An optimal neural network speed estimator using genetic algorithms

In direct torque control system, a speed-estimating method based on an optimal neural network using genetic algorithms is presented to estimate the rotor speed of asynchronous motors. According to the equations of asynchronous motors, the equation of speed is gotten, a speed estimator model based on recursive neural networks is also gotten, and genetic algorithms optimize the speed estimator. The simulation result shows the optimized speed estimator can track accurately the variation of speed, and the system has better dynamic performance.