The design of Inverse pseudo-random binary sequence (IPRBS) for the plant identification is presented. There exist two problems, the burden of identification experimental working amount based on the frequency response method and the identification precision sensitivity to the input-excitation signal based on Recursive Least Square with forgetting factor (RFF). We solve the problems using a novel identification method, which combines RFF algorithm through frequency response analysis. The nominal model parameters are identified on the basis of frequency response analysis, in which the frequency bandwidth of harmonic input is limited, and frequency resolution is coarse. Based on this, the shift-impulse cycle of IPRBS as identification input can be calculated with the iteration-optimum method, and the model parameters can also be identified optimally in the procedure. The method is applied to the design of motion control system based on disturbance observer (DOB) of SCARA-type manipulator using Windows NT and VenturCom''s real-time extension module (RTX) environment.
展开▼
