Quantitative Analysis And Design Of The Decoupling Control Systems For Multivariable Processes With Multiple Time Delays

The research on advanced process control system continuously receives the domestic and foreign scholars'widespread attention, and decoupling control system is an important branch of advanced process control system. Although a lot of achievements in the development of decoupling control system have been achieved, the research on the decoupling control system design for multivariable processes with multiple time delays is still not perfect. Multivariable processes with multiple time delays inherit the complexities of delay-free multivariable processes, and furthermore, the time delays in processes made these complexities more challenging. The decoupling control methods developed for delay-free multivariable processes and the control methods developed for delayed single-input single-output (SISO) processes can not obtain satisfactory system performance when they are applied directly to multivariable processes with multiple time delays. It has significant theory value and practical significance to develop decoupling control schemes for multivariable processes with multiple time delays, which not only can provide excellent decoupling performance but also can be conveniently applied to practical application.Based on frequency-domain decoupling technology, and using multivariable control theory and robust control theory, this dissertation studied the problem of quantitatively designing the decoupling control systems for multivariable processes with multiple time delays. By analyzing the known information about controlled process and control structure, the negative effects of process interaction and time delays on the system performance are quantitatively depicted, and subsequently, several advanced decoupling control schemes are developed. The dominant merits of the proposed methods are that the design procedure is simple and the tuning of the controller is convenient. In order to achieve the effective implementation of the proposed control schemes, this dissertation also developed the conditions for evaluating the proposed systems'stabilities and the rules for tuning the adjustable parameters of the proposed controllers.The main contributions of this dissertation are as follows:1. For the unity feedback control system of multivariable processes with multiple time delays, the effects of the process's interaction and time delays on the system outputs are clearly depicted by using the concept of coupling matrix, then a new analytical decoupling PID controller design method is proposed. Besides, the intuitive conditions for evaluating the nominal system stability and the system robust stability are developed, and the rules for tuning the controller's adjustable parameters are proposed based on the analytical controller design procedure.2. For the unity feedback control system of multivariable processes with multiple time delays, the effective equivalent diagonal transfer matrices are developed for the controlled process by using the quantitative depiction of the interaction effects on the dominant diagonal subsystem of the controlled process that is formed based on the effective relative interaction analysis method, then the analytical design procedure of the decentralized PID controller are given, in which the frequency-dependent property of the interaction are utilized. Besides, this dissertation developed the conditions for evaluating the system stability and the rules for tuning the controller's adjustable parameters.3. Through adding an additional decentralized disturbance controller into conventional decoupling internal model control structure, a new two-degree-of-freedom decoupling internal model control structure is developed for multivariable processes with multiple time delays. The new structure realized the decoupling between the nominal system set-point tracking responses and the system disturbance rejection responses. The proposed novel disturbance controller design method can achieve remarkable improvement on the performance of disturbance rejection other than the decoupling system outputs. Besides, this dissertation analyzed and developed the conditions for evaluating the system stability.4. A new two-degree-of-freedom decoupling control scheme is proposed for integrating two-by-two systems with multiple time delays. Firstly, a novel inverse-model based decoupler is proposed. This decoupler can realize the effective decoupling between the set-point tracking responses of the nominal system and the system disturbance rejection responses. Subsequently, the analytical design procedures of the set-point tracking controller and the disturbance controller are provided based on the decoupled process. Moreover, the nominal system stability and the robust system stability are analyzed. The new structure can effectively overcome the negative effects of the integrating terms on system outputs, can realize the effective decoupling between the nominal system set-point tracking responses and the system disturbance rejection responses and can effective reject ramp-type disturbances.5. A new multi-degree-of-freedom decoupling control scheme is proposed for general non-self-regulating multivariable processes with multiple time delays. Firstly, the general design formulas of the controllers in this control structure are developed based on the parameterized formulas of the controller that can stabilize general multivariable processes with multiple time delays. Secondly, the detailed analytical design formulas of these controllers are derived from the point of view of optimizing the system performance. This control structure not only can realize the complete decoupling between the nominal system set-point tracking responses and the system disturbance rejection responses but also can tuning the dynamic performance of the disturbance rejection responses.