四旋翼无人机编队变换能耗优化仿真教学

Simulated teaching for energy consumption optimization in quadcopter unmanned aerial vehicle formation change

针对四旋翼无人机编队变换能耗优化仿真教学,该文以多无人机编队切换应用中存在无人机能耗不平衡的案例作为典型教学内容,提出了基于无人机编队能耗优化的仿真教学模式,设计了无人机集群编队队形变化能耗最优仿真教学总体方案,阐明各个无人机无碰撞地驶向目标点的实现思路。搭建的仿真教学实例有效地解决了无人机编队飞行时间短的问题。此设计已经用于研究生“无人机控制技术”的课程教学,可激发学生在传统算法上进行优化的创新思维。

[Objective]The course content for unmanned aerial vehicle(UAV)control technology is extensive and complex,encompassing a broad spectrum of theoretical knowledge from various disciplines.It involves strong mathematical logic relationships and many formula derivations.The traditional teaching method,however,has proven to be inadequate in cultivating students’innovative and critical thinking abilities,resulting in disappointing academic results.Therefore,the challenge and focus of this course lie in discovering an effective teaching strategy.One that encourages each student’s active participation and provides them with opportunities to demonstrate their understanding and skills.To cultivate the students’problem-solving abilities within the context of UAV control courses,it is imperative to stimulate their creative thinking.This can be achieved through research guidance on course design and by implementing simulation teaching research on UAV formation energy consumption optimization.[Methods]The concept of multidrone formation switching presents unique challenges.Existing solutions,such as the Hungarian algorithm,can solve the assignment problem of the optimal total switching distance assignment.However,these solutions often result in certain UAVs being assigned excessively long flight paths or individual UAVs being tasked with high climbs.This invariably leads to higher power consumption during flight than during hovering,causing a rapid decrease in power consumption during formation switching.The outcome is a shorter flight time for the entire formation compared to other UAVs.To this end,we have refined the allocation plan.Our goal is to ensure that the flight paths of the UAVs during the formation switching process are similar and the flight times to the target waypoints are consistent,thereby avoiding the above problems.In this context,we employ the particle swarm optimization(PSO)algorithm to design the flight distance and climbing distance costs for the UAVs.We set an appropriate objective function and solve the linear programming in each iteration to find the optimal assignment solution for the current position of each particle.The fitness of the partic le swarm in the iterative process is calculated,and the optimal solution is obtained by comparing the fitness of the particle swarm.After determining the optimal solution for UAV cluster formation switching,it is also necessary to ensure that each UAV reaches its respective target point safely,without collision.Therefore,during the formation switching,the task for each UAV can be decomposed into two subtasks:moving to the target point and avoiding collisions.However,this process can lead to multitask conflicts owing to potential collisions between UAVs.To effectively manage these conflicts,we introduce the null space behavior control algorithm.This algorithm ensures that each UAV avoids collisions while navigating to the target point,thus providing a comprehensive resolution for multitask conflicts.[Results]Experimental results from simulations indicate that under the traditional particle swarm algorithm,half of the UAVs remain static during multi-UAV formation.This leads to a significant disparity in energy consumption among the UAVs,with some consuming excessive energy.This phenomenon,known as the“barrel effect,”greatly diminishes the task execution efficiency of UAV clusters.By introducing an optimized particle swarm algorithm that balances overall flight energy consumption,we can mitigate these issues.Under this optimization,the entire UAV formation will shift slightly,and multiple UAVs will prioritize target points reachable through descending motion.In addition,the importance of safety measures during task execution is evident when comparing different behavioral control frameworks.Without collision avoidance tasks,the distances between drones exceed the safe limit,jeopardizing task execution safety.However,when collision avoidance tasks are implemented within the behavioral control framework,the drones maintain a safe distance from each other.This ensures safety in dynamic formation switching and achieves the goal of preventing collisions between UAVs.[Conclusions]Considering the unique characteristics and teaching needs of UAV control technology courses,we have conducted simulation teaching research.This approach specifically addresses the energy optimization issue in UAV formation switching.We have designed a solution for the linear assignment problem and improved the objective function of the PSO algorithm.These enhancements effectively address the energy consumption challenge during UAV formation switching,thereby extending the flight time of the formation.Additionally,we have employed a behavioral control method to ensure the simultaneous arrival of the UAV cluster at the destination.This method,which includes reasonable task output based on each UAV’s path in the planned formation reconstruction,also enables collision avoidance.MATLAB numerical simulations were performed to compare the performance of different solutions.This design is incorporated into the teaching of graduate-level UAV control technology courses.By integrating theoretical analysis with engineering application experiments,we stimulate innovative thinking among students and heighten their interest in subject learning.