Unmanned helicopter is a nonlinear and strongly coupling complex system.There are some problems when using the traditional PID method,such as low control precision,poor robustness and so on.Aiming at these problems,an adaptive neural network PID control method,with the ability of tuning control parameter intelligently,is proposed and it is easy to implement on the practical project.Firstly,the dual closed-loop controller is designed based on the idea of time-division separation.The outer loop is the path-following controller and the inner one is the attitude-stabilizing controller.Then,taking the real-time forward velocity as the input of the neural network,the optimal control parameters and the control reference point are obtained,which makes the controller performance be optimal.Finally,the numerical simulation of the route is carried out,by using the unmanned helicopter nonlinear model,the effectiveness and superiority are verified.In the route simulation,the attitude response of the unmanned helicopter is smoother and more stable,and the path tracking is more accurate when using the adaptive control law.The simulation result shows that the improved neural network PID controller overcomes the shortcomings of poor nonlinear control ability of the traditional PID controller,improves the control accuracy and speeds up the response and has stronger robustness and intelligent.%无人直升机是一种非线性、强耦合的复杂系统,使用传统的PID进行控制,存在着控制精度低,鲁棒性差等问题.针对上述问题,设计了一种具有控制参数智能整定功能且在工程上易于实现的自适应神经网络PID控制方法.首先,基于时标分离思想设计了双回路闭环控制器,外回路为路径跟踪控制器,内回路为姿态稳定控制器;然后,利用神经网络,以系统实时前飞速度信息作为输入,实时输出最优控制参数与控制基准点,从而实现控制性能最优化.最后,利用无人直升机非线性模型进行航线仿真,验证了自适应控制方法的有效性与优越性.航线仿真中,使用自适应控制律的无人直升机姿态控制响应更加平滑稳定,路径跟踪更为准确.仿真结果表明,改进后的神经网络PID控制方法克服了传统PID控制方法非线性控制能力差的缺点,提高了控制精度,加快了响应速度,具有更强的鲁棒性与智能性.
展开▼
