Solving path planning problems in urban environments based on a priori sensors availabilities and execution error propagation

Abstract

This paper addresses safe path planning problem in urban environments under onboard sensor availability uncertainty. In this context, an approach based on Mixed-Observability Markov Decision Process (MOMDP) is presented. Such a model enables the planner to deal with a priori probabilistic sensor availability and path execution error propagation, the which depends on the navigation solution. Due to modelling particularities of this safe path planning problem, such as bounded hidden and fully observable state variables, discrete actions and particular transition function form, the belief state update function becomes a complex step that cannot be ignored during planning. Recent advances in Partially Observable Markov Decision Process (POMDP) solving have proposed a planning algorithm called POMCP, which is based on Monte-Carlo Tree Search method. It allows the planner to work on the history of the action-observation pairs without the need to compute belief state updates. Thereby, this paper proposes to apply a POMCP-like algorithm to solve the addressed MOMDP safe path planning problem. The obtained results show the feasibility of the approach and the impact of considering different a priori probabilistic sensor availability on the result policy.

Safe path planning for UAV urban operation under GNSS signal occlusion risk

“…the number of trials). However, the computation 5 In [46], the authors actually compared the planning performance with the heuristic policy following the shortest path obtained by Dijkstra algorithm over a discretized position space. However, it is not a fair comparison as the heuristic policy does not take into account the collision risk issued from the sensor availability and the navigation uncertainty.…”

Effective and Safe Trajectory Planning for an Autonomous UAV Using a Decomposition-Coordination Method

“…Looking through recent papers, we find that there are still several open questions that need to be resolved, like the modeling and control of various autonomous UAVs [5], target tracking [6], and UAVs wireless communication networks as in [7]. In this work, we propose a solution to the problem of safe trajectory planning for autonomous quadrotor, which is one of the most challenging topics in this field [8,9]. Several mathematical methods have been suggested to solve this problem.…”

“…Fig 9. The variation of the control u k = (u 1,k , u 2,k , u 3,k , u 4,k ) associated to the first optimal sub-trajectory T d,1 (case.…”

Solving Partially Observable Stochastic Shortest-Path Games

“…Throughout the years, many variants of the original problem have emerged. We focus on two of them with significant practical implications: (i) partially observable SSP [Patek, 1999;Egorov et al, 2016;Horák et al, 2018;Delamer et al, 2019] and (ii) planning in the presence of an adversary and SSP games [Patek and Bertsekas, 1999;Neu et al, 2012;Rosenberg and Mansour, 2020;Chen et al, 2020]. The partially observable SSP (also referred to as Goal-POMDP) generalizes the model to better reflect real-world scenarios where perfect information is not always available (e.g., robotic sensors are imprecise, and the true position of the robot in an environment may be unknown).…”