论文部分内容阅读
针对飞机舵机电液加载系统存在的多余力干扰不易抑制的问题,提出了结合橡胶-金属缓冲弹簧、飞机舵机位移指令前馈、基于动态RBF神经网络在线辨识的单神经元PID输出负载力反馈的复合控制器结构功能及控制策略。利用蚁群聚类算法优化的动态RBF神经网络对系统进行在线辨识,获得Jacobian信息,由单神经元PID控制器完成控制参数的在线自整定。仿真结果表明,该方法可以在实验室条件下对于模拟飞机舵机所受到的力载荷实现快速、准确的加载,能保证系统的稳定性且具有较强的鲁棒性。
Aiming at the problem that the surplus force of aircraft electro-hydraulic loading system is not easy to be restrained, this paper presents a single neuron PID output load force feedback based on dynamic RBF neural network with rubber-metal buffer spring and aircraft servo steering command feedforward Complex Controller Structure and Control Strategy. The dynamic RBF neural network optimized by ant colony clustering algorithm is used to identify the system online and acquire Jacobian information. The single neuron PID controller realizes online self-tuning of control parameters. The simulation results show that this method can load the simulated force of the aircraft steering gear quickly and accurately under laboratory conditions, and ensure the system stability and strong robustness.