring the continuity and reliability of task execution. Through timely task reallocation, it can quickly adapt to different environments and situation changes, reducing the risk of being affected by external interference or internal failures.

[0096] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A distributed task reconstruction method based on historical information backtracking, characterized in that: include: S1. Establish a distributed multi-UAV collaborative task allocation model based on the constraints of the information interaction architecture; S2, generate a preliminary decision plan, including task construction and task negotiation; S3. In the task negotiation phase, an information backtracking mechanism is designed for failed drones; S4. Construct a task priority chain storage structure based on the information backtracking results; S5. Quantitatively analyze the allocation model with random node failures, determine the failure locations of drones, and perform mission reconstruction.

2. A distributed task reconstruction method based on historical information backtracking according to claim 1, characterized in that: In step S1, there are n drones, and the upper limit of each drone completing the target task is , while given target points, the formula of the distributed multi-UAV collaborative task allocation model is expressed as: The constraints are set as: In the formula, Indicates drone The corresponding relationship mark with the task point f, when task f is assigned to the drone When the corresponding relationship is marked ,on the contrary ; Indicates a dimensional vector; Indicates all the tasks occupied by drone i The execution order of represents the profit calculation function; Represents the raw revenue value.

3. A distributed task reconstruction method based on historical information backtracking according to claim 2, characterized in that: In step S2, the task construction link constructs the task package of each drone by adopting a greedy strategy. The construction process includes: Enter new task j and find the current benefit value after adding new task j. The formula is: In the formula, represents the existing mission path of drone i, It means that the UAV i is in the mission path The total revenue value at Indicates that in the existing task path Add task j to the nth position of Indicates at location Find the position n with the maximum profit, and the current profit value is ; Compare the current revenue value with the original revenue value, and determine whether task j should be added to the task package based on the comparison result. , then use Alternative , indicating that task j is added to the task package; if , it means that task j is not added to the task package.

4. A distributed task reconstruction method based on historical information backtracking according to claim 1, characterized in that: In step S2, the task negotiation link resolves the task construction conflicts between drones by constructing a drone task negotiation algorithm based on a market auction mechanism. The task negotiation link includes agreed marking symbols and task negotiation.

5. A distributed task reconstruction method based on historical information backtracking according to claim 4, characterized in that: The agreed marking symbols include: Winner bid values ​​list vector , the vector represents the bid value of each task in the storage unit of drone i; wherein the bid value represents the task benefit value that drone i needs to achieve to win the task; Winners List Vector , the vector Indicates the ownership of each task in the storage unit of drone i; Timestamp vector , the vector Used to mark the information update time of drone i.

6. A distributed task reconstruction method based on historical information backtracking according to claim 1, characterized in that: The process of performing the task negotiation includes: when , , update the information, that is, the information mark about task j in drone i is replaced by the information mark of the corresponding task j in drone k; when , 0, reset the information, that is, all the information about task j in drone i is reset to the initial empty state; when , , information non-operation is performed, that is, the information about task j in UAV i is retained; In the formula, represents the bid value of the jth task in the storage information of drone i; Indicates the serial number value of the drone to which the jth task belongs in the storage information of drone i.

7. A distributed task reconstruction method based on historical information backtracking according to claim 1, characterized in that: In step S3, the process of designing an information backtracking mechanism for a failed UAV includes: S31, given task T; S32, analyzing the bidding process of multiple iterations between each drone, and independently bidding for task T by the drones following the greedy optimization principle, performing the first iteration, comparing the corresponding benefit values ​​of the drones following the greedy optimization principle to evaluate the bids, and determining the winner of the first round; S33, setting a drone that does not follow the greedy optimization principle to join the auction, and comparing the profit value of the current drone with the winner of the previous round, performing the current iteration, and determining the winner of the current iteration; S34. Repeat step S33 until all drones complete the bidding and determine the execution right of task T.

8. The distributed task reconstruction method based on historical information backtracking according to claim 1 is characterized in that: In step S4, the process of constructing the task priority chain storage structure includes: Build priority-based chain storage logic; Record the node of the previous priority for each point in the chain node; Get the execution priority sorting chain of task T.

9. A distributed task reconstruction method based on historical information backtracking according to claim 1, characterized in that: In step S5, the process of quantitatively analyzing the allocation model with random node failures includes: Set the initial network of multi-UAV coordination task as , indicating a A statically connected initial network of nodes; Get the set of sub-networks generated when one of the nodes, UAV R, fails: If node R is directly connected to two or more other nodes, the number of subnetworks generated , task reconstruction is required; If there is an adjacent node in node R, the number of subnetworks generated , no task reconstruction is required.

10. The distributed task reconstruction method based on historical information backtracking according to claim 1 is characterized in that: In step S5, the task reconstruction process includes: Get the location of the failed node; Traverse all drones and obtain the next-level drone of the failed node according to the priority chain storage structure; Reassign the task to the next level drone and insert the task into the execution sequence.

Images