Advancements in Motion Planning and Control

The field of motion planning and control is witnessing significant developments, with a focus on improving the efficiency, safety, and adaptability of systems. Researchers are exploring innovative approaches to integrate planning and control, such as using model predictive control (MPC) and Bayesian optimization to improve the performance of autonomous systems. Additionally, there is a growing interest in developing risk-aware and robust control methods that can handle uncertainties and non-stationary environments. Noteworthy papers in this area include: The Optimistic Risk-averse Actor Critic (ORAC) approach, which prevents convergence to sub-optimal policies in risk-averse constrained reinforcement learning. The Path Feasibility Governor (PathFG) framework, which integrates path planners with nonlinear MPC to ensure constraint satisfaction and stability. The Risk-Aware Adaptive Robust MPC (RAAR-MPC) framework, which orchestrates a synthesis of proactive risk assessment and reactive risk regulation to satisfy chance constraints with a user-defined probability.

Sources

Optimistic Exploration for Risk-Averse Constrained Reinforcement Learning

Integrating Planning and Predictive Control Using the Path Feasibility Governor

Informed Hybrid Zonotope-based Motion Planning Algorithm

Symptom-Driven Personalized Proton Pump Inhibitors Therapy Using Bayesian Neural Networks and Model Predictive Control

Intersection of Reinforcement Learning and Bayesian Optimization for Intelligent Control of Industrial Processes: A Safe MPC-based DPG using Multi-Objective BO

Polygonal Obstacle Avoidance Combining Model Predictive Control and Fuzzy Logic

Mixed Discrete and Continuous Planning using Shortest Walks in Graphs of Convex Sets

A Robust Controller based on Gaussian Processes for Robotic Manipulators with Unknown Uncertainty

MPC-based Coarse-to-Fine Motion Planning for Robotic Object Transportation in Cluttered Environments

A Risk-Aware Adaptive Robust MPC with Learned Uncertainty Quantification

Built with on top of