Improved multiverse optimizer-based anti-saturation model free adaptive control and its application to manipulator grasping systems

Abstract

To address the stable grasping control issue in manipulator grasping systems, this manuscript proposes an improved multiverse optimizer-based anti-saturation model-free adaptive control (IMVO-AS-MFAC) algorithm. Initially, the manuscript converts the manipulator grasping system into an equivalent data model through dynamic linearization techniques. Then, based on the dynamic linearization model, the IMVO-AS-MFAC controller is designed. To address the actuator saturation problem that commonly occurs during the clamping process of manipulator grasping systems, a saturation parameter is introduced into the IMVO-AS-MFAC algorithm. Meanwhile, the controller parameters are optimized using an improved multiverse optimizer algorithm, which involves modifications to the initial population distribution and location update strategy. The improved algorithm demonstrates more competitive optimization performance compared to the traditional multiverse optimizer. The major advantage of the IMVO-AS-MFAC algorithm lies in the fact that only the input and output data of the manipulator grasping system are required throughout the entire control process, and the controller parameters are derived using an optimization algorithm rather than relying on empirical knowledge. Furthermore, rigorous mathematical analysis confirms the stability of the IMVO-AS-MFAC approach, and its effectiveness is validated through semi-physical experiments conducted in an environment integrating the MATLAB/Simulink module and the RecurDyn platform.