基于神经网络PID控制的液压驱动Stewart平台稳定性研究

Study on stability of hydraulic driven stewart platform based on neural network PID control

Stewart平台运动轨迹容易受到外界波形的干扰,导致其运动轨迹输出误差较大,稳定性较差。对此,创建了液压驱动Stewart平台简图模型,推导了连杆动力学方程式。设计了液压驱动机构,给出了液压流量控制方程式。引用PID控制器并进行改进,设计了神经网络PID控制器。采用Matlab软件对Stewart平台两种控制方法进行仿真,将仿真结果进行对比和分析。结果显示:在无干扰环境中,两种控制器都能较好地实现Stewart平台运动轨迹的跟踪任务,差别不大;在有干扰环境中,采用PID控制器的Stewart平台运动轨迹输出误差较大,稳定性较差,而采用神经网络PID控制器,Stewart平台运动轨迹输出误差较小,稳定性较好。采用神经网络PID控制器,Stewart平台能够自适应调节控制参数,降低外界波形对平台运动轨迹的影响,提高Stewart平台运动的稳定性。

At present,the trajectory of Stewart platform is easily disturbed by external waveforms,which results in large error and poor stability of its trajectory output. To solve this problem,a sketch model of hydraulic drive Stewart platform is established,and the dynamic equation of connecting rod is deduced. The hydraulic driving mechanism is designed and the hydraulic flow control equation is given. The PID controller is introduced and improved,and the neural network PID controller is designed. The two control methods of Stewart platform are simulated with Matlab software,and the simulation results are compared and analyzed. The results show that the two kinds of controllers can achieve the tracking task of Stewart platform’s trajectory better in the non-interference environment,and there is little difference between them. In the interference environment,the output error of Stewart platform’s trajectory is larger and the stability is worse when using the PID controller,while the output error of Stewart platform’s trajectory is smaller and the stability is better when using the neural network PID controller. Using the neural network PID controller,the Stewart platform can adjust the control parameters adaptively,reduce the influence of external waveforms on the platform trajectory,and improve the stability of the Stewart platform motion.