Sensorless position control for a single DOF system

This paper proposes a new method for sensorless position control for a mass which is connected with a servo motor through an elastic element, the system was assumed to have unknown parameter, so recursive least square (RLS) method was used to estimate these parameters depending on the reflected torque on the motor's shaft. Kalman filter (KF) is also used as an observer to estimate the position of the mass, depending on system dynamics and the measured states (angular displacement for motor's shaft and the armature current). Finally, PID controller was used to control the position of the mass depending on the estimated state (mass position). This method is very useful in applications that contain unmeasured states with unknown parameters. The system was implemented experimentally, and driven using Arduino Uno and L298, the current through the motor was measured using ACS712 module.