Optimization Method of UWB Indoor Positioning Based on Two-Layer Bi-Directional GRU Network

Ultra-wideband technology is characterized by fast transmission speed and strong anti-interference ability, so it is widely used in high-precision indoor positioning. However, when ultra-wideband technology is used for indoor environment localization, the presence of non-line-of-sight, multi-path and other interference factors seriously affects the ultra-wideband ranging accuracy and signal stability. In order to improve the problem of declining localization accuracy of ultra-wideband technology in complex environments, this paper proposes a localization method based on the bi-layer bi-directional gated cyclic cell network model, which utilizes the time-position sequence characteristics of the locating target’s dynamic motion, the motion features are extracted from historical data, and the network structure parameters are designed to verify its effectiveness. The experimental results show that the ultrawideband localization accuracy based on this GRU model is improved by 56.75%, 50.64%, and 18.2% compared with polynomial fitting algorithm, unidirectional GRU model, and bidirectional GRU model, respectively, and the localization effect is better.