Dynamic Parameters Estimation for a 6‐DOF Manipulator Using the Recursive Least Squares Method in Practice

This paper presents a recursive least squares (RLSs) technique for identifying the payloads acting on the 6‐degree‐of‐freedom (DOF) manipulator during the robot operation. The least squares (LSs) methods were commonly used to identify offline physical parameters. Nevertheless, in scenarios involving time‐varying systems with sudden parameter alterations, the LS method may no longer be suitable for estimating dynamic parameters in real time because the LS method is only used once all the measurements are made. Hence, the RLS technique was developed in online scenarios to address this disadvantage. The RLS method is utilised to identify unknown payloads in real‐time applications, enabling adaptive estimation during manipulator operation. Besides, the LS method is also applied for determining the initial parameters. Experiments are carried out on the 6‐DOF robot with an excitation trajectory designed for it. Afterwards, the data is acquired using a propagation derivative (PD) controller via the Simulink Desktop Real‐Time toolbox. Furthermore, the Fourier series and the zero‐phase low‐pass filter are used to analyse and process the collected signals to estimate the dynamic parameters of the real target by using the RLS method. The estimation of the mass of the payload added to the manipulator is also computed through the variance on some estimated inertial parameters with and without payload. Finally, using the proposed method, the root mean squared error (RMSE) criterion is used to evaluate the accuracy of the estimated results of two case studies.