Trajectory Generation with Endpoint Regulation and Momentum-Aware Dynamics for Visually Impaired Scenarios

Yuting Zeng; Manping Fan; You Zhou; Yongbin Yu; Zhiwen Zheng; Jingtao Zhang; Liyong Ren; Zhenglin Yang

cs.ROFeb 25, 2026

Trajectory Generation with Endpoint Regulation and Momentum-Aware Dynamics for Visually Impaired Scenarios

Authors: Yuting Zeng, Manping Fan, You Zhou, Yongbin Yu, Zhiwen Zheng, Jingtao Zhang, Liyong Ren, Zhenglin Yang

Organizations: School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, Sichuan, P.R. China. · School of Medicine, University of Electronic Science and Technology of China, Chengdu, Sichuan, China. · Sichuan Provincial Key Laboratory for Human Disease Gene Study, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan, China.

Abstract

Trajectory generation for visually impaired scenarios requires smooth and temporally consistent state in structured, low-speed dynamic environments. However, traditional jerk-based heuristic trajectory sampling with independent segment generation and conventional smoothness penalties often lead to unstable terminal behavior and state discontinuities under frequent regenerating. This paper proposes a trajectory generation approach that integrates endpoint regulation to stabilize terminal states within each segment and momentum-aware dynamics to regularize the evolution of velocity and acceleration for segment consistency. Endpoint regulation is incorporated into trajectory sampling to stabilize terminal behavior, while a momentum-aware dynamics enforces consistent velocity and acceleration evolution across consecutive trajectory segments. Experimental results demonstrate reduced acceleration peaks and lower jerk levels with decreased dispersion, smoother velocity and acceleration profiles, more stable endpoint distributions, and fewer infeasible trajectory candidates compared with a baseline planner.

Figures & tables

Panel of Figure 1 — Fig. 1 : Schematic illustration of endpoint regulation in trajectory sampling. (a) Unconstrained sampling yields dispersed terminal states within the trajectory cluster. (b) Endpoint regulation constrains terminal deviation with respect to a reference sample, producing well-conditioned boundary states for subsequent local optimization.

Fig. 2 : Illustration of momentum inconsistency in heuristic trajectory generation. (a) Spatial trajectory composed of independently optimized segments. (b) Spatial trajectory with momentum-aware regulation, showing similar geometric layout. (c) Velocity profile without momentum consistency, exhibiting a slope discontinuity at the segment boundary. (d) Velocity profile with enforced momentum consistency, yielding smooth temporal evolution.

Fig. 3 : Velocity, acceleration, and jerk profiles along the longitudinal displacement across three low-speed scenarios. The results compare the baseline and the proposed method, highlighting consistent trends in motion smoothness and jerk attenuation.

Fig. 4 : Boxplots of longitudinal and lateral jerk across three scenarios.

Panel of Figure 5 — Fig. 5 : Comparison of trajectory cluster generation with and without endpoint regulation at a representative time step, where (a) corresponds to the baseline and (b) shows the result with endpoint regulation.

Scenario	Method	Mean $\pm$ Std	Min	Max
S1	B [ 43 ]	0.556 $\pm$ 0.214	0.232	0.907
	P	0.556 $\pm$ 0.212	0.226	0.875
S2	B [ 43 ]	0.663 $\pm$ 0.108	0.566	0.863
	P	0.673 $\pm$ 0.084	0.566	0.818
S3	B [ 43 ]	0.678 $\pm$ 0.165	0.480	0.926
	P	0.700 $\pm$ 0.135	0.479	0.910

TABLE I : Endpoint density statistics based on nearest-neighbor distance.

Fig. 6 : Nearest-neighbor distance distribution of trajectory endpoints in a representative scenario.

Panel of Figure 8 — Fig. 7 : Bar charts summarizing feasibility and infeasibility statistics across three scenarios, where (a) reports the overall feasibility ratio and (b) compares the rates of infeasibility due to maximum curvature, yaw rate, and curvature rate constraints.

Visually Impaired Navigation Assistance

See all 22 papers

Momentum-constrained Hybrid Heuristic Trajectory Optimization Framework with Residual-enhanced DRL for Visually Impaired Scenarios
Apr 16, 2026Yuting Zeng, Zhiwen Zheng, Jingya Wang +6Trajectory Optimization Assistive Robotics
School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, Sichuan, P.R. China · Sichuan Provincial Key Laboratory for Human Disease Gene Study, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan, China. · Research Unit for Blindness Prevention, Chinese Academy of Medical Sciences, Sichuan Academy of Medical Sciences and Sichuan Provincial People’s Hospital, Chengdu, Sichuan, China.
Online Generation of Collision-Free Trajectories in Dynamic Environments
Feb 28, 2026Nermin Covic, Bakir LacevicConstrained Motion Planning Trajectory Optimization
Faculty of Electrical Engineering, University of Sarajevo, Bosnia and Herzegovina
DRIFT: Diffusion-based Rule-Inferred For Trajectories
Mar 1, 2026Jinyang Zhao, Handong Zheng, Yanjiu Zhong +3Robot Motion Generation Diffusion-Based Planning
Hefei University of Technology, Hefei, China · School of Management, Ministry of Education Engineering Research Center for Intelligent Decision-Making & Information System Technologies, Hefei University of Technology, Hefei, 230009, Anhui, China. · Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China +1

Apr 16, 2026J/K move · Enter open · S save

Momentum-constrained Hybrid Heuristic Trajectory Optimization Framework with Residual-enhanced DRL for Visually Impaired Scenarios

Yuting Zeng, Zhiwen Zheng, Jingya Wang, You Zhou+5

School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, Sichuan, P.R. China · Sichuan Provincial Key Laboratory for Human Disease Gene Study, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan, China. · Research Unit for Blindness Prevention, Chinese Academy of Medical Sciences, Sichuan Academy of Medical Sciences and Sichuan Provincial People’s Hospital, Chengdu, Sichuan, China.

Trajectory Optimization Assistive Robotics Dynamic Obstacle Avoidance

Open paper AskPDF arXiv

Safe and efficient assistive planning for visually impaired scenarios remains challenging, since existing methods struggle with multi-objective optimization, generalization, and interpretability. In response, this paper proposes a Momentum-Constrained Hybrid Heuristic Trajectory Optimization Framework (MHHTOF). To balance multiple objectives of comfort and safety, the framework designs a Heuristic Trajectory Sampling Cluster (HTSC) with a Momentum-Constrained Trajectory Optimization (MTO), which suppresses abrupt velocity and acceleration changes. In addition, a novel residual-enhanced deep reinforcement learning (DRL) module refines candidate trajectories, advancing temporal modeling and policy generalization. Finally, a dual-stage cost modeling mechanism (DCMM) is introduced to regulate optimization, where costs in the Frenet space ensure consistency, and reward-driven adaptive weights in the Cartesian space integrate user preferences for interpretability and user-centric decision-making. Experimental results show that the proposed framework converges in nearly half the iterations of baselines and achieves lower and more stable costs. In complex dynamic scenarios, MHHTOF further demonstrates stable velocity and acceleration curves with reduced risk, confirming its advantages in robustness, safety, and efficiency.

Abstract

Visually Impaired Navigation Assistance

Explore similar work

Momentum-constrained Hybrid Heuristic Trajectory Optimization Framework with Residual-enhanced DRL for Visually Impaired Scenarios

Online Generation of Collision-Free Trajectories in Dynamic Environments

DRIFT: Diffusion-based Rule-Inferred For Trajectories