The Research And Implementation Of Data Fusion Indoor Positioning Technology Based On BLE And IMU

In the past decades,researchers have invented many indoor positioning algorithms,and people's movement and daily life are inseparable from the carrying of mobile terminals.Therefore,indoor pedestrian positioning methods for mobile terminals have been widely used by researchers.Such as trilateration,fingerprints positioning,pedestrian dead reckoning(PDR)and so on.However,various positioning technologies have their bottlenecks,and the current technology is still difficult to handle these positioning problems in indoor scenes.Fingerprint positioning and trilateration based on Received Signal Strength Index(RSSI)usually require sliding windows and median filtering to stabilize RSSI,which will cause signal filtering delays and lead to positioning delays,which is harmful for real-time positioning.The PDR in the smartmphone equipped with an inertial measurement unit(IMU)usually requires that the Y-axis direction of the smartphone is parallel to the direction of motion in order to obtain a more reliable estimated positioning result.However,pedestrians always freely swing their arms back and forth during normal walking,which is completely ignored in traditional PDR-related studies.In this case,the step detection algorithm in the traditional PDR method will face huge challenges,and the deviation of the heading estimation algorithm is also inevitable.These defects will make the traditional PDR unsuitable when pedestrians walk freely.In view of the above problems,this thesis studies how to integrate the two information sources of BLE and IMU to obtain stable and reliable positioning results.The main contents of this article are as follows:Firstly,this thesis study the widely researched information sources including IMU,WLAN,Bluetooth,UWB,Lidar and vision and explain why this thesis uses these two kinds of information source for data fusion.Then the traditional RSSI-based triangulation and fingerprint positioning algorithm details and the dead reckoning system using IMU are described in detail.Finally,this thesis innovatively proposes a data fusion positioning algorithm based on Bluetooth RSSI and IMU measurements.In the BLE positioning module of this algorithm,this thesis optimizes the performance of BLE positioning by smoothing the RSSI signal,using the area information of the BLE layout to support the positioning cell selection,and using weighted multilateration positioning to improve trilateration.In the PDR improvement part of this algorithm,this thesis analyzes the characteristics of walking posture and innovatively proposes a new method of step detection.Then the heading estimation module is improved by using the result of step detection as synchronization signal to deeply couple with the BLE positioning result.These improvement will significantly improve the robustness of the PDR.In the data fusion part,this thesis linearizes the nonlinear step noises,and uses the Extended Kalman Filter(EKF)to fuse the data of each module of BLE positioning and PDR,which effectively reduces the positioning delay and positioning error of indoor positioning services.Experiments show that compared with the traditional triangulation positioning algorithm,the weighted multilateration algorithm proposed in this thesis can reduce the static positioning error by 52.36%,but this will increase the calculation time to 433.43%of the traditional algorithm.Since the time cost of traditional algorithm is almost microsecond level,this multiple increase in computational complexity has little impact on system performance.Compared with the BLE positioning algorithm proposed in this thesis,the dynamic performance of the data fusion positioning algorithm based on EKF is better,which reduces the positioning error,route deviation and filtering delay by 11.76%,14.00%and 13.98%,respectively.Effectively improve the service quality of indoor positioning applications.In practice,this thesis apply the fusion positioning algorithm to several large underground parking lots,and fully realize commercial use on WeChat application.Among them,the largest positioning scene has a total area of 600,000m2.Our indoor positioning system has successfully provided car-finding navigation services in this scenario for a year,and this commercial positioning scenario further proves the accuracy and robustness of the algorithm in this thesis.