基于CAPSO-BPNN的铁路信号运行状态预警方法研究  

Study on the Precautionary Method for Railway Signal Running State Based on CAPSO-BPNN

作　　者：喻喜平[1] YU Xi-ping(Wuhan Railway Vocational College of Technology,Wuhan 430205,China)

机构地区：[1]武汉铁路职业技术学院,湖北武汉430205

出　　处：《山东农业大学学报（自然科学版）》2019年第2期281-284,共4页Journal of Shandong Agricultural University：Natural Science Edition

摘　　要：为提高铁路信号运行状态预警的准确率,克服BPNN模型存在收敛速度慢和局部最优的缺点及其性能易受网络的初始权值、阈值等参数选择的影响,本文提出一种基于CAPSO-BPNN的铁路信号运行状态预警模型。结果表明,与PSO-BPNN和BPNN相比较,CAPSO-BPNN模型具有更高的预警准确率和更优的性能,为铁路信号运行状态预警提供了新的方法和途径。In order to improve the precautionary accuracy for railway signal running state and overcome disadvantages of slow convergence speed and local optimization in BPNN model,and its performance was easily affected by the initial weight value and threshold value and so on,this paper proposed a precautionary model based on CAPSO-BPNN for railway signal running state.The results showed that the CAPSO-BPNN model had higher accuracy and better performance than PSO-BPNN and BPNN,which provided new method and approach for forewarning of the operating state of railway signals.

关 键 词：铁路信号 云自适应粒子群优化算法 BP神经网络 运行状态 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]