A sensorless speed control strategy for a low-cost full-closed-loop servo driver system

In digital system, a full-closed-loop configuration using a tachometers and an analog-digital converter with low resolution results in bad speed control and therefore influences the positioning performance in a high-precision positioning system. However, on the other hand, speed information cannot be accurately obtained through the encoder using traditional estimations. In this paper, we proposed a speed estimator - Kalman Filter by getting the position information through the encoder. The rules of selecting key parameters such as noise covariance in Kalman Filtering were modified to avoid trial-and-error. Experimental results in a control point of view showed a good performance both in normal and low speed range so that a full-closed-loop high-precision servo drive system could be achieved with a low-cost configuration.