基于前馈多层神经网络的无传感器.docVIP

基于前馈多层神经网络的无传感器 稳态速度辨识研究 摘要：研究了应用前馈多层神经网络，在无传感器条件下辨识变频器供电的稳态异步电机转速的方法。该方案将电机定子电流和频率作为输入，电机转速作为输出，按照“离线训练，在线辨识”的思想，训练神经网络，使模拟异步电机的稳态行为，实现电机转速的无传感器实时在线监测。并计算出转速辨识的平均误差，与其他无传感器转速辨识方法比较，验证了此种方法的优越性。 关键词 无传感器 前馈多层神经网络 变频器供电 稳态运行 Research of Sensorless Speed Estimation in Steady State Based on Feed For ward Multiplayer Neural Network Abstract：Method of Sensorless speed estimation for inverter-fed induction motor in steady state based on Feedforward Multilayer Neural Network has been studied. With stator currents and frequency as input and the motor speed as output, a neural network has been trai