GA based PI control of BLDCM Drive

The brushless DC motor (BLDCM) has a good control performance. The classical PI controller is widely used controller. But the determining of its parameter needs the mathematical model of the system. Usually the exact mathematical model of system is hard to known. This places a restriction on the applying of the commonly used Z-N method to determine the parameters. Also the parameter determined by the Z-N method is not the optimum one. GA is an optimization technique that performs a parallel, stochastic, but directed search to evolve the most fit population. And it dose not need the gradient information of object function. But only dependent on the evaluation function, it can guide the optimum process. Except this, the GA algorithm also has the property of robustness and parallel computation. Therefore, in this paper, the GA algorithm is applied to optimum the parameters of PI controller. Through this method, the response of the system is improved and overshoot is reduced and steady state error is eliminated. The determining of parameters of PI controller is simplified and the optimum response of system can be reached. Experimental results demonstrate that a good control performance is achieved. The system responses quickly with little overshoot. There is no error in steady state. The system shows a strong ability to reject disturbance and a strong characteristic of robustness.