流体中物体运动的PID控制方法的改进

An improved PID control algorithm to body movement in incompressible fluid

针对不可压缩流体中物体运动的特点,采用计算流体力学理论和重叠网格技术,改进了物体所受流体阻力的计算方法,利用物体周围流场的数值解,计算流体对物体运动的反作用力,在此基础上提出了借助数值解设计物体的运动控制系统的方法.该方法充分利用神经网络对非线性函数的逼近能力,通过在线学习调节PID控制器参数,实现了对物体运动的控制.对该方法进行了数值试验,试验结果验证了该方法的可行性和有效性.

To characterize a body movement in incompressible fluid, theory of computational fluid dynamics and technology of overlap grid are adopted to improve the method for calculating fluid resistance. This paper uses numerical solution of fluid field to attain anti-active-force to a body. And a method is proposed to design movement control system of a body based on numerical result. The control of body movement is realized by the approximating ability of neural network to nonlinear function and by adjusting PID controller coefficients of neural network on-line study. Simulation is conducted through this method, and results have proved its feasibility and effectiveness of this method.