Research on intelligent controller design for MIMO spatially -Distributed systems with applications
Spatially dynamic distributed systems have been attracting increasing attention from researchers in the field of system modelling and control since their introduction as an alternative to simple systems to meet the ever-greater requirements to make industrial systems more precise and energy-efficient and to overcome process complexities. An approach whereby complex systems with multi-dimensional parameters, inputs or outputs are simply disregarded or simplified with the help of convenient mathematical models is no longer feasible. Therefore, the purpose of the present study is to contribute to the advancement of both theoretical and empirical knowledge in this field through the means of theoretical analysis, application simulations and case studies. From a theoretical perspective, this study focuses primarily on the design methodology of control systems. To this end, the first step is identification of requirements from the applications, followed by the implementation of an original approach underpinned by data prediction for type-2 T-S fuzzy control with the purpose of making the control system design more convenient. With this aim in mind, the study creates an interface/platform to link or anticipate spatially dynamic distributed system output from lumped system data by taking advantage of the threedimensional character of type-2 fuzzy sets. Moreover, on the basis of a decoupled spatially dynamic distributed system, this study applies Mamdani-type and interval type-2 T-S type fuzzy control, and extends a discussion about the results of simulation and analysis. With regard to application examination, the study contributes to primarily with system analysis and modelling. Along with the progress of physical analysis, a MIMO model is customized for the plant by expanding from the lumped physical character to a distributed system. Furthermore, the coupling feature of the object is addressed based on the decoupling approach and the pole placement approach, while the SISO approach is expanded to a universally acknowledged MIMO approach and Matlab is used to produce the simulation results.As a conclusion, in this research, firstly a state space model was established to expand the SISO system into a MIMO system and the interacted inputs and outputs have been decoupled using decoupling method; and then a Mamdani-type fuzzy control was designed for temperature control and an Interval Type-2 fuzzy control was designed for pressure control, using a simple state-space model instead of a fuzzy model, accordance with the practical plant in use, and very satisfied, very robust control performances were obtained.
Wang, Y. Research on intelligent controller design for MIMO spatially -Distributed systems with applications. (Thesis). University of the West of England
|Keywords||Spatially dynamic distributed systems; State-space approach;
Decoupling; Fuzzy Logic Systems; Interval Type-2 T-S Fuzzy Control, Biochemical Process
Final Approved thesis-Yizhi Wang v4.pdf