Nonlinear model identification and statistical verification using experimental data with a case study of the UR5 manipulator joint parameters

Abstract

The identification of nonlinear terms existing in the dynamic model of real-world mechanical systems such as robotic manipulators is a challenging modeling problem. The main aim of this research is not only to identify the unknown parameters of the nonlinear terms but also to verify their existence in the model. Generally, if the structure of the model is provided, the parameters of the nonlinear terms can be identified using different numerical approaches or evolutionary algorithms. However, finding a non-zero coefficient does not guarantee the existence of the nonlinear term or vice versa. Therefore, in this study, a meticulous investigation and statistical verification are carried out to ensure the reliability of the identification process. First, the simulation data are generated using the white-box model of a direct current motor that includes some of the nonlinear terms. Second, the particle swarm optimization (PSO) algorithm is applied to identify the unknown parameters of the model among many possible configurations. Then, to evaluate the results of the algorithm, statistical hypothesis and confidence interval tests are implemented. Finally, the reliability of the PSO algorithm is investigated using experimental data acquired from the UR5 manipulator. To compare the results of the PSO algorithm, the nonlinear least squares errors (NLSE) estimation algorithm is applied to identify the unknown parameters of the nonlinear models. The result shows that the PSO algorithm has higher identification accuracy than the NLSE estimation algorithm, and the model with identified parameters using the PSO algorithm accurately calculates the output torques of the joints of the manipulator.

A Comprehensive Multibody Model of a Collaborative Robot to Support Model-Based Health Management

“…Good et al [35] investigated the robot control algorithm and underlined the importance of taking into account the interactions between the robotic arm and the electromechanical drives. The model presented in [36] enhanced the mathematical representation of the robot structure, considering a simplified electric motor model. Tarn et al [37] developed a thirdorder dynamic model of the robot obtained by combining the manipulator and the motor dynamics to study the effect of a nonlinear feedback control.…”

“…Such an approach offers promising results, but requires the use of additional sensors such as accelerometers to measure the motion of each joint. An alternative way to identify the dynamic parameters involves the use of statistical techniques [36]. This method can provide excellent results in terms of overlapping of the model with the experimental results, but suffers from a very loose representation of the system physics.…”

Dynamic parameter identification based on improved particle swarm optimization and comprehensive excitation trajectory for 6R robotic arm

“…Li et al (2023) developed a simplified flexible dynamic model for a ball screw drive system, incorporating RLS estimation of parameter search space and PSO was used to identify the parameters. The accuracy is higher than WLS, PSO, GA and the convergence speed is also faster than PSO and GA. Abedinifar et al (2023) used PSO to identify critical parameters of nonlinear models, showing a higher accuracy compared with nonlinear least squares errors methods. Shao et al (2020) improved the Bouc-Wen model parameters through modifications in constriction factors, learning factors and inertia weights in PSO, enhancing precision as well as global convergence speed.…”