Research On Indoor Location Algorithm Based On Bluetooth AOA/INS Fusion

With the extensive construction of global digital infrastructure and the vigorous development of the industrial Internet,hundreds of millions of intelligent terminal devices are gradually integrated into public life,and the accompanying demand for Location Based Services is also increasing exponentially.The mature Global Navigation Satellite System has long provided people with highly reliable and accurate outdoor positioning services.However,it cannot meet the indoor positioning needs that occupy most of human activity time.Therefore,finding an indoor positioning method with more advantages has become a current research hotspot.Most indoor intelligent terminals integrate Bluetooth and inertial sensor modules,both of them meet the advantages.Relevant indoor positioning methods have been supported by mature theoretical systems.Therefore,this article conducts research on indoor positioning technology that integrates Bluetooth AO A and INS.The main contents and innovations are as follows:(1)In order to ensure that the single base station can provide accurate Bluetooth AOA estimation results,this paper analyzes the error interference problem in Bluetooth AOA estimation.Firstly,a Bluetooth 5.1 packet communication model containing direction finding information is established,and then the error models that affect the AOA accuracy are modeled from three aspects:frequency,array,and channel.To compensate for the above errors,a weighted phase interferometry(WPI)based on grouping nonlinear least square fitting(GNLS)and a channel error correlation suppression method based on reverse compensation are proposed.The experimental results show that the accuracy of the proposed algorithm is improved by 36.5%compared to traditional phase interferometry under single channel conditions,and the average error of multi-channel estimation is reduced by 25.6%after reverse compensation.The experimental results show that the proposed algorithm effectively reduces the impact of various errors in AOA estimation.(2)In order to improve the heading measurement accuracy of inertial navigation systems,this paper proposes an improved Mahony complementary filtering algorithm to solve the problem that traditional complementary filtering algorithms cannot identify abnormal states of reference vectors,resulting in a decrease in the accuracy of attitude data.By setting a correction threshold for sensor data,the error states are identified,thereby filtering abnormal data,and improving the robustness of attitude calculation.Simulation of actual interference scenarios has verified that the algorithm in this paper effectively suppresses the heading error caused by abnormal acceleration and magnetic field strength,enhances the robustness of the attitude solution system,and provides more reliable heading data for INS.(3)This paper analyzes the positioning error distribution of a dual base station Bluetooth AOA positioning system through accuracy factors,and finds that there is a problem where the error of a single coordinate axis approaches infinity near the base station connection line.In order to effectively utilize the positioning results of other coordinate axes and the effective information of other positioning regions,an adaptive EKF based on ADOP is proposed for Bluetooth AOA/INS position fusion.The advantages of smooth INS data and high precision positioning of Bluetooth AOA complement each other,while eliminating the problem of uneven distribution of INS cumulative error and uneven distribution of AOA errors.Finally,the combined algorithm in this paper in an indoor experimental scene improves the positioning accuracy by 65.3%and 73.8%compared with the positioning accuracy of AOA and INS alone,and achieves high precision indoor positioning.