基于分布式麦克风阵列协同定位的三维-TDOA定位方法

By using an adaptive positioning method with a distributed microphone array, combined with quantum entangled state sensors and deep learning models, and optimizing TDOA estimation, the problem of insufficient accuracy of traditional positioning technology in complex environments is solved, and high-precision three-dimensional positioning is achieved.

CN121186707BActive Publication Date: 2026-07-17SOUTHWESTERN INST OF PHYSICS

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SOUTHWESTERN INST OF PHYSICS
Filing Date
2025-09-25
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Traditional positioning technologies suffer from insufficient positioning accuracy in complex environments, severe noise interference, and inadequate data fusion and collaborative positioning mechanisms, making it difficult to achieve high-precision three-dimensional positioning.

Method used

An adaptive localization method based on a distributed microphone array is adopted, which combines quantum entangled state sensors, deep learning models, blockchain data fusion and reinforcement learning algorithms to optimize TDOA estimation and localization strategies, and performs three-dimensional spatial search through quantum annealing algorithm.

Benefits of technology

It achieves sub-centimeter-level positioning accuracy, resists complex electromagnetic interference and noise, is suitable for various indoor and outdoor environments, and has high versatility and expandability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121186707B_ABST
    Figure CN121186707B_ABST
Patent Text Reader

Abstract

本发明提供了一种基于分布式麦克风阵列协同定位的三维‑TDOA定位方法,涉及TDOA定位技术领域,方法包括:构建自适应动态麦克风阵列;在麦克风节点信号采集前端引入量子纠缠态传感器,测量目标信号到达时间差;构建深度学习模型,根据环境因素预测声速变化,在TDOA测量中动态调整声速值;麦克风阵列节点根据自身测量数据和与其他节点交互信息,通过强化学习算法优化定位策略;对数据进行融合,基于融合结果数据,执行TDOA估计改进算法;基于TDOA计算结果,采用量子退火算法进行三维空间搜索求解目标最优位置;融合多源信息为定位算法添加约束条件,优化定位结果。本发明方法能够实现亚厘米级定位精度,满足对定位精度要求极高的应用场景。
Need to check novelty before this filing date? Find Prior Art