摘要
针对总线条件下具有双通道多包传输的短时变时延无人机(Unmanned Aerial Vehicle,UAV)发动机分布式控制系统,提出一种基于改进蝙蝠优化算法(Bat Algorithm,BA)的发动机故障检测观测器优化设计方法。将双通道多包传输转化为切换系统,短时变时延视为一种系统不确定性。把残差信号对噪声信号和故障信号的传递函数比值作为适应度函数,通过改进蝙蝠算法对该适应度函数进行优化得到最优观测器增益矩阵。在网络控制系统半物理平台上进行仿真,仿真结果表明,该方法能够同时抑制噪声信号和放大故障信号,从而提高故障诊断的正确率,降低虚警率。
Aiming at the engine distributed control system of the unmanned aerial vehicle(UAV)with dual-channel and multiple packet transmission under the bus condition,we propose an optimization design method of the engine fault detection observer based on the improved bat algorithm(BA).The dual-channel multiple packet transmission was transformed into a switched system,and the short-term time-varying delay was considered as a kind of system uncertainty.The ratio of transfer function of residual signal to noise signal and fault signal was taken as fitness function,and the optimal observer gain matrix was obtained by improved bat algorithm to optimize the fitness function.The method was simulated on the semi-physical platform of NCS.The simulation results show that our method can suppress the noise signal and amplify the fault signal at the same time,thus improving the accuracy of fault diagnosis and reducing the false alarm rate.
作者
王旭
路凯
肖玲斐
Wang Xu;Lu Kai;Xiao Lingfei(Xuchang Electrical Vocational College,Xuchang 461000,Henan,China;Xuchang College,Xuchang 461000,Henan,China;School of Energy and Dynamics,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2020年第10期249-255,共7页
Computer Applications and Software
基金
国家自然科学基金项目(51876089)。
关键词
无人机
分布式控制系统
蝙蝠算法
故障诊断
Unmanned aerial vehicle
Distributed control system
Bat algorithm
Fault diagnosis
