Design And Research Of The Motion Control Board And Control Algorithm Based On DSP

Motion Control is an important branch of modern automatic control technique. With its wide use, it plays a more and more important role in industrial production. Motion control system is developing towards generalization, intelligence and miniaturization along with the rapid development of micro-electronics and power electronic technology in recent years. Open Motion Control Architecture will be the main trend of motion control technology.On the other hand, advanced control technology depends on advanced control algorithm. Intelligent control technology represented by Neural Network gets extensive applications in control system because of its capabilities of the powerful nonlinear mapping and self-adaptive learning.In this paper, we chosen TMS320LF2407A, which is specially used in TI, as the kernel chip of motion control board. The specific implementing scheme of motion control system is designed. And the hardware structure and control algorithms are proposed and analyzed in detail, which include clock circuits, D/A converter circuits, CPLD circuits etc. For the control algorithm, the conventional PID control algorithm is expounded. So we present a new control algorithm with simplicity and high convergence rate, which combines single neuron PID with CMAC to control system. The single neuron PID is applied to accomplish the online self-adjustment of PID control parameters when environment changes. CMAC is used to control the inverse model of complex nonlinearity object. It achieves less operation complexity and faster convergence rate, and have no local minimal value. Finally, simulative results implemented by MATLAB show thatthe control algorithm integrating advantages of both single neuron P1D and CMAC has strong robustness and anti-disturbance ability.