Genetic-based fuzzy clustering for DC-motor friction identification and compensation

A fuzzy-logic-based model describing the friction present in a de-motor system is derived in this paper. Based on fuzzy clustering techniques, the structure, as well as the premise and consequence parameters are inferred in an off-line manner. The fine-tuning of these parameters is accomplished through a genetic algorithm which minimizes a system modeling relevant functional, This genetic algorithm encodes these parameters as chromosomes, and creates the next generation of fuzzy models through natural selection and survival of the fittest chromosome. This model is used as a feedforward term for tracking purposes of the de-motor's angular velocity. The proposed feedforward compensation scheme, coupled to a classical feedback controller improves the system's response in typical dc-motor micromaneuvers, Experimental results are offered to validate the performance of the proposed friction's fuzzy model and the control technique.