概率神经网络多历元残差RAIM算法  

Probabilistic neural network multi-epoch residual RAIM algorithm

作　　者：武明 许承东[1] 黄国限 孙睿 鲁智威 WU Ming;XU Chengdong;HUANG Guoxian;SUN Rui;LU Zhiwei(School of Aerospace Engineering,Beijing Institute of Technology,Beijing 100081,China)

机构地区：[1]北京理工大学宇航学院,北京100081

出　　处：《系统工程与电子技术》2023年第12期3967-3974,共8页Systems Engineering and Electronics

摘　　要：民用航空领域中,为提高接收机自主完好性监测(receiver autonomous integrity monitoring,RAIM)算法对故障偏差的检测能力,降低最小可检测偏差,提出了一种概率神经网络多历元残差RAIM算法。该算法基于概率神经网络构建4层故障卫星检测模型,利用方差膨胀模型建立伪距残差故障类与无故障类训练样本,通过粒子群优化算法优化概率神经网络的平滑参数以满足误警率要求,从而计算输入多历元伪距残差与故障类和无故障类训练样本的相似程度,判断卫星是否出现故障。仿真实验结果表明,优化平滑参数可提高所提算法故障检测性能。相比加权最小二乘RAIM算法和高级RAIM(advanced RAIM,ARAIM)算法,所提算法在不同故障情况下可提高小伪距偏差的检测性能,降低不同故障情况下的最小可检测偏差。We propose a probabilistic neural network multi-epoch residual receiver autonomous integrity monitoring(RAIM)algorithm for civil aviation which can improve the detection capability of RAIM algorithm for fault deviation and reduce the minimum detectable deviation.A four-layer fault satellite detection model based on probabilistic neural network is constructed.Fault class and fault-free class training samples of pseudorange residuals are established using variance inflation model.The smoothing parameter of probabilistic neural network are optimized by particle swarm optimization algorithm to meet the false alarm rate.Thus,the similarity between the input multi-epoch pseudorange residual and the fault samples and fault-free samples can be calculated to determine whether the satellite is faulty.Simulation results suggest that optimizing the smoothing parameter can improve the fault detection ability of the proposed algorithm.Compared with the weighted least squares RAIM algorithm and the advanced RAIM(ARAIM)algorithm,the proposed algorithm can improve the detection performance of small pseudorange deviation and reduce the minimum detectable deviation under different fault conditions.

关 键 词：接收机自主完好性监测 故障检测 概率神经网络 方差膨胀 

分 类 号：V249.31[航空宇航科学与技术—飞行器设计]