Risk-Sensitive Motion Planning using Entropic Value-at-Risk
This work provides a method for robotic systems to plan motions more safely in environments with unpredictable obstacles, potentially reducing collision risks for autonomous vehicles and drones.
This paper addresses risk-sensitive motion planning with randomly moving obstacles by using an MPC scheme and formulating obstacle avoidance as a distributionally robust constraint based on Entropic Value-at-Risk (EVaR). The authors propose an algorithm for waypoint following and demonstrate its feasibility and completion in finite time through numerical experiments for a 2D system and a 3D quadcopter simulation, comparing EVaR policies against Conditional Value-at-Risk (CVaR).
We consider the problem of risk-sensitive motion planning in the presence of randomly moving obstacles. To this end, we adopt a model predictive control (MPC) scheme and pose the obstacle avoidance constraint in the MPC problem as a distributionally robust constraint with a KL divergence ambiguity set. This constraint is the dual representation of the Entropic Value-at-Risk (EVaR). Building upon this viewpoint, we propose an algorithm to follow waypoints and discuss its feasibility and completion in finite time. We compare the policies obtained using EVaR with those obtained using another common coherent risk measure, Conditional Value-at-Risk (CVaR), via numerical experiments for a 2D system. We also implement the waypoint following algorithm on a 3D quadcopter simulation.