An enhanced neuro-adaptive PID sliding mode control for robot manipulators:promoting sustainable automation

The control of robot manipulators presents significant challenges, primarily due to their complex, nonlinear dynamics. Another major difficulty arises from environmental and operational disturbances. Numerous control frameworks and approaches have been proposed in the literature to address these issues. However, many of these approaches are discontinuous or non-adaptive, which negatively affects their performance and makes them unsuitable for real-time applications. This paper proposes a novel integrated control scheme that combines adaptive control with an improved continuous second-order sliding mode control (CSOS), utilizing generalized artificial neural networks (GANN) for enhanced performance in robot manipulators. The proposed control method consists of two key components: An adaptive proportional-integral-derivative (PID) controller and an adaptive CSOS-based control module (CSOSSD-APID), designed to deliver superior transient and steady-state performance. In this approach, the adaptive CSOS benefits from GANN’s strong noise handling capabilities and its ability to estimate uncertainties effectively. This integration significantly enhances the robustness of robot manipulator control across various tracking tasks, using a single pre-trained GANN model with fine-tuned weights tailored to each task. Numerical simulations demonstrate the effectiveness and versatility of the proposed control scheme, particularly in managing highly time-varying trajectories while contributing to more sustainable and efficient automation practices.