An event-driven rolling time window unmanned aerial vehicle emergency task bidding allocation method

Abstract

The application discloses an unmanned aerial vehicle emergency task bidding allocation method based on event-driven rolling time window, which comprises the following steps: firstly, acquiring unmanned aerial vehicle operation state data and task state data, establishing unmanned aerial vehicle set, task set and storage point set; when receiving an emergency task insertion request, defining the request as a trigger event, constructing an event-driven rolling time window, and forming a task pool; on the basis of frozen data, sequentially performing hard constraint screening based on load capacity and energy accessibility to obtain a candidate unmanned aerial vehicle set; constructing a bidding score model, performing weighted calculation on execution efficiency, distance, health condition, task compatibility and emergency task priority; introducing the bidding score into a multi-objective 0-1 programming allocation model, cooperatively solving unmanned aerial vehicle assignment, storage point selection and power exchange decision under relevant constraints, and outputting an optimal allocation result and generating a scheduling instruction. The application takes into account emergency response efficiency and global cost optimization, and is suitable for online allocation scenarios of continuous emergency task insertion.

Description

Technical Field

[0001] This invention relates to the technical field of drone swarm collaborative scheduling and emergency logistics delivery, and in particular to an event-driven rolling time window method for bidding and allocating emergency drone missions. Background Technology

[0002] The process of inserting and bidding for emergency drone missions generally involves modeling the operational status of the drone and the mission; establishing a drone-mission matching and decision-making mechanism; determining feasibility under constraints such as energy, payload, and time windows; and measuring and ranking the costs and benefits of candidate solutions to output executable scheduling instructions. Based on different decision-making mechanisms and solution frameworks, emergency mission allocation methods can be divided into centralized optimization allocation methods and distributed allocation methods. Centralized optimization methods construct unified models such as mixed-integer programming, incorporating factors such as benefits, energy consumption, risk, and delay into the objective function and solving for the optimal or near-optimal solution under constraints. Distributed methods typically use auction or bidding mechanisms to calculate bid values ​​for "drone-mission" combinations, achieving rapid assignment through bidding. Meanwhile, due to range limitations, some methods further introduce storage points and battery swapping mechanisms to coordinate mission assignment and refueling paths, supporting continuous drone operation and dynamic mission insertion scenarios.

[0003] Research on emergency mission allocation for drones has made some progress, but existing methods have done little work on "state consistency" and "collaborative decision-making" when dealing with frequent insertions of emergency missions. On the one hand, many methods rely directly on continuously changing real-time states for screening and solving, lacking a state freezing mechanism based on the trigger moment. This can easily lead to inconsistencies between candidate evaluation and final assignment data, resulting in execution risks such as energy unavailability or time window violation. On the other hand, bidding designs often focus on a few indicators such as distance or energy consumption, and fail to comprehensively characterize key factors such as health status, original mission conflict degree, and emergency mission priority. Furthermore, storage point selection, activation, and battery swapping decisions are often treated as additional conditions and are not solved in a unified and collaborative manner with assignment decisions, affecting the optimization of global cost and the executability of action sequences. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings and deficiencies of the prior art and provide an event-driven rolling time window method for bidding and allocating emergency missions for unmanned aerial vehicles (UAVs). This method can freeze the system state by using the arrival of an emergency mission as a trigger event, first perform hard constraint screening based on payload and energy accessibility, then characterize the suitability of candidate UAVs for emergency missions through a bidding score model, and directly embed the bidding score into a multi-objective 0-1 planning allocation model to achieve collaborative optimization of "assignment-storage point-battery swapping". This balances the emergency mission response speed with the optimal global cost of the solution and supports rolling allocation in continuously inserted scenarios.

[0005] To achieve the above objectives, the technical solution provided by this invention is: a method for bidding and allocating emergency drone missions using an event-driven rolling time window, comprising the following steps:

[0006] S1: Obtain the operational status data and task status data of all drones within the task area, establish drone sets, task sets and storage point sets, obtain the current location, current remaining energy, payload status and current task execution progress of each drone, and obtain the availability status and whether battery swapping is supported for each storage point;

[0007] S2: When a request to insert an emergency task, i.e. a delivery request, is received, the request is defined as a trigger event and the arrival time of the request is defined as the trigger time. The running status data and task status data in step S1 are frozen based on the trigger time, an event-driven rolling time window is constructed, and the emergency task to be inserted and the regular task that has not been completed at the trigger time are combined into a task pool.

[0008] S3: Based on the state data frozen in step S2, perform feasibility screening on the drone set: first, eliminate drones that do not meet the emergency mission payload requirements based on payload capacity, and then eliminate drones that cannot complete the entire emergency mission delivery process under the current state's remaining energy conditions based on energy accessibility, and retain feasible drones to form a candidate drone set.

[0009] S4: For each candidate drone in the candidate drone set, construct a bidding score calculation model, that is, take the drone operation status data, task status data and path distance information frozen at the trigger time as input, and output the bidding score of each candidate drone for the emergency task.

[0010] S5: Introduce the bidding score obtained in step S4 into a multi-objective 0-1 programming allocation model to solve for the optimal assignment scheme within the candidate UAV set; wherein, the multi-objective 0-1 programming allocation model takes "maximizing the emergency task bidding response score, maximizing the total scheme benefit, minimizing the total scheme execution cost, and minimizing the total task delay penalty" as the comprehensive objective function, and sets relevant variables and constraints in the model, and finally outputs the allocation result that takes into account both the emergency task response speed and the optimal global scheme cost;

[0011] S6: Generate and issue scheduling instructions based on the allocation results obtained in step S5, so that the assigned UAV can execute scheduling instructions containing a preset action sequence; after execution, update the UAV's operating status data and task status data. When a new emergency task insertion request is received again, return to step S2 with a new trigger time to reconstruct the newly triggered event-driven rolling time window, and repeat steps S3 to S6 to achieve rolling allocation in continuous insertion scenarios.

[0012] Furthermore, the specific steps of step S1 are as follows:

[0013] S11: Define a drone ensemble within the mission area. ,in Indicate the drone number; define the task set. ,in Indicates the task number; defines the set of storage points. ,in This indicates the storage point number. The above definition provides a unified description of the drone selection, bidding score calculation, and task allocation in the subsequent emergency task insertion process.

[0014] S12: For the first A drone, obtain its current location coordinates. ,in , , They represent drones x-coordinate, y-coordinate, and altitude coordinates in the mission area coordinate system; current remaining energy. ,in Used to identify energy state and indicate remaining state; drone Maximum load ,in Used to indicate the capacity limit; each drone Execution task number ;

[0015] S13: Obtain any storage point Available status and the storage location Does it support battery swapping? ,in ,when This indicates that the storage point is available, when This indicates that the storage location is unavailable; ,when This indicates that the storage location supports battery swapping. This indicates that the storage location does not support battery swapping; the aforementioned and The status data of the storage point is determined at the time of the triggering event and is used for subsequent storage point selection and battery swapping decisions.

[0016] Furthermore, the specific steps of step S2 are as follows:

[0017] S21: When a request to insert an emergency task is received, the request is identified as the first... The event is triggered for the [number]th time, and the arrival time of the request is defined as the [number]th time. The time when the secondary trigger event occurs ,in Indicates the event sequence number;

[0018] S22: with Build an event-driven scrolling time window as the starting point of the time window ,satisfy:

[0019] ;

[0020] In the formula, Indicates the first The event-driven scrolling time window corresponding to the next triggered event; This represents a constant for the length of the time window, where Used to indicate the preset length of the time window;

[0021] S23: with The operational and task status data obtained in step S1 are frozen at a reference time, ensuring that subsequent drone selection, bidding, and allocation are all based on data from the same moment. Furthermore, based on the frozen status data, the emergency tasks to be inserted are compiled into an emergency task set. ,in Used to identify the task type as an urgent task, and then the... and The task pool consists of routine tasks that the drones have not yet completed. The As an event-driven scrolling time window Task bidding and allocation within the organization.

