基于遗传算法的无人机飞控边缘性能评估方法研究

An Evaluation Method of UAV Flight Control Capability Based onEdge Performance Evaluation

摘要:

针对物理试验下固定翼无人机飞控能力评估效率低、成本高且存在安全风险的问题,开展了基于边缘性能评估的无人机飞行仿真测试和飞控能力评估方法研究。首先建立了无人机运动仿真模型和仿真环境,随后结合遗传算法搭建了飞行管理系统需求测试环境,实现了飞行系统批量测试用例生成,针对性提出了固定翼无人机俯仰姿态稳定性的测试判据和边缘性能评估方法,最后基于ROSflight进行了仿真试验,通过大量飞行测试,获取了飞机的飞行安全边缘,实现了俯仰姿态飞控能力分析评估。为无人机研发和应用阶段的仿真试验提供一定参考。

Abstract:

To address the issues of low efficiency, high costs, and safety risks associated with the evaluation of fixed-wing UAV flight control capabilities through physical testing, this paper investigates a UAV flight simulation testing method and an edge performance evaluation approach. Initially, a UAV motion simulation model and corresponding simulation environment were developed. Subsequently, a flight management system demand test environment was established by integrating genetic algorithms(GA) to facilitate the generation of batch test cases for the flight system. The criteria for testing and methods for evaluating pitch attitude stability in fixed-wing UAVs were proposed. Finally, simulations were conducted using ROSflight. Through extensive flight tests, insights into the aircraft"s safety margins were obtained while analyzing and assessing its pitch attitude control capabilities. This research provides valuable references for both the simulation testing phase in UAV development and its application stages.