一种顾及用户朝向的PSO-LSSVM指纹定位算法

A PSO-LSSVM Fingerprint Localization Algorithm Considering User’s Orientation

为了解决人体对WiFi信号遮蔽和最小二乘支持向量机参数优化的问题,提出了一种顾及用户朝向的粒子群优化最小二乘支持向量机指纹定位方法。建立全向指纹库,采用粒子群优化算法求出最小二乘支持向量机最优参数,通过最小二乘支持向量机训练出定位模型,将待测点指纹信息输入定位模型中,最终估算出待测点位置坐标。仿真实验结果表明所提算法在定位误差上达到0.72 m,普通的粒子群优化最小二乘支持向量机算法定位误差为0.84 m,提高了室内定位精度,具有实际的应用价值。

In order to solve the problem of the human body’s occlusion of WiFi signal and the parameter optimization of least squares support vector machine,a fingerprint localization method based on the particle swarm optimization algorithm for optimizing the parameters of least squares support vector machine which takes into account user’s orientation is proposed.The omnidirectional fingerprint database is established and particle swarm optimization algorithm is used to find the optimal parameters of the least squares support vector machine.Through particle swarm optimization and least squares support vector machine algorithm,the localization model is trained.The fingerprint information of the point to be measured is input into the localization model,and the localization coordinates of the pending point are estimated finally.The simulation results show that the proposed algorithm can reach 0.72 meters in localization error.The localization error of the ordinary particle swarm optimization combining with least squares support vector machine algorithm is 0.84 meters.The localization accuracy is improved,and it has practical value.