[0022] Furthermore, the specific steps of step S3 are as follows:

[0023] S31: At the triggering time Based on the frozen state data, for the task pool emergency mission Set up an emergency task The load requirement is ,in Used to identify urgent mission payload requirements, when Remove drones The remaining drones then proceed to energy reachability assessment;

[0024] S32: For drones retained in step S31 Perform energy reachability determination; for any storage point in the storage point set. With emergency mission Define drone Storage point Execute emergency mission Total energy consumption throughout the entire delivery process for:

[0025] ;

[0026] In the formula, Indicates drone Fly from current location to storage point Energy consumption, of which Used to identify drones Fly from current location to storage point The movement phase; Indicates from storage point Head to an emergency mission And the energy consumption for completing the delivery, of which Used to identify from the storage point Head to an emergency mission The delivery execution phase; Indicates completion of an emergency mission Afterwards, return to the storage point The energy consumption of returning to base, of which Used to identify the completion of an urgent task Afterwards, return to the storage point The return phase;

[0027] The energy consumption of each of the above is determined based on the frozen location information, path distance and UAV energy consumption at the time of the triggering event, and serves as the basis for subsequent energy hard constraint screening calculations;

[0028] S33: Setting up drones The minimum retention energy threshold is drones The minimum energy required to maintain flight, of which Used to identify the minimum energy required for the drone; store the following hard energy constraints:

[0029] When the drone does not require battery swapping, the following requirements must be met:

[0030] ;

[0031] When the drone exchanges batteries at a storage location that supports battery swapping, let the full battery capacity be... ,in The slow-point energy used to identify drones must meet the following requirements:

[0032] ;

[0033] Based on this, we retain the UAVs that satisfy the above-mentioned energy hard constraint conditions, forming the first... The set of candidate drones corresponding to the triggering event ,in:

[0034] .

[0035] Further, in step S4, the bidding score is a comprehensive score used to characterize the suitability of candidate drones for performing emergency tasks; the comprehensive score is composed of the drone's execution efficiency score, distance score, health status score, task compatibility score, and emergency task priority score; each of the above scores is normalized to obtain a normalized score, and the normalized scores are weighted and fused according to the weights of the user and decision-maker for each score to obtain the bidding score, satisfying the constraint that it is non-negative and the sum of the weights is 1; the bidding score and its corresponding normalized score for each candidate drone are output as one of the input parameters for introducing the bidding score and solving the allocation in step S5.

[0036] Further, in step S4, the bidding score calculation model performs the following operations:

[0037] S41: In the obtained... The set of candidate drones corresponding to the triggering event Inside, for each candidate drone Calculate its impact on emergency tasks Bidding score ,satisfy:

[0038] ;

[0039] ;

[0040] In the formula, the bidding score Indicates drone For emergency missions The degree of compatibility; Indicates the first The weights of each scoring indicator are weight parameters pre-defined by decision-makers and users; Indicates drone For emergency missions The Normalized scores for each scoring indicator Corresponding to drones For emergency missions Execution efficiency score Distance rating Health status score Task compatibility score Priority rating for urgent tasks ;

[0041] Normalized score s corresponding to the execution efficiency score satisfy:

[0042] ;

[0043] In the formula, Indicates drone Expected to complete the emergency mission Response completion time; Represents the maximum response completion time constant used for normalization, derived from the first... The set of candidate drones corresponding to the triggering event The drone that takes the longest time to complete an emergency mission is identified, representing the longest time a drone in the candidate drone set takes to complete the emergency mission.

[0044] Normalized score corresponding to distance score satisfy:

[0045] ;

[0046] Among them, the shortest path length index satisfy:

[0047] ;

[0048] In the formula, Indicates drone Execute emergency mission At that time, gather at the storage point. The shortest path length index corresponding to the selection of the optimal storage point; Indicates drone to storage point distance, Indicates storage point Distance to emergency mission C;

[0049] Normalized score corresponding to health status score satisfy:

[0050] ;

[0051] In the formula, Indicates drone The health risk rate, with a range of values ​​being [value missing]. , indicating drone The probability of failure; the lower the risk, the lower the probability of failure. The larger;

[0052] Normalized score corresponding to task compatibility score satisfy:

[0053] ;

[0054] In the formula, Indicates the time when the drone triggered the event. Routine tasks currently in progress; Indicates routine tasks With emergency mission The conflict degree function value is determined by factors such as the type of goods and the degree of overlap between the original task and the emergency task. The greater the conflict degree, the lower the score.

[0055] Normalized score corresponding to emergency task priority score satisfy:

[0056] ;

[0057] In the formula, Indicates an emergency task Priority parameters; Indicates drone Current routine tasks The priority parameter; the normalized score This is used to characterize the priority of an emergency task relative to the routine tasks currently being performed by the drone; when the priority of an emergency task is higher than that of the routine task, the normalized score is correspondingly higher.

[0058] S42: Obtain each candidate drone according to step S41. For emergency missions Bidding score and output the and its corresponding normalized score As part of step S5, when establishing the multi-objective 0-1 planning allocation model, the "candidate UAV set" One of the input parameters within the range.

[0059] Furthermore, in step S5, in the... The set of candidate drones corresponding to the triggering event Emergency Task Collection Set of storage points The following decision variables are established:

[0060] a. Assigning variables:

[0061] ;

[0062] in, Indicates candidate drones Are you assigned to perform an emergency mission? ,when Indicates assignment; otherwise, no assignment is made. (Variable) , ;

[0063] b. Variable selection for storage point:

[0064] ;

[0065] in, Indicates drone Execute emergency mission Choose storage location As a transit storage point, variables ;

[0066] c. Enable variables at storage points:

[0067] ;

[0068] in, Indicates drone Enable storage point Store or temporarily save its current routine tasks;

[0069] d. Decision variables for battery swapping:

[0070] ;

[0071] in, Indicates drone Choosing a storage location Execute emergency mission Battery swapping is performed before, and vice versa. Then battery swapping will not be performed; similarly Indicates drone Complete the emergency mission Return to storage point Then, to continue the original task, battery swapping was performed; otherwise... Then battery swapping will not be performed, among which Used to identify the battery swapping phase when returning to continue the original task;

[0072] e. Timeout variable:

[0073] ;

[0074] in, This indicates the cumulative timeout amount incurred by subsequent regular tasks after an emergency task was inserted; Indicates an emergency task The timeout amount.

[0075] Furthermore, in step S5, the construction and solution of the comprehensive objective function are as follows:

[0076] a. Constructing the objective function for the bidding response This is used to directly embed the bidding score obtained in step S4 into the optimized allocation, so that the assignment result tends to select the drone with the higher bidding score and more suitable for performing the emergency mission. The bidding response objective satisfies:

[0077] ;

[0078] In the formula, The candidate drones obtained in step S4 For emergency missions The bidding score;

[0079] b. Construct the objective function for the scheme's revenue. This is used to characterize the benefits of completing urgent tasks, ensuring that the model prioritizes the effective execution of high-reward tasks, provided it is feasible. The benefit objective of the proposed solution satisfies:

[0080] ;

[0081] In the formula, Indicates an emergency task The benefit parameters are determined by the priority level of the emergency task when it is inserted;

[0082] c. Construct the objective function for the total cost of the solution. This solution comprehensively reflects the energy consumption costs incurred by candidate UAVs when performing emergency missions, the costs of transit storage and handling at storage points, and the costs of battery swapping, aiming to minimize costs while completing emergency missions.

[0083] ;

[0084] Total cost target of the plan From the activated storage point Storage and handling costs Total cost of a single item Composition, in which , These are used to identify the costs incurred in storage and handling, as well as the total cost of each item:

[0085] To enable storage points The cost parameters for storage and loading / unloading are obtained from the storage point in step S1. Given a state;

[0086] Indicates "candidate drones" Select storage location Execute emergency mission The corresponding total cost per item includes energy consumption converted costs. Total cost of battery swapping , The calculation and definition are as follows:

[0087] ;

[0088] In the formula, energy consumption is converted into cost. The conversion factor based on unit energy consumption With drones via storage point Execute emergency mission Energy consumption throughout the entire process The sum of the products determines the superscript. Used to indicate the cost calculated based on unit energy consumption; total cost of battery swapping. The cost of a single battery swap With drones In carrying out emergency missions At the storage point Should the battery be replaced? And to complete emergency missions Return to storage point To continue the original mission Should the battery be replaced? It was jointly determined that, among which This is used to indicate that the cost is incurred by battery swapping;

[0089] d. Construct the objective function for task delay penalty This is used to suppress situations where urgent tasks time out, enabling the model to detect delays caused to the original tasks by the insertion of urgent tasks. and drones Exceeding the deadline for completing urgent tasks "A penalty is given, and the objective of the task delay penalty is satisfied as follows:"

[0090] ;

[0091] In the formula, , They represent emergency missions. Insert for regular tasks The relationship between the resulting delay penalty and the severity coefficient of the delay penalty for completing an urgent task is as follows: This indicates that the penalty for delays in emergency tasks is greater than the penalty for delays in regular tasks.

