基于RBF神经网络的BDS接收机作战效能评估  

Operational effectiveness evaluation of BDS receivers based on RBF neural network

作　　者：李奎 张侹[1,2] 王华 廖斌[1,2] 吴娟 LI Kui;ZHANG Ting;WANG Hua;LIAO Bin;WU Juan(State Key Laboratory of Geo-Information Engineering,Xi’an 710054,China;Xi’an Research Institute of Surveying and Mapping,Xi’an 710054,China)

机构地区：[1]地理信息工程国家重点实验室,西安710054 [2]西安测绘研究所,西安710054

出　　处：《导航定位学报》2023年第6期57-63,共7页Journal of Navigation and Positioning

摘　　要：针对传统作战效能评估方法存在主观性强,依赖专家经验等问题,提出一种基于径向基函数(RBF)神经网络的北斗卫星导航系统(BDS)接收机作战效能评估方法:梳理BDS接收机作战效能评估需求,并构建BDS接收机作战效能评估指标体系;然后对网络基础原理、学习算法和评估流程进行研究。实验结果表明,提出的方法能够有效完成BDS用户机的作战效能评估,虽然RBF神经网络在收敛速度上比反向传播(BP)神经网络要慢5.52倍,但是损失函数和准确率相比BP神经网络分别提升了65.7%和8%,而且与传统的装备作战效能评估算法相比,评估结论更加客观,具有一定实用性。Aiming at the problem that the traditional operational effectiveness evaluation methods have defects such as over-reliance on expert experience and strong subjectivity,the paper proposed an operational effectiveness evaluation method of BeiDou navigation satellite system(BDS)receivers based on radial basis function(RBF)neural network:the operational effectiveness evaluation requirements of BDS receivers were sorted out,and the operational effectiveness evaluation index system was built;then,the basic principle,learning algorithm and evaluation process of network were studied.Experiment result showed that the proposed method could be effectively competent for effectiveness evaluation.Although the convergence speed of RBF neural network would be 5.52 times slower than that of the back propagation(BP)neural network,the loss function value and accuracy rate could be improved by 65.7%and 8%,respectively,compared with that of the BP neural network;moreover,compared with the traditional methods,the adopted method could have certain practicability by more objective evaluation conclusion.

关 键 词：北斗卫星导航系统(BDS)接收机 作战效能 径向基函数(RBF)神经网络 评估指标体系 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] P228[自动化与计算机技术—计算机科学与技术]