CN-121657745-B - Post-disaster rescue task cooperative scheduling method considering unmanned aerial vehicle performance difference

CN121657745BCN 121657745 BCN121657745 BCN 121657745BCN-121657745-B

Abstract

The invention discloses a post-disaster rescue task collaborative scheduling method considering unmanned aerial vehicle performance difference, which comprises the steps of constructing unmanned aerial vehicle energy vectors and task demand vectors, calculating matching degree of unmanned aerial vehicles and tasks by means of weighted cosine similarity, establishing a bidirectional preference sequence of a task side and an unmanned aerial vehicle side, designing and introducing an improved Gale-shape stable matching algorithm to conduct multi-to-multi stable matching between the unmanned aerial vehicles and the tasks to obtain an initial scheduling scheme, introducing environment perception factors, rescheduling the initial scheduling scheme when rescheduling conditions are triggered due to dynamic environment changes, judging immediate switching or delayed switching based on smooth switching execution strategies, and distributing the unmanned aerial vehicles to execute all rescue tasks according to a final scheduling scheme. According to the invention, the performance difference of the unmanned aerial vehicle and the task demand can be automatically matched, the accuracy, the instantaneity and the stability of rescue task scheduling are greatly improved, and the method is suitable for an emergency rescue scene of cooperation of multiple unmanned aerial vehicles.

Inventors

HAN JIALI
Xie Naiming
SHEN YANG

Assignees

南京航空航天大学

Dates

Publication Date: 20260508
Application Date: 20260209

Claims (7)

