基于预测误差格型滤波器的航迹关联算法研究

A Track Association Algorithm Based on Lattice Predictor

在K近邻域(K-NN)航迹关联算法的基础上,提出一种基于预测误差格型滤波器的关联算法,当融合中心仅获得状态估计值,而没有状态估计协方差矩阵时,将状态估计值通过格型滤波器,利用得到的预测误差代替协方差,为判决门限的选择提供依据,仿真结果表明,这种算法是有效的。

For multi-sensor tracking systems,track association is a crucial operation.A method based on K-NN algorithm is proposed.When only state estimation is available in the fusion center,these values are processed using a lattice predictor to get predicted errors.They act as the references for critical values.Simulation results illustrate that this algorithm is valid.