Affiliations: [a]
Department of Electrical Engineering, Instituto Tecnológico de Querétaro, Querétaro Qro., México | [b] Instituto Politécnico Nacional, CICATA IPN, Querétaro Qro., México | [c] Universidad Nacional Autónoma de México, Centro de Física Aplicada y Tecnología Avanzada, Querétaro Qro., México
Abstract: This work, presents a new tuning algorithm for type-I fuzzy logic controllers (FLCs), called Change Symmetry in Input Membership Functions (CSIMFS). The algorithm uses an iterative method that applies specific criteria to reduce mathematical operations during the tuning process to develop embedded FLCs. It uses one-input-one-output fuzzy inference system with five control rules and no more than two tuning parameters. The effectiveness of the proposed algorithm was validated experimentally using a real nonlinear pneumatic positioning system and verified with a simulation using a second order model. In both cases, CSIMFS algorithm exhibited a better performance compared with heuristically tuned FLCs. Moreover, a new combined performance index was defined and compared to traditional ITAE index offering better results, reducing the maximum overshoot and a faster convergence of settling time.
