基于多阶段去噪与双分支时序网络的井下RSSI定位方法

打开文本图片集

中图分类号：TD655 文献标志码：A

Abstract： The underground Received Signal Strength Indicator （RSSI） signal exhibits non-stationary characteristics such as sharp spikes，high-frequency jiter，and trend drift under the influence of multipath propagation，occlusion，andelectromagnetic interference，resulting inlargepositioning errors.Existing positioning methods lack collaborative suppressionof multi-source interference，and their feature extraction and multi-scale feature fusion are insufficient.To addressthese problems，an underground RSSI positioning method based on multi-stage denoising and a dual-branch temporal network was proposed.Multi-stage denoising suppresed spike interference，high-frequency jiter，and trend driftthrough outliereliminationand interpolationrepair，adaptive Kalman filtering，and wavelet-domain adaptive gating，respectively，thereby producing a more stable RSSI sequence with preserved details.The dual-branch temporal network introduced the first-order difference as an auxiliary disturbance prior， extracted features in paralel through a trend branch and a disturbance branch，and adaptively fused themvia a channel attention mechanism.ABidirectional Long Short-Term Memory （Bi-LSTM） network was then used to capture contextual temporal dependencies， ensuring trajectory smoothness and continuity in complex dynamic environments.Test results showed that the RSSI signal became more stable after multi-stage denoising while preserving local dynamic features without excessve smoothing. The dual-branch temporal network achieved high accuracy， F1 -score，precision，and recall with fast convergence;in tests under different scenarios， both accuracy and F1 -score exceeded 85% ，demonstrating good generalization. In continuous positioning tasks under dynamic environments，the average positioning error of the proposed method was only 0.12m ：

Key words： underground coal-mine positioning; Received Signal Strength Indicator;RSSI; multi-stage denoising; dual-branch temporal network; Bi-LSTM

0引言

随着煤矿智能化建设的深人推进，井下作业模式正逐步由传统的“人工巡检 + 经验判断”向“感知-决策-执行一体化”的智能体系转变。（剩余12080字）