频域徙动运动参数闭频繁项集挖掘算法

Closed Frequent Itemsets Mining Algorithm of Frequency Domain Migration Movement Parameters

高速多目标运动状态监测过程,运动参数的频域徙动导致参数估计和挖掘困难,传统方法对高速频率徙动运动目标参数的挖掘采用MapReduce框架的并行FP-Growth算法,算法需要对运动目标的速度和加速度进行预估计,实现困难。基于多普勒扩散的项集期望支持数模型,提出一种改进的基于闭频繁项集挖掘的高速多目标的运动参数挖掘算法,构建高速多目标运动参数信号模型,采用普勒频率模糊数搜索的方法完成高速多目标的频域徙动动态平滑,准确挖掘出运动参数的相位、时延、速度和频率等相关信息。研究结果表明,该算法能准确拟合时延、速度等运动参数,拟合值与真实值相同,对高速运动目标的运动参数估计精确,在高速运动目标参数挖掘和精确制导等方面具有较高的应用价值。

In high speed multi target motion state monitoring process, the frequency domain migration movement of moving parameters is difficult to be estimated, traditional method takes parallel FP-Growth algorithm of MapReduce framework, it needs speed and acceleration parameters pre estimation, it is difficult to realize. On the basis of itemsets support model of Doppler diffusion, an improved algorithm for mining closed frequent motion parameters of high speed multi targets is pro-posed, high speed multi parameters of moving target signal model is constructed, and Doppler frequency fuzzy number search is used for frequency domain migration speed multi-objective dynamic smooth. The movement parameters such as phase, time delay, speed and frequency and other related information are mined accurately. The simulation results show that it can estimate the motion parameter accurately, and the fitted values and true values are the same, the result is accu-rate, it has high application value in the high speed moving target parameters mining and accurate guidance.