Enhanced predictive control strategy for unstable parallel cascade processes with time delay

Abstract

The dynamics of runaway processes are unstable in nature and to control such processes with measured disturbances, an advanced regulatory control scheme such as cascade control is required. In this research, a control structure is suggested based on a generalized predictor scheme for controlling unstable processes in parallel cascade control architecture. The scheme is developed with two loops. An IMC controller is designed for the inner (secondary) loop and a PID controller with a lag compensator is designed for the outer (primary) loop. Along with a servo tracking controller, the primary loop consists of a generalized predictor structure having two filters capable of providing enhanced disturbance rejection and time delay compensation. The complexity associated with the suggested predictor is lesser on account of the fact that only a single tuning parameter needs to be tuned so as to attain a trade-off between regulation capabilities and performance under uncertainties. Simulation analysis is performed on different integrating as well as unstable parallel cascade processes. Obtained performances with the devised control strategy are compared with the existing methods. Quantitative evaluation in terms of the corresponding costs for output and input are carried out. The developed predictive strategy is found to deliver enhanced closed loop performances with a lesser control cost.

Keywords:

• Disturbance rejection
• generalized predictor
• parallel cascade process
• robustness
• time delay systems
• unstable systems