[0092] Among them, the total delay penalty caused by the insertion of an emergency task to subsequent regular tasks. Determine as follows: Assume an unmanned aerial vehicle (UAV) When no urgent task is inserted, its subsequent regular tasks The expected completion times according to the established sequence are ,Task The deadline is ,in Used to identify the deadline for urgent tasks; when an urgent task is inserted... Subsequently, regarding the drone at the time of insertion. An unfinished set of follow-up routine tasks Incremental insertion time If this causes the overall postponement of subsequent routine tasks, then the timeout amounts of the subsequent routine tasks are summed and the non-negative portion is taken, defined as:

[0093] ;

[0094] The extra time inserted According to drones The insertion procedure for executing an emergency mission has been determined, and the insertion procedure includes: the drone moving from its current location to the storage point. Save the current task and proceed from the save point to perform the emergency task. Return to the original task save point And continue to complete the original delivery task; the insertion of an emergency task results in additional time for the original delivery task. satisfy:

[0095] ;

[0096] In the formula, For drones Select storage location And the total time to complete the emergency task inserted into the delivery process;

[0097] The emergency task delay penalty Determine as follows: Assign an emergency task The latest completion time is When drones Assigned to perform emergency tasks ,Right now At that time, the estimated completion time for its emergency mission is ,like Then no delay will occur, if This results in a delay; the emergency task delay penalty is defined as the non-negative portion of the timeout amount by which the emergency task completion time exceeds the time limit.

[0098] ;

[0099] in, The insertion process time is obtained, and it satisfies:

[0100] ;

[0101] In the formula, For drones The current moment when the decision to insert the order is made; For drones via storage point Complete the emergency mission Total time spent;

[0102] e. Quantize the above objective functions, since... , , , The units of measurement may differ and the numerical scales may also differ. Min-max dimensionality is used to normalize each objective function to an interval. ,use Represents any objective function, for any objective function Given upper and lower bounds and The normalized objective function satisfies:

[0103] ;

[0104] In the formula, Describe the objective function Dimensional results; As the target identifier, respectively corresponding to , , , ; and These represent the candidate range for the currently triggered event. , , The lower and upper bounds within the candidate drone set are defined before each insertion. Emergency Task Collection Set of storage points Calculated within the candidate range; To prevent positive constants with a denominator of zero;

[0105] f. Construct a dimensionalized comprehensive objective function and solve for its maximization:

[0106] ;

[0107] In the formula, This represents the dimensionalized comprehensive objective function; , , , They are respectively , , , Dimensional results; Configure the settings in advance before the event is triggered.

[0108] Furthermore, in step S5, the following constraints are applied, and the decision results are solved and output:

[0109] a. Apply a unique task assignment constraint to ensure that each emergency task is uniquely assigned to a single candidate drone under this triggering event:

[0110] ;

[0111] The above formula represents the expression for each emergency task. The sum of the assignment variables for all candidate drones must equal 1, and this constraint applies to all... All are true;

[0112] b. Apply a drone mission capacity constraint to ensure that each candidate drone undertakes at most one emergency mission in the solution of this triggered event. The constraint satisfies:

[0113] ;

[0114] The constraint applies to all All are true;

[0115] c. Apply storage point selection and allocation constraints to ensure that "only drones assigned to perform emergency missions are allowed to select storage points," satisfying:

[0116] ;

[0117] The constraint applies to all , All are true;

[0118] d. Apply storage point path logic constraints to establish a consistent relationship between "whether to execute an emergency mission through a certain storage point" and "whether to use the storage point to store a regular mission," and limit each drone to using at most one storage point, satisfying:

[0119] ;

[0120] ;

[0121] Ensure the selection of storage location During transit The storage point has been enabled. And it limits each drone to using a maximum of one storage location; this constraint applies to all... , , All are true;

[0122] e. Apply battery swapping logic constraints to ensure that: only when the candidate drone... In carrying out emergency missions When the storage point was selected As a transit and storage point Furthermore, this storage location supports battery swapping. Only when this condition is met is battery swapping permitted at the storage location. or The battery swapping logic constraints include battery swapping constraints before executing an emergency task and battery swapping constraints before returning to the storage point to continue the original task after completing the emergency task, as detailed below:

[0123] 1) Battery swapping decisions before carrying out emergency missions This is used to constrain whether candidate drones can be stored at a designated location before heading to perform an emergency mission. Battery swapping is performed to meet the following requirements:

[0124] ;

[0125] The above constraint means that if the candidate drone No storage location selected As an emergency mission The transit storage point, or the storage point Battery swapping is not supported. Then the battery swapping decision variables before executing the emergency task The value can only be 0, which means that battery swapping is not allowed at this storage point;

[0126] ;

[0127] The above constraint means that the storage point is selected if and only if the storage point supports battery swapping. Battery swapping decision variables before executing an emergency mission Only a value of 1 is allowed, corresponding to the battery swapping constraint before executing an emergency mission for all candidate drones. Emergency mission and storage points All are true;

[0128] 2) Decisions regarding battery swapping before returning to the storage point to continue the original task after completing an emergency mission. This is used to constrain candidate drones to complete emergency missions and return to their storage locations. Afterwards, before continuing the original routine task, can the battery be swapped at this storage point, satisfying the following conditions:

[0129] ;

[0130] The above constraint means that if the candidate drone No storage location selected As an emergency mission The transit storage point, or the storage point Battery swapping is not supported. Then the returned battery swapping decision variables The value can only be 0, which means that battery swapping is not allowed at this storage point;

[0131] ;

[0132] The above constraint means that the storage point is selected if and only if the storage point supports battery swapping. The returned battery swapping decision variables Only a value of 1 is allowed, corresponding to the battery swapping constraint before returning to the storage point to continue the original task after completing the emergency mission, for all candidate drones. Emergency mission and storage points All are true;

[0133] f. To ensure that when candidate drones are available... Assigned to perform emergency tasks And choose a storage location As a transit and storage point At that time, it was from the storage point Depart to complete the emergency delivery and return to the storage point. Energy consumption shall not exceed the available energy threshold. The following energy constraints shall be imposed for emergency mission delivery:

[0134] ;

[0135] Wherein, the constraint applies to all , , All are true. A pre-set large constant is used in The constraint is relaxed at this time; the energy constraint for emergency mission delivery is determined by the intermediate variable of energy consumption for the round trip of the emergency mission delivery. ,in Intermediate variables used to identify the round-trip delivery phase of an emergency delivery mission and the available energy threshold. , "Combined composition, its definition is as follows:"

[0136] Indicates from storage point Depart to complete the emergency mission Delivery energy consumption Add return storage point energy consumption The sum of them satisfies:

[0137] ;

[0138] Indicates drone Without swapping batteries, the remaining energy in the current state of the drone... Arrival at storage point Consume energy And deduct the safe energy threshold The available energy threshold that can then be used to continue executing the task satisfies:

[0139] ;

[0140] This represents the increase in available energy relative to the threshold of not swapping batteries when swapping them at the storage location; in other words, the battery capacity after swapping. Plus storage point Consume energy And offset the remaining energy of the current drone state. The energy, satisfying:

[0141] ;

[0142] g. Apply energy reachability constraints to the storage point, since the battery swapping occurs at the storage point. Regardless of whether the battery is swapped or not, the drone must be guaranteed to function properly. Can reach the storage point using the current remaining energy. The energy reachability constraints to the storage point are applied as follows:

[0143] ;

[0144] Wherein, the constraint applies to all , , Both are true, when At that time, the above formula is used to limit the energy consumption of movement to the storage point to not exceed the "remaining energy". Deducting the safe energy threshold The threshold of "".

[0145] h. To ensure the drone Complete the emergency mission Return to storage point Remaining usable energy :

[0146] ;

[0147] It can satisfy the need to continue executing the temporarily stored subsequent regular tasks, impose energy constraints on continuing the original tasks, and determine the remaining available energy after returning to the storage point after completing the emergency task. With drones Minimum available energy required to complete subsequent routine tasks Divided into "no battery replacement after return" "and battery swapping upon return" Two scenarios:

[0148] No battery swapping will be performed upon return. At that time, the remaining available energy is required. Meet the needs of drones Minimum available energy required to complete subsequent routine tasks :

[0149] ;

[0150] The above constraints are used to characterize whether the drone still has the energy feasibility to continue performing its original conventional mission without battery swapping. This applies to all... , , All are true;

[0151] Battery swapping will be performed upon return. At that time, the required battery capacity after battery swapping is... Deducting the safe energy threshold Post-satisfaction drones Minimum available energy required to complete subsequent routine tasks :

[0152] ;

[0153] The above constraints characterize whether the UAV has the energy feasibility to continue performing its original conventional mission under battery swapping conditions. For all... , , All are true;

[0154] i. Apply an urgent task time limit to ensure that urgent tasks are completed before the deadline, satisfying the following:

[0155] ;

[0156] In the formula, Indicates drone Select storage location Execute emergency mission Delivery time; Indicates storage point Battery swapping service time; Indicates an emergency task The deadline, given by the urgent task insertion request, applies to all... , , All are true;

[0157] j. Solve and output the decision results. The solution is used to obtain the optimal allocation scheme, i.e., the scheduling instruction set, under the combined effect of the above comprehensive objective function and constraints, and outputs the emergency task assignment results, storage point selection results, and battery swapping decision results. The emergency task assignment results are determined by... The value of is determined, and the storage point selection result is determined by . and The value of is determined, and the battery swapping decision is made by . , The value of is determined.

