Abstract:In contemporary control theory, self-adaptive methodologies are highly esteemed for their inherent flexibility and robustness in managing modeling uncertainties. Particularly, robust adaptive control stands out owing to its potent capability of leveraging robust optimization algorithms to approximate cost functions and relax the stringent constraints often associated with conventional self-adaptive control paradigms. Deep learning methods, characterized by their extensive layered architecture, offer significantly enhanced approximation prowess. Notwithstanding, the implementation of deep learning is replete with challenges, particularly the phenomena of vanishing and exploding gradients encountered during the training process. This paper introduces a self-adaptive control scheme integrating a deep MPC, governed by an innovative weight update law designed to mitigate the vanishing and exploding gradient predicament by employing the gradient sign exclusively. The proffered controller is a self-adaptive dynamic inversion mechanism, integrating an augmented state observer within an auxiliary estimation circuit to enhance the training phase. This approach enables the deep MPC to learn the entire plant model in real-time and the efficacy of the controller is demonstrated through simulations involving a high-DoF robot manipulator, wherein the controller adeptly learns the nonlinear plant dynamics expeditiously and exhibits commendable performance in the motion planning task.

What problem does this paper attempt to address?

The paper attempts to address the issues of external disturbances and modeling uncertainties faced by high-degree-of-freedom (High-DoF) robotic manipulators in motion planning tasks. Specifically, the paper proposes an adaptive robust motion planning method that utilizes Deep Model Predictive Control (Deep MPC) to enhance the robot's control performance and robustness. ### Main Issues: 1. **External Disturbances**: High-DoF robots in practical applications are subject to various external disturbances, which can lead to a decline in control performance. 2. **Modeling Uncertainty**: The dynamic model of robotic systems often has uncertainties, and traditional adaptive control methods struggle to effectively handle these uncertainties. 3. **Real-time Learning**: How to enable the controller to learn and adapt to environmental changes in real-time to improve the accuracy and robustness of motion planning. ### Solution: The paper proposes an adaptive control method that combines deep learning and Model Predictive Control (MPC), specifically including: - **Deep MPC Algorithm**: Utilizing deep neural networks to approximate complex nonlinear system dynamics and performing real-time optimization through model predictive control. - **Innovative Weight Update Rule**: Designing a new weight update mechanism that alleviates the vanishing and exploding gradient problems by using only the gradient sign. - **Adaptive Dynamic Inverse Control**: Combining enhanced state observers and auxiliary estimation circuits to achieve real-time learning and adaptation of the entire system model. ### Experimental Validation: Experiments conducted on the MuJoCo simulation platform validate the effectiveness of the proposed adaptive Deep MPC algorithm in various motion planning tasks. The experimental results show that this method can not only quickly learn the dynamics of nonlinear systems but also exhibit good robustness and control performance. ### Conclusion: The paper aims to address the issues of external disturbances and modeling uncertainties encountered by high-DoF robots in motion planning tasks by combining deep learning and model predictive control, thereby improving the robustness and real-time adaptability of the control system.

Self-Adaptive Robust Motion Planning for High DoF Robot Manipulator using Deep MPC

Adaptive robust posture control of a parallel manipulator driven by pneumatic muscles

Get the design right, the rest is automatic: Towards an adaptive MPC controller

Unified Motion/Force/Impedance Control for Manipulators in Unknown Contact Environments Based on Robust Model-Reaching Approach

Learning-Based Neural Dynamic Surface Predictive Control for MMC

Enhanced Robust Motion Control based on Unknown System Dynamics Estimator for Robot Manipulators

Motion Planning and Obstacle Avoidance for Robot Manipulators Using Model Predictive Control-based Reinforcement Learning

A Hierarchical MPC for End-Effector Tracking Control of Legged Mobile Manipulators

Real-Time Dynamic Obstacle Avoidance for Robot Manipulators Based on Cascaded Nonlinear MPC With Artificial Potential Field

Deep Model Predictive Optimization

Online-Learning Deep Neuro-Adaptive Dynamic Inversion Controller for Model Free Control

Hierarchical multiloop MPC scheme for robot manipulators with nonlinear disturbance observer

Adaptive Model Predictive Control with Data-driven Error Model for Quadrupedal Locomotion

Deep Model Predictive Variable Impedance Control

Robust tube-based MPC with smooth computation for dexterous robot manipulation

Adaptive Contact-Implicit Model Predictive Control with Online Residual Learning

A Model Weighted Adaptive Neural Backstepping Sliding-Mode Controller for Cooperative Manipulator System

A Robust Tube-Based Smooth-MPC for Robot Manipulator Planning

Adaptive Hybrid Optimization Learning-Based Accurate Motion Planning of Multi-Joint Arm

Whole-Body Control of an Autonomous Mobile Manipulator Using Model Predictive Control and Adaptive Fuzzy Technique

Deep Koopman Operator Based Model Predictive Control for Nonlinear Robotics Systems