Multiple models adaptive feedforward decoupling controller

• 出 处: Journal of Systems Engineering and Electronics
• 发表日期: 2005年04期
• 关键词: decoupling troller instant satisfying entirely posed switching varied matrices r

When the parameters of the system change abruptly, a new multivariable adaptive feedforward decoupling controller using multiple models is presented to improve the transient response. The system models are composed of multiple fixed models, one freerunning adaptive model and one reinitialized adaptive model. The fixed models are used to provide initial control to the process. The reinitialized adaptive model can be reinitialized as the selected model to improve the adaptation speed. The freerunning adaptive controller is added to guarantee the overall system stability. At each instant, the best system model is selected according to the switching index and the corresponding controller is designed. During the controller design, the interaction is viewed as the measurable disturbance and eliminated by the choice of the weighting polynomial matrix. It not only eliminates the steadystate error but also decouples the system dynamically. The global convergence is obtained and several simulation examples are presented to illustrate the effectiveness of the proposed controller.

When the parameters of the system change abruptly, a new multivariable adaptive feedforward decoupling controller using multiple models is presented to improve the transient response. The system models are composed of multiple fixed models, one freerunning adaptive model and one reinitialized adaptive model. The fixed models are used to provide initial control to the process. The reinitialized adaptive model can be reinitialized as the selected model to improve the adaptation speed. the freerunning adaptive controller is added to guarantee the overall system stability. According to the controller design, the interaction is viewed as the measurable disturbance and eliminated by the choice of the weighting polynomial matrix. It not only eliminates the steadystate error but also decouples the system dynamically. The global convergence is obtained and several simulati on examples are presented to illustrate the effectiveness of the proposed controller.