Combining neural network and PID control, a PID control strategy based on diagonal recursive neural network tuning is proposed and applied to the control of AC servo system. The parameters of PID controller are adjusted online by diagonal recursive neural network, so that the static and dynamic performance indicators of the system are relatively ideal. Experimental results show that the AC servo system controlled by PID based on diagonal recursive neural network tuning has the characteristics of fast response speed, high steady-state accuracy and strong robustness. AC servo system is a nonlinear system with time-varying parameters, strong coupling and multivariable. The design of its controller directly affects the operating state of the servo motor, thus determining the performance of the whole system to a large extent. Traditionally, PID control is mostly used. This is because the structure of PID controller is simple. For the occasions where the structure and parameters of the controlled object are known, good control effect can be obtained by adjusting the controller parameters, but its parameters are not easy to adjust online. Therefore, it is often difficult to obtain satisfactory control effect for some complex systems with slow time-varying parameters and random interference. In recent years, with the development of computer technology, people use artificial intelligence methods to store the adjustment experience of operators as knowledge in computers. According to the actual situation on site, the computer can automatically adjust the PID parameters. In this way, intelligent PID controllers have emerged and have been successfully applied in actual industrial control. Most adaptive control schemes based on neural networks use multi-layer feedforward neural networks. Feedforward neural networks are static networks. However, in the processing of AC motor servo systems, it is necessary to introduce time-delay links to describe the dynamic characteristics of the system, but this requires a large number of neurons to represent the dynamic response. The dynamic recursive network uses the internal state feedback of the network to describe the nonlinear dynamic characteristics of the system, which can more directly reflect the dynamic characteristics of the system. Therefore, it is more suitable for the control problems of dynamic systems than the forward neural network. The diagonal recursive neural network has the characteristics of the general dynamic network that is easy to handle dynamic nonlinear problems, and has the advantages of simple structure and easy construction of training algorithms. Therefore, this paper uses the diagonal recursive neural network to adjust the parameters of PID control, and the simulation results prove the effectiveness of the control scheme.
You Might Like
Recommended ContentMore
Open source project More
Popular Components
Searched by Users
Just Take a LookMore
Trending Downloads
Trending ArticlesMore