轨迹规划方法及电子设备

By using a visual language model for multimodal joint modeling, the problem of fragmented multimodal information in trajectory planning is solved, achieving end-to-end integrated trajectory planning and improving environmental perception and robustness.

CN122412965APending Publication Date: 2026-07-17INSPUR SUZHOU INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
INSPUR SUZHOU INTELLIGENT TECH CO LTD
Filing Date
2026-06-16
Publication Date
2026-07-17

Smart Images

  • Figure CN122412965A_ABST
    Figure CN122412965A_ABST
Patent Text Reader

Abstract

本申请公开了一种轨迹规划方法及电子设备，涉及车辆技术领域，将车辆的视觉感知数据、查询文本和历史轨迹数据输入预训练的视觉语言模型，通过视觉语言模型输出感知特征向量，视觉语言模型包括视觉、文本、轨迹编码器和预训练语言模型，分别将视觉感知数据编码为视觉词元序列，将查询文本编码为文本词元序列，将历史轨迹数据编码为轨迹词元序列，将视觉、文本和轨迹词元序列输入预训练语言模型，预训练语言模型输出感知特征向量；将历史轨迹数据编码为轨迹特征向量，将车辆状态数据编码为状态特征向量，根据感知特征、轨迹特征和状态特征向量生成目标规划轨迹，解决了相关技术中轨迹规划对环境的感知与语义的理解能力不足和鲁棒性较差的技术问题。
Need to check novelty before this filing date? Find Prior Art