A Neural Network Adaptive Control Scheme

ZHANG Sen,HAN Zhi-tao,ZHANG Hua-guang
DOI: https://doi.org/10.3321/j.issn:1005-3026.2000.02.004
2000-01-01
Abstract:For unknown nonlinear uncertain systems, a new adaptive control scheme based on neural network was proposed. It only needs the forward model of the plant. In this method, the neural network was combined with optimum method to solve the control law. It makes the output of plants follow the desired output very well. Control algorithm was deduced for SISO systems and MIMO systems respectively. The optimum method is gradient and Gauss-Newton method respectively. It can follow desired output precisely and has small overshoot. Its response is faster and has no steady error. The effectiveness of these algorithms was verified by simulation.
What problem does this paper attempt to address?