Occlusion-Aware Consistent Model Predictive Control for Robot Navigation in Occluded Obstacle-Dense Environments

TL;DR

Occlusion-aware CMPC addresses safe navigation under partial observability by modeling occluded regions and risk cones using $v_{obs,max}$ to generate multiple trajectory branches; it solves them in parallel with ADMM, enforcing a consensus segment to maintain motion consistency. It optimizes over planning horizon $T_h=6\,\text{s}$ with consensus length $T_c=2\,\text{s}$ and guidance from Visual-PRM. Experiments show real-time performance on Ackermann vehicles with improved safety margins and reduced lateral velocity fluctuations compared with baselines.

Abstract

Ensuring safety and motion consistency for robot navigation in occluded, obstacle-dense environments is a critical challenge. In this context, this study presents an occlusion-aware Consistent Model Predictive Control (CMPC) strategy. To account for the occluded obstacles, it incorporates adjustable risk regions that represent their potential future locations. Subsequently, dynamic risk boundary constraints are developed online to enhance safety. Based on these constraints, the CMPC constructs multiple locally optimal trajectory branches (each tailored to different risk regions) to strike a balance between safety and performance. A shared consensus segment is generated to ensure smooth transitions between branches without significant velocity fluctuations, preserving motion consistency. To facilitate high computational efficiency and ensure coordination across local trajectories, we use the alternating direction method of multipliers (ADMM) to decompose the CMPC into manageable sub-problems for parallel solving. The proposed strategy is validated through simulations and real-world experiments on an Ackermann-steering robot platform. The results demonstrate the effectiveness of the proposed CMPC strategy through comparisons with baseline approaches in occluded, obstacle-dense environments.

Figures (5)

• Figure 1: Robot navigation in an occluded, obstacle-dense environment using the proposed occlusion-aware CMPC. The occlusion-aware CMPC generates three trajectories (orange, red, and blue) with different considerations of risk regions. All trajectories share an initial consensus segment (green) to enable smooth transitions between trajectories and ensure motion consistency.
• Figure 2: Modeling of occluded regions and risk regions. Occluded regions are bounded by two tangent lines. Risk regions are located along two tangent lines, each with center $(c_x,c_y)$ and radius $r_{\text{risk}}$.
• Figure 3: Simulation snapshots, where the robot navigates in an occluded, obstacle-dense environment. The robot, shown as an orange rectangle with black wheels, optimizes three trajectories in parallel. Dark-gray boxes represent visible obstacles, and light-gray boxes represent occluded obstacles. The arrow indicates the current velocity vector of the robot. Curves in different colors represent trajectories with different risk region configurations.
• Figure 4: The longitudinal and lateral velocity profiles of three approaches in the same scenario. When the occluded dynamic obstacles suddenly appear, both longitudinal and lateral velocities of our CMPC are more stable.
• Figure 5: Real-world experiment snapshots on the TianRacer robot. The robot successfully navigates through obstacles and avoids collision with the occluded dynamic obstacle.

Theorems & Definitions (5)

• Remark 1
• Remark 2
• Remark 3
• Remark 4
• Remark 5