Sensor less control of DC motor using Kalman filter for low cost CNC machine

Abstract

DC motor is a common component used in automated systems, almost in every industry. In this paper, the focus is on the machine tool and precision system control design as DC motors are used as prime movers for all the stages. S ophisticated CNC machine tools use high end closed or open architecture control and sensors to run complex CAM routines. Linear or rotary encoders and glass scales are commonly used for rate sensing as a feedback system in CNC machine tools. The intelligent control design is necessary to operate these flexible CNC machine tool systems to achieve high accuracies. To make such systems in low cost, it is necessary to build machine tool system with reduced number of sensors. In this paper, an intelligent speed control of DC motor in a sensor less environment is followed in order to avoid costly hardware. This is a stochastic estimation method for estimating DC motor angular rate and current without using any rate sensors in the closed loop feedback. In this simple and cost effective approach, we have used Kalman Filter (KF) for the speed control of DC motor where it is applied to stochastic time varying system. KF is found robust against external disturbances and has the ability to predict such states and parameters which are even not easy to measure through deployed sensors. The employed scheme uses the state estimation technique by analyzing the currents in the windings and re-constructs the states. This way, the machine tool can get high accuracies using low cost control alternative.

Sensorless speed measurement for brushed DC motors

“…Various models have been proposed to convert the motor current [9] or voltage into a speed estimate and hence be usable in a motor speed control system. Interesting methods have included fuzzy algorithms [10,11], wavelets [12], genetic algorithms [13,14] and Kalman filters [15][16][17]. Saurav and Potluri [5] showed that for better accuracy the model should include mechanical factors, such as friction and any mechanical or electrical non-linearities.…”