基于神经网络数据融合的目标跟踪简化算法

分析了基于神经网络数据融合的目标跟踪算法 ,指出了传统的融合算法计算量大 ,神经网络目标向量不易选取等缺点 ,并提出了一种简化的算法。应用理论分析和蒙特卡洛仿真方法 ,对标准卡尔曼滤波算法和简化的滤波算法进行了比较 ,并给出了均方根误差的统计值。该简化算法原理简单 ,数据处理量小 ,速度快 ,误差小 ,特别适用于多传感器的处理 ,将融合结果反馈给单传感器 ,可提高各单传感器的跟踪精度

A Simplified Algorithm of Target Tracking Based on Neural Network Data Fusion

In this paper, we analyse an algorithm of target tracking based on neural network data fusion, explain that the traditional fusion algorithm has some shortcomings such as heavy calculation burden and the difficult selection of target vectors of neural network, and present a simplified algorithm. The theoretical analysis and Monte Carlo simulation methods are used to compare the traditional fusion algorithm with the new one. The simplified algorithm is more simple in principle, ess in data, faster in processing and less in error. It is suitable for the multisensors. The feedback of the fusion result to the single sensor can enhance the single sensor's precision.