ATRS: Adaptive Trajectory Re-splitting via a Shared Neural Policy for Parallel Optimization
2026-04-24 • Robotics
Robotics
AI summaryⓘ
The authors address a problem in planning long paths by breaking it into smaller parts solved in parallel, but fixed splitting can cause slowdowns when some parts get stuck. They created ATRS, a system that uses a shared deep learning policy to decide when and where to split path segments dynamically during optimization. This method treats each segment like an agent sharing the same neural network, allowing it to adapt to different path lengths and new environments without extra training. Their approach makes the solver faster and stable, as shown in simulations and real-world tests where it improved speed and worked for both offline and real-time planning.
Trajectory OptimizationAlternating Direction Method of Multipliers (ADMM)Deep Reinforcement LearningMulti-Agent SystemsMarkov Decision ProcessParallel ComputingMotion PlanningZero-Shot GeneralizationDynamic Re-splittingReal-Time Replanning
Authors
Jiajun Yu, Guodong Liu, Li Wang, Pengxiang Zhou, Wentao Liu, Yin He, Chao Xu, Fei Gao, Yanjun Cao
Abstract
Parallel trajectory optimization via the Alternating Direction Method of Multipliers (ADMM) has emerged as a scalable approach to long-horizon motion planning. However, existing frameworks typically decompose the problem into parallel subproblems based on a predefined fixed structure. Such structural rigidity often causes optimization stagnation in highly constrained regions, where a few lagging subproblems delay global convergence. A natural remedy is to adaptively re-split these stagnating segments online. Yet, deciding when, where, and how to split exceeds the capability of rule-based heuristics. To this end, we propose ATRS, a novel framework that embeds a shared Deep Reinforcement Learning policy into the parallel ADMM loop. We formulate this adaptive adjustment as a Multi-Agent Shared-Policy Markov Decision Process, where all trajectory segments act as homogeneous agents and share a unified neural policy network. This parameter-sharing architecture endows the system with size invariance, enabling it to handle dynamically changing segment counts during re-splitting and generalize to arbitrary trajectory lengths. Furthermore, our formulation inherently supports zero-shot generalization to unseen environments, as our network relies solely on the internal states of the numerical solver rather than on the geometric features of the environment. To ensure solver stability, a Confidence-Based Election mechanism selects only the most stagnating segment for re-splitting at each step. Extensive simulations demonstrate that ATRS accelerates convergence, reducing the number of iterations by up to 26.0% and the computation time by up to 19.1%. Real-world experiments further confirm its applicability to both large-scale offline global planning and real-time onboard replanning within 35 ms per cycle, with no sim-to-real degradation.