A new identification approach for closed loop dc electrical drive using prior knowledge

Abstract

Monitoring robots and machine-tools is a subject of increasing interest because of industrial needs of control, reliability and availability of these machines. The main devices in these systems are the electrical drives (axis and feed-drives). Improvements in the monitoring of these devices can be realised by estimating the parameters of the electrical and mechanical parts of the system. In this paper we propose a new approach to estimate these parameters using system prior knowledge. This approach can be applied when drives are controlled by three cascaded loops with a current controller, a velocity controller and a position controller and when a velocity have a trapezoidal profile along point to point motion. Estimation is done off-line when system is unloaded and during point to point drives motions.