Abstract
Due to the recent development of DSP (Digital Signal Processors), we can control a servomotor with a very short sampling period, e. g. 100 [μs]. The speed information is usually calculated from the increased pulse number of a shaft encoder within one sampling period. However, in this simple method, the speed resolution in lower speed range is easily lost. If we use a longer sampling period for maintaining the speed resolution still high, the system will be unstable. To overcome this problem, we proposed the instantaneous speed observer to estimate the instantaneous speed at every 100μs sampling point and greatly improved the system stability. Also, we realized the disturbance rejection control using the disturbance observer inherently included in the speed observer. However, this speed observer needed the exact value of the inertia moment. In this paper, we propose an identification method of the inertia moment by applying some adaptive methods. We implemented the basic least square method, the recursive least square method and the fixed trace method, and compared their identification performances. We realized the stable operation of the speed observer. Further, we designed STR (Self Tuning Regulator) type speed controller by feeding back the identified inertia moment for further performance improvement. The proposal in this paper is one suggestion to combine the robust and adaptive methods in servo motor control.
