Research On Fusion Positioning System Based On Ultra-Wide Band And Bluetooth Technology

Indoor positioning technology has always played a vital role in people’s everyday lives.While Ultra Wide Band(UWB)positioning systems offer high accuracy,they come with higher costs.On the other hand,Bluetooth positioning systems are more cost-effective but provide lower accuracy.Addressing the challenge of using a single sensor to simultaneously meet high accuracy and low-cost requirements in complex indoor environments,this study conducts in-depth research on a fusion positioning system based on UWB and Bluetooth technologies.The research covers various aspects,such as the Received Signal Strength Indication(RSSI)ranging method for Bluetooth,Bluetooth positioning algorithms,and the fusion positioning model combining UWB and Bluetooth.The main research findings are as follows:1.We propose a Received Signal Strength Indication(RSSI)ranging method based on Dynamic Adaptive Sparrow Search Algorithm(DASSA)optimized Back Propagation(BP)neural network,called DASSA-BP.This method addresses the issue of larger errors encountered when using the path loss model for RSSI ranging.By utilizing the DASSA,the BP neural network’s weights and thresholds are adjusted dynamically to avoid getting stuck in local minima.Experimental results demonstrate that,compared to other ranging methods,the RSSI ranging method based on the DASSA-BP neural network model has a more concentrated error distribution and higher interference resistance in complex environments.2.We propose a Minimum Maximum(MinMax)localization algorithm based on the Two-Step Residual Network and K-Nearest Neighbor(TSRK)called TSRK-MinMax,which addresses the issue of larger localization accuracy errors in traditional MinMax algorithms due to fluctuations in maximum and minimum boundary values when calculating tag coordinates.The algorithm first uses the K-nearest neighbor method to select base stations near the tag.It then calculates the boundary points using the MinMax method,and derives the tag coordinates by weighting the residual between the distance from the boundary points to the base stations and the RSSI ranging values.Finally,it filters the coordinates with the smallest residual between the distance from the tag coordinates to the base stations and the RSSI ranging values as the final location of the tag.Experimental results show that the TSRK-MinMax localization algorithm improves the overall accuracy by 11.7% compared to the traditional MinMax localization algorithm.3.We propose a fusion positioning model combining Ultra Wide Band(UWB)and Bluetooth technologies,which caters to the varying positioning accuracy requirements in different areas within complex indoor environments.This model employs a precise partitioning strategy to divide the positioning area into high-precision positioning zones,regular positioning zones,and transition zones,with different positioning algorithms being utilized in each zone for combined positioning.To address the data fusion issue between UWB and Bluetooth in the transition zones,we propose a UWB and Bluetooth residual-weighted fusion positioning algorithm.Simulation experiments show that the proposed algorithm offers higher positioning accuracy and stability compared to the original Bluetooth positioning algorithm.4.To validate the effectiveness of the proposed UWB and Bluetooth hybrid positioning model,we have established an experimental platform for the fusion positioning system.In real-world environments,the proposed hybrid positioning system demonstrates higher accuracy and robustness compared to the Bluetooth-only positioning system.It ensures the required positioning accuracy in high-precision zones,while reducing the overall system cost compared to a UWB only positioning system.