基于二次多项式运动建模的WSN目标跟踪预测

针对无线传感器网络(WSN)目标跟踪线性预测模型误差较大的问题,提出一种基于二次多项式运动建模的WSN目标跟踪预测新方法(PQPMM),该方法根据机动性目标运动学原理建立预测模型,利用最小二乘法拟合目标定位坐标、定位时间的二次多项式函数来逼近目标运动模型.结果表明,PQPMM方法的总体预测准确度相比线性预测法明显提高,当拟合点数N=14时,PQPMbl方法均方根误差RMSE比线性预测法减小53%.

Target tracking prediction in WSN based on quadratic polynomial motion modeling

Aiming at the problem of large prediction model error for target tracking in wireless sensor networks (WSN) based on linear prediction method, a novel method of target tracking prediction in WSN based on quadratic polynomial motion modeling (PQPMM) is proposed. According to the method, the prediction model is built for maneuvering target based on kinematics theory. The quadratic polynomial function of estimated target's coordinates and localization time is fitted using least square method and the function can be used for approximating target's motion model. Experimental results show that the general prediction accuracy is improved highly using PQPMM method compared with linear prediction method. When number of target's fitting locations equals to 14, RMSE of PQPMM method decreased by 53% in compare with linear prediction method.