In this paper, Genetic Algorithm optimization and fuzzy logic for diagnostic and fault detection of rotor broken bar in induction motor is developed. Fuzzy logic variables ranges are generally divided into set of membership functions where the form and interferences are regular. In this paper, fuzzy membership functions are not limited and can vary in the full range of corresponding variable. A Meta heuristic (Genetic algorithm) is then used to optimize membership parameters. This method is applied in diagnostic of a nonlinear system (Induction motor) for the rotor broken bar default. Simulation results are presented to show the effectiveness of the proposed approach.