Abstract
In this paper, dynamic inversion method is used to decouple a multivariable system. Then by canceling traditional fuzzy quantification operation and based on analytical optimal correction function and on-line interpolation, two improved design methods for fuzzy controllers are provided. A class of nonlinear PID controllers is formed when traditional PI controllers are joined after the improved fuzzy controllers immediately. Both methods can essentially remove the system’s steady-state errors and flutter phenomenon, which are caused by the fuzzy quantification errors and regulating dead bands in the traditional fuzzy control system. Finally, by combining the improved fuzzy control methods with the dynamic inversion method, a kind of dynamic decoupling fuzzy integration control method for multivariable systems is constructed. The integration control system designed by this method possesses not only excellent decoupling results and better control performance, but also stronger adaptability to the parameter changes in the system. In addition, an example and its simulation results are also provided.
