The thesis presents a Genetic Algorithm with Adaptive Search Space (GAASS) proposed to improve both convergence performance and solution accuracy of traditional Genetic Algorithms (GAs). The proposed GAASS incorporates an adaptive mechanism to control the size of the search space as well as the rates of crossover and mutation. The proposed GAASS method has been hybridized to a real-coded genetic algorithm to perform hysteresis parameters identification and hysteresis inverse compensation of an electromechanical-valve actuator installed on a pneumatic system. The experimental results have demonstrated the supreme performance of the proposed GAASS in the search of optimum solutions.