A Two-Layer Trajectory Tracking Control Scheme of Manipulator Based on ELM-SMC for Autonomous Robotic Vehicle
Xingyu Wang,Dazhi Wang,Mingtian Du,Keling Song,Yongliang Ni,Yanming Li
DOI: https://doi.org/10.1109/tase.2023.3238349
IF: 6.636
2024-01-01
IEEE Transactions on Automation Science and Engineering
Abstract:With the development of autonomous driving technology, some special vehicles should also be developed into autonomous driving vehicles. The robotic vehicles are widely used in engineering operations, and the robotic manipulator is a common tool mounted on robotic vehicles. The autonomous driving of robotic vehicles requires not only automatic movement, but also the automation of the robotic manipulator drive system. Therefore, this paper proposes an adaptive trajectory tracking control of manipulator based on sliding mode control (SMC), which can more accurately describe the joint space movement of the manipulator system under multi task coupling and external interference. Considering the multi-tasking situation, the influence of obstacle avoidance and base tilt on the trajectory of manipulator is investigated by using extreme learning machine (ELM), and the framework of machine learning is established. The output based on ELM-SMC is used as the bottom-layer control. Then an effective trajectory compensation method is proposed as the top-level control to avoid the error accumulation in the periodic repeated operation of the manipulator. Finally, the application effect of trajectory error tracking and compensation of manipulator in complex tasks of dynamic environment is verified by simulation and experiments, which lays a foundation for the stability control of manipulator in practical complex engineering applications. Note to Practitioners —The motivation of this paper is to solve the problem of poor trajectory tracking control accuracy of the manipulator caused by the change of the direction of the manipulator base. It is suitable for the high-precision position control of the manipulator assembled on the mobile robotic vehicle. Nowadays, the autonomous driving technology is developing rapidly, and the automation of traffic will be widely popularized in the future. Therefore, we should not only pay attention to the autonomous driving function of passenger vehicles, but also give consideration to the autonomous driving technology of robotic vehicles applied in engineering. It not only needs to complete the driving function, but also needs to ensure the reliable operation of the manipulator. In this paper, we mathematically deduce the control model of the manipulator. Then, we use SMC as the trajectory tracking control method of the manipulator, and use ELM to estimate the trajectory error caused by the end trajectory of the manipulator in the non-horizontal state of the base. Furthermore, we propose an optimization algorithm to optimize the compensation coefficient of cubic spline and give a reasonable compensation scheme. Through simulation and experiments, it is preliminarily verified that the schemes proposed in this paper are feasible, but they lack the application verification in the actual robotic vehicles, and the model fitting part of machine learning is completed with the help of the host computer, and the calculation ability of the controller of the actual manipulator may also be the limitation of completing this scheme.