[0158] Furthermore, the specific steps of step S6 are as follows:

[0159] S61: In the The time when the secondary trigger event occurs Based on the solution result of step S5, a scheduling instruction is generated and issued to the assigned UAV. Insert the arrival time of the request for urgent tasks;

[0160] S62: After receiving the dispatch instruction, the assigned UAV executes a sequence of actions to insert a task. The sequence of actions includes: interrupting the current routine task, going to the storage point to perform storage, making a battery swap decision, going to the emergency task point to complete the emergency task, and returning to the storage point to resume the execution of the original task. The sequence of actions is determined by the dispatch instruction output at the trigger time step S5.

[0161] The storage action of the storage point in the insertion action sequence satisfies: if and only if there exists Make At that time, drones Go to the corresponding storage point Perform regular tasks and store data. The solution is obtained from step S5;

[0162] The battery swapping action in the insertion action sequence satisfies: if and only if there exists Make At that time, drones At the corresponding storage point Battery swapping is performed; among which The result is obtained from step S5, and it is determined whether the storage point supports battery swapping. The storage point status data obtained in step S1 is given;

[0163] The emergency task execution action in the order insertion action sequence satisfies: if and only if there is an emergency task. Make At that time, drones Head to the emergency mission point and complete the emergency mission. ;in The urgent task set is obtained from step S5. Given by the emergency task insertion request;

[0164] If and only if there exists Make At that time, drones After completing the urgent mission Return to the corresponding storage point A battery swap is then performed to allow the temporarily stored subsequent routine tasks to continue. The result is obtained from step S5, and whether the storage point supports battery swapping is also determined by whether it supports battery swapping status. Given;

[0165] S63: After the insertion action sequence is completed, update the status data of the UAV and the storage point, including the current position coordinates of the UAV. Current remaining energy And the current task execution progress, as the basis for freezing state data when the next event is triggered;

[0166] S64: When the next emergency task insertion request arrives, the first... When the event is triggered for the second time, its arrival time is defined as ,in Calculated from the trigger time; and the freeze time in step S2 is returned. The operation and mission status of drones are reconstructed, and the event-driven rolling time window is restructured. And repeat steps S3 to S6; wherein:

[0167] ;

[0168] In the formula, Indicates the first The event-driven scrolling time window corresponding to the next triggered event.

[0169] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0170] 1. This invention adopts an event-driven rolling time window method to freeze the system state at the trigger moment, so that the entire process of screening, bidding and allocation is based on data at the same moment, thereby enhancing the consistency and executability of decision-making in online insertion scenarios.

[0171] 2. This invention considers both load capacity and energy accessibility during the hard constraint screening stage, and combines storage location and power swapping conditions to make a feasibility judgment, effectively avoiding unreachable assignments and improving the reliability of emergency task execution.

[0172] 3. This invention constructs a bidding score calculation model composed of execution efficiency, distance, health status, task compatibility and emergency priority, and directly introduces the bidding score into a multi-objective 0-1 planning allocation model. Under the constraints of uniqueness, energy closed loop and time window, it jointly solves the collaborative decision of "assignment-storage point-battery swapping", thereby taking into account both emergency response speed and optimal global cost of the solution.

[0173] 4. This invention generates an action sequence for emergency task insertion that includes "interruption-storage-power swap-execution-recovery" based on the optimal allocation result, and reconstructs the time window and solves repeatedly when a new emergency request arrives, which is suitable for online allocation of emergency tasks that arrive continuously.

[0174] In summary, this invention ensures that the entire process of emergency task insertion decision-making is based on data from the same moment by using event-triggered state freezing and rolling time windows, thereby improving the consistency and executability of online allocation. Simultaneously, it combines hard constraints such as payload and energy accessibility for screening and integrates multi-dimensional scoring based on execution efficiency, distance, health status, task compatibility, and emergency priority to enhance the characterization of candidate UAV suitability. Furthermore, the scoring results are incorporated into multi-objective 0-1 planning allocation, optimizing assignment, storage point selection, and battery swapping decisions under energy closed-loop and time window constraints, balancing emergency response speed with optimal global cost. This allows the invention to adapt to dynamic scenarios where emergency tasks arrive continuously, improving the overall system scheduling efficiency and task completion quality. Attached Figure Description

[0175] Figure 1 This is a flowchart illustrating the method of the present invention.

[0176] Figure 2 A schematic diagram illustrating the construction of an event-triggered scrolling time window.

[0177] Figure 3 This is a schematic diagram showing the results of the feasibility screening for an emergency mission.

[0178] Figure 4 This diagram illustrates the ranking of candidate drones by their bidding scores and the output of the candidate set.

[0179] Figure 5 This is a diagram illustrating the bidding scores of candidate drones for emergency missions.

[0180] Figure 6 This is a schematic diagram illustrating the execution of scheduling instructions and the restoration of the original task after an emergency task is inserted.

[0181] Figure 7 A schematic diagram of the optimal dispatch route for emergency delivery tasks.

[0182] Figure 8 This diagram illustrates the time breakdown and total time calculation for the emergency order insertion process. Detailed Implementation

[0183] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, but the implementation of the present invention is not limited thereto. For ease of explanation, this embodiment is described using "one optimization triggered by a single emergency task insertion" as an example; when new emergency task insertion requests are received continuously, the system state is re-frozen at the new trigger time, the event-driven rolling time window is reconstructed, and the subsequent steps are repeated to achieve rolling allocation in continuous insertion scenarios.

[0184] like Figure 1 As shown, this embodiment discloses an event-driven rolling time window method for bidding and allocating emergency drone missions, including steps S1 to S6.

[0185] Furthermore, the method executes steps S1 to S6 sequentially each time a triggering event occurs, and each step is triggered at the same time. The frozen data serves as a unified calculation benchmark to ensure data consistency and the feasibility of the solution in the insertion scenario.

[0186] S1: Obtain runtime status data and task status data

[0187] Within the mission area, first define the drone ensemble. Task Set and storage point set ,in Indicates the drone's serial number. Indicates the task number. This indicates the storage location number.

[0188] Based on the above definitions, a unified description is provided for the selection of drones, calculation of bidding scores, and task allocation in the subsequent emergency task insertion process.

[0189] For any drone Get its current location coordinates and current remaining energy. Full charge capacity Maximum load Health risk rate Current task number, current task priority And the task execution progress; for any storage point Get its availability status And whether battery swapping is supported. Among them, when When the storage point is available, it indicates that the storage point is available. This indicates that the storage location supports battery swapping.

[0190] In this embodiment, at the triggering time Freeze the status data of each drone, as shown in Table 1.

[0191] Table 1. Data on frozen state at trigger time

[0192]

[0193] Simultaneously, the availability status of storage points, battery swapping support status, and parameters such as storage cost, battery swapping cost, and battery swapping time are obtained, as shown in Table 2. This data serves as the unified input for filtering, bidding, and optimization within the event-driven rolling time window.

[0194] Table 2. Available Status Information of Trigger Time Storage Point

[0195]

[0196] S2: Construct an event-driven scrolling time window and form a task pool.

[0197] When an emergency task insertion request is received, the request is identified as the first... This event is triggered, and the arrival time of the request is defined as the trigger time. .by Build an event-driven scrolling time window as the starting point of the time window ,satisfy:

[0198] ;

[0199] in, This is a preset time window length constant. Then, with... The operational and task status data obtained in step S1 are frozen at a reference time, ensuring that subsequent drone selection, bidding, and allocation are all based on data from the same moment. Furthermore, based on the frozen status data, the emergency tasks to be inserted are compiled into an emergency task set. Then, the set of emergency tasks and The task pool consists of routine tasks that the drones have not yet completed. As an event-driven scrolling time window Task bidding and allocation within the organization. Figure 2 The process of constructing the triggering event and the event-driven scrolling time window is shown.

[0200] In this embodiment, take min, Min, insert one emergency task E1, the key inputs of which are shown in Table 3.

[0201] Table 3 Emergency Task Event Information Table

[0202]

[0203] S3: Perform feasibility screening on the drone ensemble to form a candidate drone ensemble.

[0204] like Figure 3 As shown, at the triggering time Based on the frozen state data, for the task pool Emergency missions First, screen based on payload capacity, then screen based on energy accessibility, retaining feasible UAVs to form the first... The set of candidate drones corresponding to the triggering event .

