Research And Implementation Of Inner Navigation To Assist Bluetooth Indoor Location Technology

Fire always threatens the life safety of fire rescue personnel,and the positioning of personnel movement in the field rescue process is a key issue to ensure safety,and has become a research hotspot.With the vigorous development of the Internet of Things technology,more and more positioning technology solutions can be selected,and the service requirements for indoor personnel positioning are also increasing.Bluetooth Low Energy(BLE)technology has low cost,low power consumption,and easy deployment.It is very conducive to popularization and has become one of the main choices for indoor positioning technology.However,because Bluetooth is susceptible to environmental factors and interference,it may cause errors in Bluetooth Signal Strength Indicator(RSSI)transitions.Although the Pedestrian Dead Reckoning(PDR)algorithm of inertial navigation can provide accurate and high-precision positioning,it is not suitable for long-term use due to the cumulative error of pedestrian dead reckoning.This thesis studies Bluetooth fingerprint positioning and inertial navigation PDR positioning technology.Based on Bluetooth indoor positioning,it can provide better positioning accuracy and stable positioning error.At the same time,with the assistance of inertial navigation technology,the fusion positioning of the two can greatly improve the positioning performance.First of all,research and analysis and use four filtering algorithms,mean filtering,median filtering,Gaussian filtering and Kalman filtering,to pre-process the Bluetooth RSSI fingerprint,so as to reduce the error caused by indoor environmental interference and improve the positioning accuracy.Secondly,three Bluetooth fingerprint matching algorithms,Nearest Neighbor(NN),K-Nearest Neighbor(KNN)and Weighted K-Nearest Neighbor(WKNN),are analyzed.Through the location experiment verification,it is determined that the optimal WKNN positioning algorithm is used,and at the same time,the fingerprint library using the Gaussian filtering algorithm for Bluetooth data preprocessing is determined.Then select PDR algorithm as the inertial navigation method,and study the positioning effect of PDR from three angles of gait recognition,step length estimation and heading estimation.Finally,a fusion positioning model based on Extended Kalman Filter(EKF)algorithm is established to achieve fusion positioning.The experimental results show that when the average positioning accuracy of fusion positioning reaches 0.89 m and the cumulative probability of positioning error reaches 90%,the error of fusion positioning reaches 1.30 m.Compared with Bluetooth fingerprint positioning and PDR positioning,the error reduction ratio reaches 18.75% and51.85%.In this thesis,the positioning method based on the fusion of Bluetooth positioning fingerprint and PDR makes full use of the advantages of the two technologies,making the final positioning result more consistent and less error,which improves the accuracy of indoor positioning,which is different from the rescue process.The requirements of indoor personnel for positioning are consistent and are highly feasible.