Permanent Magnet Synchronous Motor Drive With Single Neuron PID Speed Controller

Recently, with the advances in permanent magnetic materials, semiconductor power devices manufacturing techniques and control theories, the permanent magnet synchronous motor (PMSM) drive obtaining popularity in motion control applications. PMSMs are widely used in high-performance servo applications because of its compact size, high air-gap flux density, and high-torque/inertia ratio. But, some machine parameters (for example rotor resistance) and drag load parameters (for example moment of inertia) always change with the different work conditions. All the control theories based on the classical control theory and modern control theory face the same problems, in which they all depend on the motor model, and when the system suffered the influences of parameters varying and disturbance, its performance will be much affected. So, it is an inevitable trend that introduces intelligent control theory into AC motor control, because it can get rid of the dependence of control plants' model and has feasibility and robustness for processing the uncertainty and inexact problems.PID control is widely adopted in industrial fields because of its simple structure, high reliability and easily implementation. Conventional PID controller has good control effect if the parameters of system model have not big variation, but when controlled plant has high nonlinear trait and high uncertainty, PID control can't achieve good result. Neural network, as one of modern information process technologies, has various advantages in many applications. Neural network control became a regarded research direction in control theory. Single neuron PID Control theory combines the advantages of Conventional PID control with neural network control and has the characteristics of simple configuration, easily regulation, parameter can self-study and self- adapt.In this paper, based on analyzing of neural network, single neuron PID control and its improved structure are introduced. And based on the modeling of PMSM, the single neuron PID speed controller with advanced Hebb supervising study and single neuron speed controller based on RBF neural network identification are studied and simulated by matlab. The results demonstrate that the single neuron PID speed controller is improved greatly in rapidness and robustness.