1. The post-disaster rescue task cooperative scheduling method considering the performance difference of the unmanned aerial vehicle is characterized by comprising the following steps of: Carrying out multidimensional quantitative analysis on the performance difference of the unmanned aerial vehicle and the rescue task requirement to construct an unmanned aerial vehicle capability vector and a task requirement vector; Based on the unmanned aerial vehicle capability vector and the task demand vector, calculating the matching degree between the unmanned aerial vehicle and the rescue task, and generating an unmanned aerial vehicle side preference sequence and a task side preference sequence; Based on the unmanned aerial vehicle side preference sequence and the task side preference sequence, under the premise of considering unmanned aerial vehicle capability constraint, adopting an improved Gale-Shapley stable matching algorithm to perform many-to-many stable matching between the unmanned aerial vehicle and the rescue task, and obtaining an initial scheduling scheme, wherein the method specifically comprises the following steps: Initializing allocation states of all unmanned aerial vehicles and rescue tasks, and introducing an unmanned aerial vehicle side preference sequence of each unmanned aerial vehicle and a task side preference sequence of each rescue task; The method comprises the steps of starting request rounds, sending execution requests to the most favored unmanned aerial vehicles by each rescue task according to a task side preference sequence of each rescue task, and sequentially sending the execution requests to the unmanned aerial vehicles by the rescue tasks which need to be cooperated by the unmanned aerial vehicles according to the preference sequence until the cooperation quantity requirement is met; the method comprises the steps of selecting a single unmanned aerial vehicle from a plurality of unmanned aerial vehicles, selecting a plurality of unmanned aerial vehicles according to the unmanned aerial vehicle side preference sequence, and selecting the unmanned aerial vehicle side preference sequence according to the unmanned aerial vehicle side preference sequence, wherein the single unmanned aerial vehicle is allowed to execute a plurality of rescue tasks, and the single unmanned aerial vehicle is combined with unmanned aerial vehicle energy constraints to screen the first K execution requests which are preferred to be forward and meet the unmanned aerial vehicle energy constraints; The unmanned aerial vehicle makes a decision again after receiving the new execution request, namely, the rescue tasks corresponding to all the received execution requests are reordered, screened and accepted; Repeating the process of sending a request to the unmanned aerial vehicle decision by the rescue task until all the rescue tasks are accepted or the task side preference sequence is exhausted, stopping repeated operation to obtain an initial scheduling scheme; The unmanned aerial vehicle is controlled to execute a rescue task according to an initial scheduling scheme, the unmanned aerial vehicle state information and the environment state information are monitored in real time in the rescue task execution process, when the state information detected in real time meets a preset rescheduling triggering condition, the unmanned aerial vehicle energy vector, the task demand vector and the weight vector participating in calculating the matching degree are updated, the matching degree is recalculated, a Gale-shape stable matching algorithm is executed, a new scheduling scheme is generated, and the matching degree weight vector is updated specifically as follows: Defining a real-time environment influence factor vector, carrying out item-by-item weighting correction on an original weight vector which participates in calculating the matching degree between the unmanned aerial vehicle and the rescue task, and carrying out normalization processing to obtain an updated weight vector, wherein the environment influence factors comprise a cruising influence factor, a load influence factor, a perception capability influence factor of visibility, a communication capability influence factor of electromagnetic interference, a motor capability influence factor of wind speed and an environment adaptation capability influence factor of precipitation; when the scheduling scheme is changed, judging whether to immediately execute the scheduling switching or delay switching according to the smooth switching execution strategy, distributing the unmanned aerial vehicle to execute all rescue tasks according to the final scheduling scheme, and repeating the process until the tasks are all completed.
2. The post-disaster rescue task collaborative scheduling method considering performance differences of unmanned aerial vehicles according to claim 1, wherein the multi-dimensional quantitative analysis is performed on unmanned aerial vehicle performance and rescue task requirements, and the construction of unmanned aerial vehicle performance vectors and task requirement vectors is specifically as follows: The capacity of six dimensions of endurance capacity, loading capacity, perception capacity, communication capacity, maneuverability and environment adaptability is selected as a construction factor of the unmanned aerial vehicle capacity vector; The cruising ability is calculated according to the current residual electric quantity, the maximum flight duration and the unit energy consumption rate of the unmanned aerial vehicle; the load capacity is calculated according to the maximum load and the current residual load capacity of the unmanned aerial vehicle; the perception capability carries out weighted calculation according to the resolution ratio, the angle of view and the number and the type of the sensors carried by the unmanned aerial vehicle; The communication capacity carries out weighted calculation according to the communication bandwidth, the signal strength and the image transmission stability; The mechanical capacity is calculated according to the maximum flight speed, the climbing rate and the wind resistance level of the unmanned aerial vehicle; the environment adaptation capability is calculated according to the rainproof and dustproof capability of the unmanned aerial vehicle, the temperature adaptation range and the humidity of the flight environment where the unmanned aerial vehicle is located; And selecting the requirements of six dimensions, namely a endurance requirement, a load requirement, a perception requirement, a communication requirement, a mobility requirement and an environment adaptation requirement, as construction factors of task requirement vectors, and normalizing the requirements of all the dimensions to form the task requirement vectors.
