A neurocontroller with adaptive static state decoupling for multivariable systems

The neurocontroller with adaptive static state decoupling for multivariable systems is proposed in this paper. In this new intelligent control system, a recursive least squares method with a changeable forgetting factor is used to obtain the parameters of the low-order model of the multivariable system. The multivariable system is decoupled statically, and then the neurocontroller is used in each input-output path to control the decoupling multivariable system. The simulation test results show that good performance, strong robustness and adaptability are obtained.