基于样条模型的多传感器目标跟踪算法研究

研究武器试验中多传感器目标跟踪的数据融合问题。通过用样条函数表示目标轨迹参数,建立了问题的节省参数模型,并构造了快速的样条递推算法和参数估计算法。采用样条函数这种稀疏表示方法后,新方法能有效地减少待估参数的个数,提高参数估计精度,在减少计算量的同时降低处理系统的存储负担。仿真结果说明新算法能在很短的时间内完成数据融合,且得到的融合结果精度很高。

Algorithm Research for Multi-sensor Target Tracking Based on Spline Model

This paper discussed the problem of data fusion for multi-sensor target tracking in weapon test. It built a reduced parameter model for problem by representing the track of target by spline, and got a fast algorithm for spline iteration and parameter estimation. By using sparse spine to represent the targets' tracks, the new model can reduce the number of parameters availably, which can decrease the computation cost and debase the burden of memory. Simulation results confirmed that the proposed algorithm could finish data fusion in very short time, and has high accuracy.