3. The post-disaster rescue task collaborative scheduling method considering performance differences of unmanned aerial vehicles according to claim 1, wherein the matching degree between unmanned aerial vehicles and rescue tasks is calculated based on unmanned aerial vehicle capability vectors and task demand vectors, and the generation of unmanned aerial vehicle side preference sequences and task side preference sequences is specifically as follows: Calculating and weighting the similarity of the unmanned aerial vehicle energy vector and the task demand vector in each dimension by adopting a weighted cosine similarity function to obtain the matching degree between the unmanned aerial vehicle and the rescue task; For a single unmanned aerial vehicle, calculating the matching degree between the single unmanned aerial vehicle and all rescue tasks, and sequencing the rescue tasks from high to low according to the matching degree to obtain an unmanned aerial vehicle side preference sequence of the unmanned aerial vehicle; And for a single rescue task, calculating the matching degree between the single rescue task and all unmanned aerial vehicles, and sequencing the unmanned aerial vehicles from high to low according to the matching degree to obtain a task side preference sequence of the rescue task.
4. The post-disaster relief task collaborative scheduling method considering unmanned aerial vehicle performance differences according to claim 1, wherein the unmanned aerial vehicle capability constraints include unmanned aerial vehicle endurance constraints, unmanned aerial vehicle load capability constraints, task acceptance quantity constraints and execution sequence constraints, wherein: The unmanned aerial vehicle endurance constraint is that the total endurance requirement of a rescue task accepted by one unmanned aerial vehicle does not exceed the endurance of the unmanned aerial vehicle; The unmanned aerial vehicle load capacity constraint is that the total load requirement of a rescue task accepted by an unmanned aerial vehicle is not more than the load capacity of the unmanned aerial vehicle; the task acceptance quantity constraint is that one unmanned aerial vehicle is required to only accept a preset quantity of rescue tasks, and the rescue tasks which are in front in a preference sequence are reserved preferentially; the execution sequence constraint is that when one unmanned aerial vehicle executes a plurality of rescue tasks, the task execution sequence is ordered according to a voyage priority principle.
5. The post-disaster rescue task collaborative scheduling method considering performance differences of unmanned aerial vehicles according to claim 1, wherein the preset rescheduling triggering conditions specifically comprise: the real-time residual electric quantity of the unmanned aerial vehicle is lower than a preset safety threshold value; the real-time communication link quality of the unmanned aerial vehicle is lower than a preset stable communication threshold; The real-time wind speed of the environment where the unmanned plane is located exceeds the maximum allowable wind speed, the real-time precipitation exceeds the maximum allowable precipitation, and the real-time visibility is lower than the minimum allowable visibility; the unmanned aerial vehicle fails or the real-time health status metric is below a preset minimum health threshold.
6. The post-disaster rescue task cooperative scheduling method considering performance differences of unmanned aerial vehicles according to claim 1, wherein when the scheduling scheme is changed, judging whether to immediately execute scheduling switching or delay switching according to a smooth switching execution strategy is specifically: The smooth switching execution strategy carries out comprehensive evaluation on extra voyages, task interruption risks and residual execution time generated by continuously executing a current task and switching to a new scheduling scheme by constructing a switching cost function, delays execution of scheduling switching when the switching cost function value is larger than a preset cost threshold, and immediately executes scheduling switching when the calculation result of the switching cost function is smaller than or equal to the preset cost threshold.
7. The application system of the post-disaster rescue task collaborative scheduling method considering unmanned aerial vehicle performance difference according to claim 1, comprising: The unmanned aerial vehicle capability modeling module is used for quantifying the endurance capability, the load capability, the perception capability, the communication capability, the maneuverability and the environment adaptation capability of each unmanned aerial vehicle into multi-dimensional unmanned aerial vehicle capability vectors; The rescue task demand modeling module is used for quantifying the endurance demand, the load demand, the perception demand, the communication demand, the maneuvering demand and the environment adaptation demand of each rescue task into multi-dimensional task demand vectors; The matching degree calculation module is used for calculating the matching degree between the unmanned aerial vehicle and the rescue task based on the weighted similarity calculation method; The preference sequence construction module is used for respectively constructing a task side preference sequence and an unmanned aerial vehicle side preference sequence according to the matching degree; The stable matching scheduling module is used for generating an initial scheduling scheme between the unmanned aerial vehicle and the rescue task by an improved stable matching algorithm under the premise of considering the energy constraint of the unmanned aerial vehicle according to the task side preference sequence and the unmanned aerial vehicle side preference sequence; The dynamic scheduling module is used for monitoring the state information of the unmanned aerial vehicle and the environment state information in real time in the rescue task execution process, and dynamically updating the initial scheduling scheme when the preset triggering condition is met; And the switching control module is used for controlling the unmanned aerial vehicle to execute the current task or switch to a new scheduling task based on the smooth switching execution strategy when the scheduling scheme is updated.