[0205] First, set an emergency task. The load requirement is When drones Maximum load When this happens, the drone is removed, and the remaining drones proceed to the energy reachability assessment. For emergency mission E1 in this embodiment, its required payload is 2.0, therefore:

[0206] ;

[0207] As shown in Table 1, the maximum load of U3 is 1.59 and the maximum load of U4 is 1.74. Both are less than 2.0, so they are directly eliminated in the load capacity screening stage.

[0208] Then, the drones that were selected and retained based on payload capacity were... For any storage point in the set of storage points With emergency mission Define drone via storage point Execute emergency mission Total energy consumption throughout the delivery process for:

[0209] ;

[0210] in, Indicates drone Fly from current location to storage point energy consumption Indicates from storage point Head to an emergency mission And the energy consumption to complete the delivery, Indicates completion of an emergency mission Afterwards, return to the storage point The energy consumption for returning to base. The energy consumption values ​​mentioned above are determined at the moment the triggering event occurs based on the frozen location information, path distance, and the UAV energy consumption model.

[0211] Further establish drones The minimum retention energy threshold is When the drone does not require battery swapping, the following requirements must be met:

[0212] ;

[0213] When the drone exchanges batteries at a storage location that supports battery swapping, let the full battery capacity be... The following requirements must be met:

[0214] ;

[0215] Furthermore, regardless of whether battery swapping is performed, energy accessibility during the arrival at the storage point must be met; that is, the remaining energy constraint must not be violated before reaching the storage point. Based on the above hard constraints, this embodiment obtains a set of candidate drones:

[0216] ;

[0217] The corresponding filtering results are shown in Table 4.

[0218] Table 4 Feasibility Screening Table for Hard Constraints

[0219]

[0220] S4: Construct a bidding score calculation model and output the bidding scores of candidate drones.

[0221] After obtaining the candidate drone set Afterwards, as Figure 4 and Figure 5 As shown, for each candidate drone With emergency mission Calculate the bid score The bidding score is a comprehensive score used to characterize the suitability of candidate UAVs for performing emergency missions. The comprehensive score consists of execution efficiency score, distance score, health status score, mission compatibility score, and emergency mission priority score. Each of these scores is normalized to obtain a normalized score, which is then weighted and fused according to preset weights to satisfy the following conditions:

[0222] ;

[0223] ;

[0224] in, Indicates the first The weight of each scoring indicator These correspond to the normalized results of execution efficiency score, distance score, health status score, task compatibility score, and emergency task priority score, respectively. The specific definitions of each scoring item are consistent with the claims: the execution efficiency score is related to the expected response completion time for the emergency task; the distance score is related to the shortest path length indicator; and the health status score is determined by the health risk rate. It is determined that the mission compatibility score is determined by the conflict degree function between the current routine mission and the emergency mission, while the emergency mission priority score is used to characterize the priority of the emergency mission relative to the routine mission currently being performed by the UAV.

[0225] In this embodiment, the top bidding scores of each candidate UAV at the optimal storage point are shown in Table 5. As can be seen from Table 5, U1 achieved the highest bidding score of 0.856520 when executing emergency mission E1 via D1, indicating that U1 has the highest overall suitability for this emergency mission.

[0226] Table 5 Top Bidding Scores

[0227]

[0228] S5: Incorporate the bidding scores into a multi-objective 0-1 programming allocation model and solve for the optimal assignment scheme.

[0229] In the The set of candidate drones corresponding to the triggering event Emergency Task Collection Set of storage points Establish joint decision variables, including: assignment variables Storage point selection variable Enable variables at storage points Battery swapping decision variables before performing emergency tasks Decision variables for battery swapping before returning to the storage point to continue the original task after performing an emergency mission. and the cumulative timeout of regular tasks and emergency mission overtime Then, the bidding score obtained in step S4 will be used. A multi-objective 0-1 programming allocation model is introduced to solve for the optimal assignment scheme within the candidate UAV set.

[0230] In this embodiment, the multi-objective optimization includes the following objectives:

[0231] Bidding response objectives:

[0232] ;

[0233] In the formula, The candidate drones obtained in step S4 For emergency missions The bidding score;

[0234] Profit target:

[0235] ;

[0236] In the formula, Indicates an emergency task The benefit parameters are determined by the priority level of the emergency task when it is inserted;

[0237] Cost Objective:

[0238] ;

[0239] In the formula, To enable storage points Parameters of storage and handling costs Indicates "candidate drones" Select storage location Execute emergency mission "The corresponding total cost per item includes the sum of energy consumption conversion cost and battery swapping cost;

[0240] Delayed penalty target:

[0241] ;

[0242] In the formula, , They represent emergency missions. Insert for regular tasks The severity coefficient of the delay penalty and the delay penalty for completing the emergency task; Insert the total delay penalty for subsequent regular tasks into the emergency task. Delay penalty amount for emergency tasks.

[0243] After the above objectives are dimensionalized, a comprehensive objective function is constructed and solved. In this embodiment, all weights are the same.

[0244] ;

[0245] To ensure the feasibility of the solution, the model is further constrained by the following constraints: unique assignment of emergency tasks, UAV capacity constraints, storage point selection and path logic constraints, battery swapping logic constraints, energy closed-loop and reachability constraints, time-limited completion constraints, and variable domain constraints. Among these:

[0246] 1) Unique assignment constraint for emergency tasks:

[0247] For any emergency task in the emergency task set, constrain it to be uniquely assigned in this triggering event:

[0248] ;

[0249] in, Indicates drone Execute emergency mission ,otherwise ;in .

[0250] 2) Drone capacity constraints:

[0251] For any drone in the candidate drone set The number of emergency tasks it undertakes in this triggered event is limited to no more than 1:

[0252] ;

[0253] 3) Storage point selection and path logic constraints

[0254] To ensure that "transfer storage points are selected only when assigned" and that "storage point activation is consistent with storage point selection", the following constraints are established:

[0255] (1) Consistency between the assigned location and the selected storage point:

[0256] ;

[0257] in, Indicates drone Execute emergency mission Choose storage location As a transit point; among which , .

[0258] (2) Selection and activation consistency:

[0259] ;

[0260] in, Indicates drone Enable storage point ;in , , .

[0261] (3) For any UAV in the candidate UAV set Only one storage point can be used at a time.

[0262] ;

[0263] 4) Battery swapping logic constraints

[0264] To ensure that "battery swapping is only allowed when a storage location is selected and that location supports battery swapping," the following constraint is established:

[0265] (1) Battery swapping logic before the completion of an emergency task:

[0266] ;

[0267] ;

[0268] (2) Battery swapping logic after emergency mission completion:

[0269] ;

[0270] ;

[0271] in, This indicates the location at the storage point before carrying out an emergency mission. Battery swapping; This indicates the location at the storage point after returning from an emergency mission. Battery swapping; Indicates storage point Does it support battery swapping?

[0272] 5) Regarding energy closed-loop and accessibility, such as Figure 6 As shown, the action chain "arrive at storage point - execute emergency mission - return to base - resume regular mission" is written as a solvable 0-1 constraint. The energy consumption for flying to storage point is defined. Energy consumption during emergency round trip Total energy consumption And the energy required to resume regular tasks And combined with the minimum energy retention threshold Remaining energy Fully charged Establish constraints such as arrival at the storage point, no power replacement, power replacement before execution, and restoration of the original task after execution.

[0273] (1) Accessibility at the storage point stage:

[0274] ;

[0275] in , .

[0276] (2) Energy constraints throughout the entire process without battery swapping:

[0277] ;

[0278] Among them, when and When, the above expression is not Relax.

[0279] (3) Energy constraints for emergency round trips during battery swapping before an emergency mission:

[0280] ;

[0281] (4) Energy closed-loop constraint for resuming regular tasks after an emergency task is completed:

[0282] When not changing batteries after returning to base:

[0283] ;

[0284] When swapping batteries after returning to base:

[0285] ;

[0286] 6) Regarding time constraints, define the time required to fly to the storage point. Emergency round-trip flight and service time Battery swapping service hours Storage / loading / unloading service time Then drone via storage point Execute emergency mission The total time taken is:

[0287] ;

[0288] Further establish time-limited completion constraints and introduce sufficiently large constants. The deadline for the emergency mission is And set an emergency task timeout variable. :

[0289] ;

[0290] 7) Domain constraints:

[0291] ;

[0292] ;

[0293] Based on satisfying the above comprehensive objective function and all constraints, the optimal allocation scheme of this embodiment is obtained: the optimal assignment is U1, which is transferred through D1, without power switching before and after execution, satisfying the deadline and achieving the optimal comprehensive objective. The results are shown in Table 6.

