遗传算法在逆相馈电感应电机中的应用

TRAINING NEURAL NETWORK USING GENETIC ALGORITHM FOR INVERTER-FED INDUCTION MACHINES

直接转矩控制（DTC）是工业应用感应电机的最简单的控制．它的关键部件是状态选择器．本文将讨论用神经网络来模拟通常的DTC的状态选择器，并区用遗传算法训练该神经网络．然后用训练好的神经网络代替状态选择器，在同一台机器上进行了仿真比较，得到满意结果．另外，还粗略地试验了遗传算法中一些参数对算法性能的影响．

Direct torque control (DTC) is the simplest torque control of induction machines for industrial application.The key components of DTC is the state selector.In this paper,we will discuss to train the neural network using a genetic algorithm.The neural network is used to simulate the state selector of the conventional DTC.The simulations that have been performed were obtained using the trained state selector neural network instead of the conventional DTC for the same machine.The simulation results are satisfying.In addition,we have been tried simply the effect of change in crossover rate and mutation rate.