Description

Post-disaster rescue task cooperative scheduling method considering unmanned aerial vehicle performance difference Technical Field The invention belongs to the technical field of unmanned aerial vehicle cluster cooperative scheduling, and particularly relates to a post-disaster rescue task cooperative scheduling method considering unmanned aerial vehicle performance differences. Background Along with the rapid development of unmanned aerial vehicle technology, the collaborative operation of multiple unmanned aerial vehicles is gradually applied to post-disaster emergency scenes such as flood control and disaster relief, earthquake rescue, forest fire monitoring and the like. Compared with the traditional ground rescue mode, the unmanned aerial vehicle has the advantages of rapid deployment, flexibility, wide visual field and the like, and can execute multi-type rescue tasks such as disaster investigation, communication relay, material delivery and the like in a complex environment. However, in practical applications, unmanned aerial vehicles participating in rescue are usually sourced from different models or different task groups, and have significant differences in the aspects of endurance, loading capacity, perception capacity, communication capacity, maneuvering performance, environmental adaptability and the like, and the unmanned aerial vehicle scheduling problem is more complicated due to multi-task parallelism and environmental dynamic change. The existing multi-unmanned aerial vehicle scheduling method is based on manual experience, static rules or heuristic allocation strategies, accurate quantitative description of unmanned aerial vehicle performance differences is difficult, and differentiated requirements of rescue tasks in different capacity dimensions are difficult to comprehensively describe. In a high dynamic scene of post-disaster rescue, environmental factors and unmanned aerial vehicle states are continuously changed along with time, the traditional scheduling method often lacks real-time adjustment capability, and problems such as capability mismatch, task interruption or resource waste are easy to occur. Therefore, a collaborative scheduling scheme capable of improving the accuracy of post-disaster rescue task scheduling of multiple unmanned aerial vehicles and simultaneously considering stability and overall execution efficiency is needed in the technical field of collaborative scheduling of unmanned aerial vehicle clusters. Disclosure of Invention Aiming at the problems in the prior art, the invention provides a post-disaster rescue task collaborative scheduling method considering unmanned aerial vehicle performance difference, which realizes accurate adaptation between an unmanned aerial vehicle and a task by carrying out vectorization modeling on unmanned aerial vehicle capability and task requirements, builds a scheduling mechanism based on matching degree, introduces an improved Gale-shape stable matching algorithm and a dynamic scheduling mechanism, updates a scheduling scheme in real time under environment and state change conditions, and effectively reduces task interruption risk by combining a smooth switching execution strategy. The invention aims to improve the accuracy, stability and overall execution efficiency of the rescue task scheduling after the disaster of the multi-unmanned aerial vehicle. In order to achieve the technical purpose, the invention provides the following technical scheme: A post-disaster rescue task cooperative scheduling method considering unmanned aerial vehicle performance difference specifically comprises the following steps: Carrying out multidimensional quantitative analysis on the performance difference of the unmanned aerial vehicle and the rescue task requirement to construct an unmanned aerial vehicle capability vector and a task requirement vector; Based on the unmanned aerial vehicle capability vector and the task demand vector, calculating the matching degree between the unmanned aerial vehicle and the rescue task, and generating an unmanned aerial vehicle side preference sequence and a task side preference sequence; Based on the unmanned aerial vehicle side preference sequence and the task side preference sequence, under the premise of considering unmanned aerial vehicle capability constraint, adopting an improved Gale-Shapley stable matching algorithm to perform many-to-many stable matching between the unmanned aerial vehicle and the rescue task, so as to obtain an initial scheduling scheme; When the state information detected in real time meets the preset rescheduling triggering condition, updating the unmanned aerial vehicle energy vector, the task demand vector and the weight vector participating in calculating the matching degree, recalculating the matching degree and executing an improved Gale-shape stable matching algorithm to generate a new scheduling scheme; When the scheduling scheme is changed, judging whether to immediately e