多传感器协同辨识自校正加权观测融合Kalman滤波器

对于带未知噪声统计的多传感器系统，利用最小二乘法将观测方程统一处理，形成新 的跟踪系统，处理后的观测结果之差可以产生多组新的白噪声序列，利用各组白噪声的相关 函数阵解矩阵方程组，可解得各传感器观测噪声方差〖WTHX〗R〖WTBX〗i。通过状态 方程和观测方 程以及观测噪声估值，利用相关函数，可求得〖WTHX〗ΓQ〖WTBX〗w〖WTHX〗Γ〖WTBX 〗 T 的估计，进而得到自校正 加权观测融合Kalman滤波器。一个带有3传感器目标跟踪系统的仿真例子说明了其收敛速度 快，估计精确等特点。

Self-tuning Weighted Measurement Fusion Kalman Filter with Cooperating Identification for Multisensor System

For the multisensor system with unknown noise statistics, the measurement function can be dealt with in a unified way to form a new tracking system by least square method. The differences between these measurements that dealt with many group of new white noise sequences. Using the correlated functions matrix of these sequences, the measurement noise variances Ri of the subsystems can be estimated. And the estimates of ΓQwΓT can be obtained from the state functions, the measurement function and the estimates of the measurements noise variances by correlated functions matrix and then the self-tuning weighted measurement fusion Kalman filter is obtained. A simulation example for a tracking system with 3 sensors shows its fast convergence and exactness.