本文關(guān)鍵詞： 重尾分布 中心差分法 幅值信息 極大似然估計(jì) 虛警 非線性系統(tǒng) OSPA距離 信雜比　出處：《北京航空航天大學(xué)學(xué)報(bào)》2017年03期 　論文類型：期刊論文

【摘要】：針對(duì)低信雜比環(huán)境下的多機(jī)動(dòng)目標(biāo)跟蹤問(wèn)題,提出了一種基于極大似然(ML)背景參數(shù)估計(jì)的中心差分卡爾曼-勢(shì)概率假設(shè)密度濾波(BE-CDKF-CPHD)算法。算法采用ML法實(shí)時(shí)估計(jì)重尾分布模型參數(shù),計(jì)算檢測(cè)概率和虛警概率。運(yùn)用極大似然-恒虛警(MLCFAR)算法對(duì)信號(hào)進(jìn)行處理,提取有效量測(cè)值,將幅值似然函數(shù)與勢(shì)概率假設(shè)密度濾波器(CPHD)中的目標(biāo)位置似然函數(shù)相結(jié)合,通過(guò)中心差分法遞歸更新得到后驗(yàn)均值與協(xié)方差,達(dá)到對(duì)多機(jī)動(dòng)目標(biāo)進(jìn)行跟蹤的目的。仿真結(jié)果表明,在低信雜比環(huán)境中,所提算法提高了跟蹤精度與目標(biāo)數(shù)目估計(jì)準(zhǔn)確度。
[Abstract]:According to the low signal to clutter ratio tracking problem of multiple maneuvering target environment, propose a method based on maximum likelihood (ML) estimation of background parameters of the central difference Calman Cardinalized probability hypothesis density filter (BE-CDKF-CPHD) algorithm. The algorithm uses the ML method to estimate the heavy tailed distribution of model parameters, calculate the detection probability and false alarm probability. The use of maximum likelihood - CFAR (MLCFAR) algorithm for signal processing, extraction of effective measurements, the amplitude of the likelihood function and potential probability hypothesis density filter (CPHD) target position in the combination of the likelihood function, the central difference method recursive update posterior mean and covariance, achieve the goal of tracking for multiple maneuvering targets. The simulation results show that in the low SNR environment, the proposed algorithm improves the tracking precision and the target number estimation accuracy.

【作者單位】： 西北工業(yè)大學(xué)自動(dòng)化學(xué)院;
【基金】：航空科學(xué)基金(20152853029)~~
【分類號(hào)】：TN713

【參考文獻(xiàn)】

相關(guān)期刊論文 前4條

1 陳里銘;陳U，

本文編號(hào)：1535146