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

TL;DR

The paper addresses robust motion planning for high-DoF robot manipulators under unknown disturbances. It proposes a self-adaptive deep MPC framework that uses a gradient-sign based update law to train a neural observer and a dynamic-inversion controller, enabling real-time learning of the plant dynamics. The approach integrates a robust MPC formulation with an augmented state observer and horizon-based optimization to achieve setpoint tracking under uncertainty. In simulation on a UR5 manipulator in MuJoCo, the method demonstrates fast learning and accurate motion planning across multiple disturbance scenarios. Future work aims to enhance decision-making in dynamic environments by conditioning the value function on robot-centric maps for obstacle avoidance.

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.

Figures (4)

• Figure 1: Different motion planning tasks for a high-DoF robot arm(UR5) with deep MPC under different external forces applied on different parts to examine the self-adaptive performance.
• Figure 2: Tracking errors for the end-effector of the UR5 robot arm in different path planning tasks. From (a) to (f) denotes the tracking error concerning the scenarios Fig.\ref{['fig:overview']}
• Figure 3: Validation mean absolute error (MAE) during the training process.
• Figure 4: Training losses for each batch per epoch.