火炮发射与控制学报

为提高炮口磁场数据的测量精度，通过多传感器组合阵列对炮口磁场进行数据采集以减小随机干扰误差。基于多传感器阵列测量模型，设计了炮口多路磁场信号的卡尔曼融合处理方法。该方法通过BP神经网络对数据拟合得到炮口磁场信号的数学模型，使用卡尔曼滤波算法对炮口磁场数据进行滤波，在滤波的基础上，针对各传感器的空间分布特点，利用最小均方算法自适应调整最优加权因子，进而通过卡尔曼融合获得最优估计值。仿真实验表明，多传感器卡尔曼融合后数据与测量数据相比，误差减小了9.5%,因而将该方法用于炮口磁场信号处理是可行的。

In order to improve the measurement accuracy of muzzle magnetic field data, a multi-sensor combination array is used to collect the data to reduce random interference error. Based on the multi-sensor array measurement model, a Kalman fusion processing method of multiple magnetic field signals at the muzzle is designed. This method uses the BP neural network to fit the data to obtain a mathema-tical model of the muzzle magnetic field signals. By using the Kalman filtering algorithm to filter the data, the optimal weighting factor is adaptively adjusted on its basis by the Least Mean Square(LMS) algorithm according to the spatial distribution characteristics of each sensor. Then the optimal estimation is obtained by Kalman fusion. Simulated experiments show that the multi-sensor Kalman fusion data has 9.5% less error than the measured data, so it is feasible to use this method for muzzle magnetic field signal processing.