针对不同型号的智能移动终端设备，由软件和硬件的异构性而导致不同设备在同一采集点处采集到的同一蓝牙源接入点的蓝牙信号强度观测值存在显著差异而影响定位精度的问题，该文提出了一种基于混沌麻雀搜索算法优化BP神经网络模型的室内蓝牙RSSI(received signal strength indicator)标定算法。该标定算法应用混沌麻雀搜索算法的全局搜索能力和快速收敛性来帮助BPNN模型选取最优的初始权值和阈值。实验结果表明:该标定方法得到的平均RSSI误差相较于未标定之前降低了 87。6％，有效地降低了软硬件异构性对采集到的蓝牙信号强度观测值的精度。
The Indoor Bluetooth RSSI Calibration Method Based on Chaotic Sparrow Search Algorithm
Due to the heterogeneity of software and hardware of different types of intelligent mobile terminal devices,there are significant differences in the observed Bluetooth signal strength of the same Bluetooth source access point collected by different devices at the same collection point,which affects positioning accuracy.The indoor Bluetooth RSSI calibration algorithm based on the chaotic sparrow search algorithm optimizing the BP neural network model is proposed.This calibration algorithm uses the global search capability and fast convergence of the Chaos Sparrow Search algorithm to help the BPNN model select the optimal initial weights and thresholds.Experimental results show that the average RSSI error obtained by this calibration method is reduced by 87.6％compared with that before calibration,effectively reducing the impact of software and hardware heterogeneity on the accuracy of collected Bluetooth signal strength observations.
software and hardware heterogeneityBP neural networkchaotic sparrow search algorithmbluetooth RSSI calibration model
万方数据
维普
