Absorption Principle in Process Control Applications

The broad class of industrial processes has similar dynamical behavior that may be described by simple mathematical models with the dead time. The most popular, a very effective and usual structure with a long dead time compensator in use today, is the Smith predictor. However, during the last 20 years, three principal problems of the Smith predictor controlling structures have been analyzed by many authors: (1) the robustness, (2) the disturbance rejection possibilities, and (3) the extension of the idea of the Smith predictor to the case of integrative plants. Furthermore, in order to effectively use control in industrial applications, simple tuning procedures must be developed. The mentioned problems may be solved more successfully than before by use of internal model principle and control together (IMPACT) structure. In this paper, the previous modification of the Smith predictor based on the IMPACT structure is improved and generalized for process control applications with the long dead time. The crucial part of the structure synthesis is implementation of the absorption principle that is derived and implemented in the general case of the continuous SISO systems with the dead time. The structure enables the extraction of the known class immeasurable disturbances and easy setting of controller parameters in order to achieve robust stability and performance. Both analytical analysis and simulation results show that tuning of the proposed structures is extremely simple due to relatively small number of tuning parameters, all having clear physical meanings.