基于注意力引导与关键节点选择的无人系统轨迹预测方法与装置
By adding and evaluating perturbations to the trajectory prediction model of unmanned systems based on attention guidance and key node selection, the problems of uninterpretability and safety hazards of deep learning models in unmanned systems are solved, and the reliability and decision accuracy of the model are improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING INST OF SPACECRAFT SYST ENG
- Filing Date
- 2025-07-03
- Publication Date
- 2026-07-17
AI Technical Summary
Existing deep learning trajectory prediction models have uninterpretability and safety risks in unmanned systems, especially in the trajectory prediction process, which is easily affected by disturbances, leading to incorrect collision determinations.
An attention-guided and key node selection-based approach is adopted to add perturbations to the historical trajectories of candidate agents to generate reference adversarial examples. Adversarial examples are then generated through node importance measurement and local node perturbation to evaluate the model's credibility.
It improves the reliability assessment efficiency of unmanned system trajectory prediction models, reduces unnecessary disturbances, and ensures reasonable decision-making in complex environments.
Smart Images
Figure CN120762450B_ABST
Abstract
Citation Information
Patent Citations
Movement planning method, system and equipment of disabled-helping robot, storage medium and computer program
CN120085653A
Method for Interaction between a Vehicle and Road Users
DE102016210760A1