1 Optimisation of Fuzzy Predictive Functional Control using coordination techniques

Abstract — Predictive functional control (PFC), belongs to the family of predictive control techniques. It has been demonstrated as a powerful algorithm for controlling process plant. In this paper, PFC strategy is extended to nonlinear processes. The predictive functional control is combined with a fuzzy model of the process and formulated in the state space domain. The process model is decomposed into subsystems each described by a fuzzy rules. In controller design, prediction errors and control energy is minimised trough a two layered iterative optimization process. The lower layer finds local control policies for each sub system. The objective of the upper layer is to find a near optimum for the overall system trough coordinating the subsystems. The performance of the FPFC approach is demonstrated trough an example of CSTR process. Index Terms—Fuzzy identification, predictive control, decomposition-coordination, nonlinear systems. I