Authors : Y. Bensalem , W. Abboud , L. Sbita and M.N. Abdelkrim
Abstract: Development trends in industrial electrical drives indicate that the next generation of electrical drives will include some type of sensorless control. Controlled induction motor (IM) drives without speed sensors have the attractions of low cost and high reliability due to the absence of the mechanical component and its sensor cable. Speed estimation schemes that allow high dynamic performances are based on IM vector control. However, volt per Hertz (V/f) IM drives law produces satisfactory precision in speed sensorless control and is adequate for low dynamics applications. The proposed speed control scheme presented in this study using a simple low cost IM scalar control consists of a neural network controller (NNC) and a neural network speed estimator (NNSE). The NNC is used to produce a control force so that the motor speed can accurately tracks the reference command. The NNSE is trained off line by using the error back-propagation algorithm. The estimated speed is then fed back in the speed control loop and the speed-sensorless is then realized. A back-propagation algorithm is used as the learning algorithm to automatically adjust the weights of the NNC and NNSE in order to minimize the performance functions. The proposed sensorless control scheme has shown good performance in the transient and steady states and also at either variable-speed operations and load torque disturbances. Both computer simulations and experimental results demonstrate that the proposed control scheme is able to obtain robust speed sensorless IM control.
Y. Bensalem , W. Abboud , L. Sbita and M.N. Abdelkrim , 2008. A Sensorless Neural Network Speed Control of Induction Motor Drive. International Journal of Signal System Control and Engineering Application, 1: 150-158.