Safe and Fast Tracking on a Robot Manipulator: Robust MPC and Neural Network Control

Abstract

Fast feedback control and safety guarantees are essential in modern robotics. We present an approach that achieves both by combining novel robust model predictive control (MPC) with function approximation via (deep) neural networks (NNs). The result is a new approach for complex tasks with nonlinear, uncertain, and constrained dynamics as are common in robotics. Specifically, we leverage recent results in MPC research to propose a new robust setpoint tracking MPC algorithm, which achieves reliable and safe tracking of a dynamic setpoint while guaranteeing stability and constraint satisfaction. The presented robust MPC scheme constitutes a one-layer approach that unifies the often separated planning and control layers, by directly computing the control command based on a reference and possibly obstacle positions. As a separate contribution, we show how the computation time of the MPC can be drastically reduced by approximating the MPC law with a NN controller. The NN is trained and validated from offline samples of the MPC, yielding statistical guarantees, and used in lieu thereof at run time. Our experiments on a state-of-the-art robot manipulator are the first to show that both the proposed robust and approximate MPC schemes scale to real-world robotic systems.

arXiv comment8 pages, 4 figures,

References in corpus (9)

• Efficient representation and approximation of model predictive control laws via deep learning
• Learning an Approximate Model Predictive Controller with Guarantees
• Implementation of Nonlinear Model Predictive Path-Following Control for an Industrial Robot
• A nonlinear tracking model predictive control scheme for dynamic target signals
• Optimization-Based Collision Avoidance
• A computationally efficient robust model predictive control framework for uncertain nonlinear systems -- extended version
• A nonlinear model predictive control framework using reference generic terminal ingredients -- extended version
• Large Scale Model Predictive Control with Neural Networks and Primal Active Sets
• Near-Optimal Rapid MPC using Neural Networks: A Primal-Dual Policy Learning Framework

Cited by in corpus (21)

• Real-time Neural-MPC: Deep Learning Model Predictive Control for Quadrotors and Agile Robotic Platforms
• Model-Free Safety-Critical Control for Robotic Systems
• A nonlinear tracking model predictive control scheme for dynamic target signals
• Adaptive CLF-MPC With Application To Quadrupedal Robots
• Analysis and design of model predictive control frameworks for dynamic operation -- An overview
• Reliably-stabilizing piecewise-affine neural network controllers
• Optimality Principles in Spacecraft Neural Guidance and Control
• A computationally efficient robust model predictive control framework for uncertain nonlinear systems -- extended version
• Deep Model Predictive Control with Stability Guarantees
• Approximate non-linear model predictive control with safety-augmented neural networks
• Discrete Robust Control of Robot Manipulators using an Uncertainty and Disturbance Estimator
• Vision-Based Uncertainty-Aware Motion Planning based on Probabilistic Semantic Segmentation
• Safe Value Functions
• Shared Telemanipulation with VR controllers in an anti slosh scenario
• Minimal Constraint Violation Probability in Model Predictive Control for Linear Systems
• Non-Parametric Neuro-Adaptive Control Subject to Task Specifications
• On exploration requirements for learning safety constraints
• Event-triggered Learning for Linear Quadratic Control
• The Mini Wheelbot: A Testbed for Learning-based Balancing, Flips, and Articulated Driving
• Artificial-reference tracking MPC with probabilistically validated performance on industrial embedded systems
• Adaptive Non-singular Terminal Sliding Mode Fault-tolerant Control of Robotic Manipulators Based on Contour Error Compensation