The Research Of Control Method For Multivariable System With Multiple Time Delays

Aiming at multiple input-multiple output controlled Process with time delays and coupling that is common phenomenon encountered in industrial production and difficult to control, especially for multivariable non-square systems with multiple time delays. In this paper, aiming at the properties of time delay, coupling and non-square that controlled process possess, a better control scheme is proposed based on the in-depth study of the history, current status and forward development of the controlled field related to controlled process. the coupled problems of non-square systems are better solved by using pseudo-inverse of non-square process model's steady-state gain matrix; the impact on control system of time delays are effectively compensation by using internal model control approach to design smith delay compensation controller. The designed procedure of decoupling controller is given in this paper. The proposed method not only retains the good performance and robustness of the internal model controller, but also dynamically compensates for shortcoming caused by the static decoupling and overcomes the impact on system performance of model error caused by model approximation; Meanwhile, the designed controller has good control performance and robustness for the uncertainties in process model parameters. Compared to the method proposed in literature two where by the pole placement method is used to design smith delay compensation controller, the design method proposed in this paper can significantly improve dynamic quality and the robustness of process which can be seen from simulation results. Moreover, the proposed method provides good performance when mismatch exits between model and actual process. Meanwhile, this paper quantitatively evaluates the design method using ISE (integral of square error) performance criterion.In addition, aiming at nonlinear process that is common phenomenon encountered in industrial production, the research of control effect of control scheme that the least square support vector machine inverse system and internal model control is used to control nonlinear process are carried out. Compared to the method proposed in literature seventy two where the neural network inverse system and internal model control is used to control nonlinear process, the design method proposed in this paper has better robustness stability and anti-interference performance which can be seen from simulation results. The study in this paper laid the foundation for the decoupling control of the multivariable systems.