Position and Speed Detection Method Based on Adaptive Extended Moving-Window Linear Regression for Traction Machine Drives

In the traction machine system, the use of the ordinary-resolution encoder can reduce the cost and the difficulty of the installation of the motor system. However, the position and the speed obtained by the ordinary-resolution encoder cause the detection errors at low speed, resulting in reduced detection accuracy. In order to acquire the accurate position and speed at the low resolution of the encoder, an adaptive extended moving-window linear regression (AEMLR)-based position and speed detection method is proposed in this article. Based on the analysis of the characteristics of the traditional moving-window linear regression (MLR) method, the regression errors caused by MLR can be reduced by extending the sampling-window length to calculate the position regression basis function. The relationship between the speed error, the operating speed, and the sampling-window length is established by extending the basis function calculated using extended MLR. According to the relationship, the sampling-window length trajectory planning curve can be obtained. In this way, the adaptive tuning of the window length as the speed changes can be achieved by fixing the sampling error to obtain more accurate regression signals. The experimental results on the 11.7-kW traction machine drive platform verify the effectiveness of the proposed method.