基于核岭回归和粒子滤波的室内移动目标追踪算法研究

本文提出了一种基于核岭回归和粒子滤波的室内移动目标追踪算法,该算法在离线阶段采用核岭回归方法提取传感器之间的距离与RSSI(Received Signal Strength Indicator)信号值之间的非线性关系,从而训练出一种非线性回归距离模型;在线追踪阶段,利用非线性回归模型和粒子滤波算法实现室内移动目标的定位和追踪。本文在典型的室内办公环境下进行实验,并通过MATLAB对实测数据进行仿真。实验结果表明,相比WKNN算法和KF算法,本文所提出的算法能到达更好的定位精度,误差均值为1.2743 m。

Research on Moving Target Tracking Algorithm Based on Kernel Ridge Regression and Particle Filter

A based on kernel ridge regression and particle filter indoor moving target tracking algorithm is proposed in this paper. In the off-line phase, use the Kernel Ridge Regression algorithm to extract the relationship between the sensors distance and the Received Signal Strength indicators(RSSI), so that to get a nonlinear regression distance model. In the online tracking stage, nonlinear regression model and particle filter algorithm are used to track indoor moving targets. The experiment is carried out in a typical indoor office environment, and the measured data were simulated by MATLAB. The experimental results show that, compared with WKNN algorithm and KF algorithm, the algorithm proposed in this paper can achieve better positioning accuracy, and the algorithm error mean is 1.2743 m.