Control Performance Simulation of a Space Robotic Joint by Estimation of Motor Fast Shaft Position and Speed via a Seven-State Nonlinear Kalman Filter

This paper presents the results of a control performance simulation for a robotic joint, sized for space applications, via a seven-state nonlinear Kalman Filter. This leads to better estimation of joint mechanical state and in particular the motor fast shaft position and speed which are directly used for the commutation and speed control functions. In the presented simulation the sensor information assumed available to the Kalman Filter are the ones normally available in the drive electronics, like currents and voltages, motor windings temperatures and speed information of the joint output shaft, downstream the reducer, derived from an output shaft absolute position sensor. The general technique to utilize a Kalman Filter estimator to support the reconstruction of the motor shaft mechanical state have been already proposed in di erent variants and could be considered for robotic applications in hostile environments where active electronics, inclusive hall sensors, installed in the robotic arm should be reduced to the minimum and where the need to keep the arm design as light and as simple as possible suggests to reduce the number of components into the joint to the minimum. Possible scenarios are represented by Moon and Mars robotic exploration where the extreme environmental conditions of certain areas and the convenience to keep the mass of the robotic arm as low as possible suggest the use of techniques similar to the one discussed.