[0294] Table 6 Summary of Optimal Allocation Results

[0295]

[0296] S6: Generate scheduling instructions and perform rolling updates

[0297] In the The time when the secondary trigger event occurs Based on the solution result of step S5, a scheduling instruction is generated and sent to the assigned UAV. After receiving the scheduling instruction, the assigned UAV executes a sequence of actions to insert a task. The sequence of actions includes: interrupting the current regular task, going to the storage point to perform storage, making a battery swap decision, going to the emergency task point to complete the emergency task, and returning to the storage point to resume the original task execution sequence. Figure 7 The optimal scheduling path for emergency delivery is shown. Figure 8 The time breakdown and total time calculation process for the emergency order insertion procedure are shown.

[0298] Specifically, in this embodiment, based on the optimal solution... , , and The value of generates scheduling instructions, causing the assigned drone U1 to complete the emergency task insertion according to the action chain of "interrupting the original task - flying to the storage point - storage / loading / unloading - battery swapping decision - executing the emergency task - returning to base - battery swapping decision - resuming the original task". Since the solution results of this embodiment show that U1 transits through D1 and does not need to swap batteries before or after execution, U1 directly executes the emergency task E1 after completing storage / loading / unloading at D1, and returns to D1 after completing the task to resume the original task execution.

[0299] After the insertion action sequence is completed, the UAV's current position coordinates, current remaining energy, current mission execution progress, and storage point-related status data are updated to serve as the basis for freezing status data when the next trigger event occurs. The next emergency mission insertion request will then form the... When the event is triggered for the second time, its arrival time is defined as And return to step S2 to freeze. Real-time drone operation and task status, reconstructing event-driven rolling time windows Repeat steps S3 to S6 to achieve rolling allocation in continuous insertion scenarios.

[0300] In summary, this embodiment fully illustrates the execution process of the present invention in a single emergency task insertion scenario: First, the system state is frozen by triggering an event and an event-driven rolling time window is constructed; then, feasibility screening is performed based on payload capacity and energy accessibility to form a candidate UAV set; further, a bidding score model is constructed based on execution efficiency, distance, health status, task compatibility, and emergency task priority to output the bidding score of each candidate UAV for the emergency task; then, the bidding score is introduced into a multi-objective 0-1 planning allocation model to jointly solve the assignment-storage point selection-battery swapping decision under the constraints of uniqueness, battery swapping logic, energy closed loop, and time window; finally, scheduling instructions are generated based on the optimal solution and rolling updates are performed. As shown in Table 6, the scheme in this embodiment where U1 transits through D1 and does not require battery swapping is the comprehensive optimal scheme, which can balance the emergency task response speed and the optimal global cost of the scheme.

[0301] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for bidding and allocating emergency drone missions using an event-driven rolling time window, characterized in that, Includes the following steps: S1: Obtain the operational status data and task status data of all drones within the task area, establish drone sets, task sets and storage point sets, obtain the current location, current remaining energy, payload status and current task execution progress of each drone, and obtain the availability status and whether battery swapping is supported for each storage point; S2: When a request to insert an emergency task, i.e. a delivery request, is received, the request is defined as a trigger event and the arrival time of the request is defined as the trigger time. The running status data and task status data in step S1 are frozen based on the trigger time, an event-driven rolling time window is constructed, and the emergency task to be inserted and the regular task that has not been completed at the trigger time are combined into a task pool. S3: Based on the state data frozen in step S2, perform feasibility screening on the drone set: first, eliminate drones that do not meet the emergency mission payload requirements based on payload capacity, and then eliminate drones that cannot complete the entire emergency mission delivery process under the current state's remaining energy conditions based on energy accessibility, and retain feasible drones to form a candidate drone set. S4: For each candidate drone in the candidate drone set, construct a bidding score calculation model, that is, take the drone operation status data, task status data and path distance information frozen at the trigger time as input, and output the bidding score of each candidate drone for the emergency task. S5: Introduce the bidding score obtained in step S4 into a multi-objective 0-1 programming allocation model to solve for the optimal assignment scheme within the candidate UAV set; wherein, the multi-objective 0-1 programming allocation model takes "maximizing the emergency task bidding response score, maximizing the total scheme benefit, minimizing the total scheme execution cost, and minimizing the total task delay penalty" as the comprehensive objective function, and sets relevant variables and constraints in the model, and finally outputs the allocation result that takes into account both the emergency task response speed and the optimal global scheme cost; S6: Generate and issue scheduling instructions based on the allocation results obtained in step S5, so that the assigned UAV can execute scheduling instructions containing a preset action sequence; after execution, update the UAV's operating status data and task status data. When a new emergency task insertion request is received again, return to step S2 with a new trigger time to reconstruct the newly triggered event-driven rolling time window, and repeat steps S3 to S6 to achieve rolling allocation in continuous insertion scenarios.

2. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 1, characterized in that, The specific steps for step S1 are as follows: S11: Define a drone ensemble within the mission area. ,in Indicate the drone number; define the task set. ,in Indicates the task number; defines the set of storage points. ,in This indicates the storage point number. The above definition provides a unified description of the drone selection, bidding score calculation, and task allocation in the subsequent emergency task insertion process. S12: For the first A drone, obtain its current location coordinates. ,in , , They represent drones x-coordinate, y-coordinate, and altitude coordinates in the mission area coordinate system; current remaining energy. ,in Used to identify energy state and indicate remaining state; drone Maximum load ,in Used to indicate the capacity limit; each drone Execution task number ; S13: Obtain any storage point Available status and the storage location Does it support battery swapping? ,in ,when This indicates that the storage point is available, when This indicates that the storage location is unavailable; ,when This indicates that the storage location supports battery swapping. This indicates that the storage location does not support battery swapping; the aforementioned and The status data of the storage point is determined at the time of the triggering event and is used for subsequent storage point selection and battery swapping decisions.

3. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 2, characterized in that, The specific steps for step S2 are as follows: S21: When a request to insert an emergency task is received, the request is identified as the first... The event is triggered for the [number]th time, and the arrival time of the request is defined as the [number]th time. The time when the secondary trigger event occurs ,in Indicates the sequence number of the triggered event; S22: with Build an event-driven scrolling time window as the starting point of the time window ,satisfy: ； In the formula, Indicates the first The event-driven scrolling time window corresponding to the next triggered event; This represents a constant for the length of the time window, where Used to indicate the preset length of the time window; S23: with The operational and task status data obtained in step S1 are frozen at a reference time, ensuring that subsequent drone selection, bidding, and allocation are all based on data from the same moment. Furthermore, based on the frozen status data, the emergency tasks to be inserted are compiled into an emergency task set. ,in Used to identify the task type as an urgent task, and then the... and The task pool consists of routine tasks that the drones have not yet completed. The As an event-driven scrolling time window The task bidding and allocation targets within the organization.

4. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 3, characterized in that, The specific steps for step S3 are as follows: S31: At the triggering time Based on the frozen state data, for the task pool emergency mission Set up an emergency task The load requirement is ,in Used to identify urgent mission payload requirements, when Remove drones The remaining drones then proceed to energy reachability assessment; S32: For drones retained in step S31 Perform energy reachability determination; for any storage point in the storage point set. With emergency mission Define drone via storage point Execute emergency mission Total energy consumption throughout the entire delivery process for: ； In the formula, Indicates drone Fly from current location to storage point Energy consumption, of which Used to identify drones Fly from current location to storage point The movement phase; Indicates from storage point Head to an emergency mission And the energy consumption for completing the delivery, of which Used to identify from the storage point Head to an emergency mission The delivery execution phase; Indicates completion of an emergency mission Afterwards, return to the storage point The energy consumption of returning to base, of which Used to identify the completion of an urgent task Afterwards, return to the storage point The return phase; The energy consumption of each of the above is determined based on the frozen location information, path distance and UAV energy consumption at the time of the triggering event, and serves as the basis for subsequent energy hard constraint screening calculations; S33: Setting up drones The minimum retention energy threshold is drones The minimum energy required to maintain flight, of which Used to identify the minimum energy required for the drone; store the following hard energy constraints: When the drone does not require battery swapping, the following requirements must be met: ； When the drone exchanges batteries at a storage location that supports battery swapping, let the full battery capacity be... ,in The slow-point energy used to identify drones must meet the following requirements: ； Based on this, we retain the UAVs that satisfy the above-mentioned energy hard constraint conditions, forming the first... The set of candidate drones corresponding to the triggering event ,in: 。 5. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 4, characterized in that, In step S4, the bidding score is a comprehensive score used to characterize the suitability of candidate drones for performing emergency tasks. The comprehensive score consists of the drone's execution efficiency score, distance score, health status score, task compatibility score, and emergency task priority score. Each of the above scores is normalized to obtain a normalized score, and the normalized scores are weighted and fused according to the weights of the user and decision-maker for each score to obtain the bidding score, satisfying the constraint that it is non-negative and the sum of the weights is 1. The bidding score and its corresponding normalized score for each candidate drone are output as one of the input parameters for introducing the bidding score and solving the allocation in step S5.

6. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 5, characterized in that, In step S4, the bidding score calculation model performs the following operations: S41: In the obtained... The set of candidate drones corresponding to the triggering event Inside, for each candidate drone Calculate its impact on emergency tasks Bidding score ,satisfy: ； ； In the formula, the bidding score Indicates drone For emergency missions The degree of compatibility; Indicates the first The weights of each scoring indicator are weight parameters pre-defined by decision-makers and users; Indicates drone For emergency missions The Normalized scores for each scoring indicator Corresponding to drones For emergency missions Execution efficiency score Distance rating Health status score Task compatibility score Priority rating for urgent tasks ; Normalized score s corresponding to the execution efficiency score satisfy: ； In the formula, Indicates drone Expected to complete the emergency mission Response completion time; Represents the maximum response completion time constant used for normalization, derived from the first... The set of candidate drones corresponding to the triggering event The drone that takes the longest time to complete an emergency mission is identified, representing the longest time a drone in the candidate drone set takes to complete the emergency mission. Normalized score corresponding to distance score satisfy: ； Among them, the shortest path length index satisfy: ； In the formula, Indicates drone Execute emergency mission At that time, gather at the storage point. The shortest path length index corresponding to the selection of the optimal storage point; Indicates drone to storage point distance, Indicates storage point Distance to emergency mission C; Normalized score corresponding to health status score satisfy: ； In the formula, Indicates drone The health risk rate, with a range of values ​​being [value missing]. , indicating drone The probability of failure; the lower the risk, the lower the probability of failure. The larger; Normalized score corresponding to task compatibility score satisfy: ； In the formula, Indicates the time when the drone triggered the event. Routine tasks currently in progress; Indicates routine tasks With emergency mission The conflict degree function value is determined by factors such as the type of goods and the degree of overlap between the original task and the emergency task. The greater the conflict degree, the lower the score. Normalized score corresponding to emergency task priority score satisfy: ； In the formula, Indicates an emergency task Priority parameters; Indicates drone Current routine tasks The priority parameter; the normalized score This is used to characterize the priority of an emergency task relative to the routine tasks currently being performed by the drone; when the priority of an emergency task is higher than that of the routine task, the normalized score is correspondingly higher. S42: Obtain each candidate drone according to step S41. For emergency missions Bidding score and output the and its corresponding normalized score As part of step S5, when establishing the multi-objective 0-1 planning allocation model, the "candidate UAV set" One of the input parameters within the range.

7. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 6, characterized in that, In step S5, at the The set of candidate drones corresponding to the triggering event Emergency Task Collection Set of storage points The following decision variables are established: a. Assigning variables: ； in, Indicates candidate drones Are you assigned to perform an emergency mission? ,when Indicates assignment; otherwise, no assignment is made. (Variable) , ; b. Variable selection for storage point: ； in, Indicates drone Execute emergency mission Choose storage location As a transit storage point, variables ; c. Enable variables at storage points: ； in, Indicates drone Enable storage point Store or temporarily save its current routine tasks; d. Decision variables for battery swapping: ； in, Indicates drone Choosing a storage location Execute emergency mission Battery swapping is performed before, and vice versa. Then battery swapping will not be performed; similarly Indicates drone Complete the emergency mission Return to storage point Then, to continue the original task, battery swapping was performed; otherwise... Then battery swapping will not be performed, among which Used to identify the battery swapping phase when returning to continue the original task; e. Timeout variable: ； in, This indicates the cumulative timeout amount incurred by subsequent regular tasks after an emergency task was inserted; Indicates an emergency task The timeout amount.

8. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 7, characterized in that, In step S5, the construction and solution of the comprehensive objective function are as follows: a. Constructing the objective function for the bidding response This is used to directly embed the bidding score obtained in step S4 into the optimized allocation, so that the assignment result tends to select the drone with the higher bidding score and more suitable for performing the emergency mission. The bidding response objective satisfies: ； In the formula, For the candidate drones obtained in step S4 For emergency missions The bidding score; b. Construct the objective function for the project's revenue. This is used to characterize the benefits of completing urgent tasks, ensuring that the model prioritizes the effective execution of high-reward tasks, provided it is feasible. The benefit objective of the proposed solution satisfies: ； In the formula, Indicates an emergency task The benefit parameters are determined by the priority level of the emergency task when it is inserted; c. Construct the objective function for the total cost of the solution. This solution comprehensively reflects the energy consumption costs incurred by candidate UAVs when performing emergency missions, the costs of transit storage and handling at storage points, and the costs of battery swapping, aiming to minimize costs while completing emergency missions. ； Total cost target of the plan By enabling storage point Storage and handling costs Total cost of a single item Composition, in which , These are used to identify the costs incurred in storage and handling, as well as the total cost of each item: To enable storage points The cost parameters for storage and loading / unloading are obtained from the storage point in step S1. Given a state; Indicates "candidate drones" Select storage location Execute emergency mission The corresponding total cost per item includes energy consumption converted costs. Total cost of battery swapping , The calculation and definition are as follows: ； In the formula, energy consumption is converted into cost. The conversion factor based on unit energy consumption With drones via storage point Execute emergency mission Energy consumption throughout the entire process The sum of the products determines the superscript. Used to indicate the cost calculated based on unit energy consumption; total cost of battery swapping. The cost of a single battery swap With drones In carrying out emergency missions At the storage point Should the battery be replaced? And to complete emergency missions Return to storage point To continue the original mission Should the battery be replaced? It was jointly determined that, among which This is used to indicate that the cost is incurred by battery swapping; d. Construct the objective function for task delay penalty This is used to suppress situations where urgent tasks time out, enabling the model to detect delays caused to the original tasks by the insertion of urgent tasks. and drones Exceeding the deadline for completing urgent tasks "A penalty is given, and the objective of the task delay penalty is satisfied as follows:" ； In the formula, , They represent emergency missions. Insert for regular tasks The relationship between the resulting delay penalty and the severity coefficient of the delay penalty for completing an urgent task is as follows: This indicates that the penalty for delays in emergency tasks is greater than the penalty for delays in regular tasks. Among them, the total delay penalty caused by the insertion of an emergency task to subsequent regular tasks. Determine as follows: Assume an unmanned aerial vehicle (UAV) When no urgent task is inserted, its subsequent regular tasks The expected completion times according to the established sequence are ,Task The deadline is ,in Used to identify the deadline for urgent tasks; when an urgent task is inserted... Subsequently, regarding the drone at the time of insertion. An unfinished set of follow-up routine tasks Incremental insertion time If this causes the overall postponement of subsequent routine tasks, then the timeout amounts of the subsequent routine tasks are summed and the non-negative portion is taken, defined as: ； The extra time inserted According to drones The insertion procedure for executing an emergency mission has been determined, and the insertion procedure includes: the drone moving from its current location to the storage point. Save the current task and proceed from the save point to perform the emergency task. Return to the original task save point And continue to complete the original delivery task; the insertion of an emergency task results in additional time for the original delivery task. satisfy: ； In the formula, For drones Select storage location And the total time to complete the emergency task inserted into the delivery process; The emergency task delay penalty Determine as follows: Assign an emergency task The latest completion time is When drones Assigned to perform emergency tasks ,Right now At that time, the estimated completion time for its emergency mission is ,like Then no delay will occur, if This results in a delay; the emergency task delay penalty is defined as the non-negative portion of the timeout amount by which the emergency task completion time exceeds the time limit. ； in, The insertion process time is obtained, and it satisfies: ； In the formula, For drones The current moment when the decision to insert the order is made; For drones via storage point Complete the emergency mission Total time spent; e. Quantize the above objective functions, since... , , , The units of measurement may differ and the numerical scales may also differ. Min-max dimensionality is used to normalize each objective function to an interval. ,use Represents any objective function, for any objective function Given upper and lower bounds and The normalized objective function satisfies: ； In the formula, Describe the objective function Dimensional results; As the target identifier, respectively corresponding to , , , ; and These represent the candidate range for the currently triggered event. , , The lower and upper bounds within the candidate drone set are defined before each insertion. Emergency Task Collection Set of storage points Calculated within the candidate range; To prevent positive constants with a denominator of zero; f. Construct a dimensionalized comprehensive objective function and solve for its maximization: ； In the formula, This represents the dimensionalized comprehensive objective function; , , , They are respectively , , , Dimensional results; Configure the settings in advance before the event is triggered.

9. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 8, characterized in that, In step S5, the following constraints are applied, and the decision results are solved and output: a. Apply a unique task assignment constraint to ensure that each emergency task is uniquely assigned to a single candidate drone under this triggering event: ； The above formula represents the expression for each emergency task. The sum of the assignment variables for all candidate drones must equal 1, and this constraint applies to all... All are true; b. Apply a drone mission capacity constraint to ensure that each candidate drone undertakes at most one emergency mission in the solution of this triggered event. The constraint satisfies: ； The constraint applies to all All are true; c. Apply storage point selection and allocation constraints to ensure that "only drones assigned to perform emergency missions are allowed to select storage points," satisfying: ； The constraint applies to all , All are true; d. Apply storage point path logic constraints to establish a consistent relationship between "whether to execute an emergency task through a certain storage point" and "whether to use the storage point to store a regular task," and limit each drone to using at most one storage point, satisfying: ； ； Ensure the selection of storage location During transit The storage point has been enabled. And it limits each drone to using a maximum of one storage location; this constraint applies to all... , , All are true; e. Apply battery swapping logic constraints to ensure that: only when the candidate drone... In carrying out emergency missions When the storage point was selected As a transit and storage point Furthermore, this storage location supports battery swapping. Only when this condition is met is battery swapping permitted at the storage location. or The battery swapping logic constraints include battery swapping constraints before executing an emergency task and battery swapping constraints before returning to the storage point to continue the original task after completing the emergency task, as detailed below: 1) Battery swapping decisions before carrying out emergency missions This is used to constrain whether candidate drones can be stored at a designated location before heading to perform an emergency mission. Battery swapping is performed to meet the following requirements: ； The above constraint means that if the candidate drone No storage location selected As an emergency mission The transit storage point, or the storage point Battery swapping is not supported. Then the battery swapping decision variables before executing the emergency task The value can only be 0, which means that battery swapping is not allowed at this storage point; ； The above constraint means that the storage point is selected if and only if the storage point supports battery swapping. Battery swapping decision variables before executing an emergency mission Only a value of 1 is allowed, corresponding to the battery swapping constraint before executing an emergency mission for all candidate drones. Emergency mission and storage points All are true; 2) Decisions regarding battery swapping before returning to the storage point to continue the original task after completing an emergency mission. This is used to constrain candidate drones to complete emergency missions and return to their storage locations. Afterwards, before continuing the original routine task, can the battery be swapped at this storage point, satisfying the following conditions: ； The above constraint means that if the candidate drone No storage location selected As an emergency mission The transit storage point, or the storage point Battery swapping is not supported. Then the returned battery swapping decision variables The value can only be 0, which means that battery swapping is not allowed at this storage point; ； The above constraint means that the storage point is selected if and only if the storage point supports battery swapping. The returned battery swapping decision variables Only a value of 1 is allowed, corresponding to the battery swapping constraint before returning to the storage point to continue the original task after completing the emergency mission, for all candidate drones. Emergency mission and storage points All are true; f. To ensure that when candidate drones are available... Assigned to perform emergency tasks And choose a storage location As a transit and storage point At that time, it was from the storage point Depart to complete the emergency delivery and return to the storage point. Energy consumption shall not exceed the available energy threshold. The following energy constraints shall be imposed for emergency mission delivery: ； Wherein, the constraint applies to all , , All are true. A pre-set large constant is used in The constraint is relaxed at this time; the energy constraint for emergency mission delivery is determined by the intermediate variable of energy consumption for the round trip of the emergency mission delivery. ,in This is used to identify the intermediate variable between "delivery phase of emergency delivery mission" and "available energy threshold". , "Combined composition, its definition is as follows:" Indicates from storage point Depart to complete the emergency mission Delivery energy consumption Add return storage point energy consumption The sum of them satisfies: ； Indicates drone Without swapping batteries, the remaining energy in the current state of the drone... Arrival at storage point Consume energy And deduct the safe energy threshold The available energy threshold that can then be used to continue executing the task satisfies: ； This represents the increase in available energy relative to the threshold of not swapping batteries when swapping them at the storage location; in other words, the battery capacity after swapping. Plus storage point Consume energy And offset the remaining energy of the current drone state. The energy, satisfying: ； g. Apply energy reachability constraints to the storage point, since the battery swapping occurs at the storage point. Regardless of whether the battery is swapped or not, the drone must be guaranteed to function properly. Can reach the storage point using the current remaining energy. The energy reachability constraints to the storage point are applied as follows: ； Wherein, the constraint applies to all , , Both are true, when At that time, the above formula is used to limit the energy consumption of movement to the storage point to not exceed "remaining energy". Deducting the safe energy threshold The threshold of "". h. To ensure the drone Complete the emergency mission Return to storage point Remaining usable energy : ； It can satisfy the need to continue executing the temporarily stored subsequent regular tasks, impose energy constraints on continuing the original tasks, and determine the remaining available energy after returning to the storage point after completing the emergency task. With drones Minimum available energy required to complete subsequent routine tasks Divided into "no battery replacement after return" "Return and swap batteries" Two scenarios: No battery swapping will be performed upon return. At that time, the remaining available energy is required. Meet the needs of drones Minimum available energy required to complete subsequent routine tasks : ； The above constraints are used to characterize whether the drone still has the energy feasibility to continue performing its original conventional mission without battery swapping. This applies to all... , , All are true; Battery swapping will be performed upon return. At that time, the required battery capacity after battery swapping is... Deducting the safe energy threshold Post-satisfaction drones Minimum available energy required to complete subsequent routine tasks : ； The above constraints characterize whether the UAV has the energy feasibility to continue performing its original conventional mission under battery swapping conditions. For all... , , All are true; i. Apply an urgent task time limit to ensure that urgent tasks are completed before the deadline, satisfying the following: ； In the formula, Indicates drone Select storage location Execute emergency mission Delivery time; Indicates storage point Battery swapping service time; Indicates an emergency task The deadline, given by the urgent task insertion request, applies to all... , , All are true; j. Solve and output the decision results. The solution is used to obtain the optimal allocation scheme, i.e., the scheduling instruction set, under the combined effect of the above comprehensive objective function and constraints, and outputs the emergency task assignment results, storage point selection results, and battery swapping decision results. The emergency task assignment results are determined by... The value of is determined, and the storage point selection result is determined by . and The value of is determined, and the battery swapping decision is made by . , The value of is determined.

10. The event-driven rolling time window method for bidding and allocating emergency drone missions according to claim 9, characterized in that, The specific steps for step S6 are as follows: S61: In the The time when the secondary trigger event occurs Based on the solution result of step S5, a scheduling instruction is generated and issued to the assigned drone. Insert the arrival time of the request for urgent tasks; S62: After receiving the dispatch instruction, the assigned UAV executes a sequence of actions to insert a task. The sequence of actions includes: interrupting the current routine task, going to the storage point to perform storage, making a battery swap decision, going to the emergency task point to complete the emergency task, and returning to the storage point to resume the execution of the original task. The sequence of actions is determined by the dispatch instruction output at the trigger time step S5. The storage action of the storage point in the insertion action sequence satisfies: if and only if there exists Make At that time, drones Go to the corresponding storage point Perform regular tasks and store data. The solution is obtained from step S5; The battery swapping action in the insertion action sequence satisfies: if and only if there exists Make At that time, drones At the corresponding storage point Battery swapping is performed; among which The result is obtained from step S5, and it is determined whether the storage point supports battery swapping. The storage point status data obtained in step S1 is given; The emergency task execution action in the order insertion action sequence satisfies: if and only if there is an emergency task. Make At that time, drones Head to the emergency mission point and complete the emergency mission. ;in The urgent task set is obtained from step S5. Given by the emergency task insertion request; If and only if there exists Make At that time, drones After completing the urgent mission Return to the corresponding storage point A battery swap is then performed to allow the temporarily stored subsequent routine tasks to continue. The result is obtained from step S5, and whether the storage point supports battery swapping is also determined by whether it supports battery swapping status. Given; S63: After the insertion action sequence is completed, update the status data of the UAV and the storage point, including the current position coordinates of the UAV. Current remaining energy And the current task execution progress, as the basis for freezing state data when the next event is triggered; S64: When the next emergency task insertion request arrives, the first... When the event is triggered for the second time, its arrival time is defined as ,in Calculated from the trigger time; and the freeze time in step S2 is returned. The operation and mission status of drones are reconstructed, and the event-driven rolling time window is restructured. And repeat steps S3 to S6; wherein: ； In the formula, Indicates the first The event-driven scrolling time window corresponding to the next triggered event.

Images