Comparison of different Methods for Encoder Speed Signal Filtering exemplified by an Inverted Pendulum

This paper deals with the problem of filtering speed signals captured by means of an encoder in conjunction with the frequency measurement method. Filtering is necessary to cope with the quantization noise, which has an extremely negative influence on a state space controller using these signals as input. A simple low pass filter and two model based filter approaches (a Kalman filter and an adaptive Kalman filter) are presented and their effect on the performance of the controller is compared through the example of a cart-pole system.