Research On The Reliability Model Of Civil Aviation Passenger Information System Based On Grey Neural Network

To ensure the civil aviation passenger information system be safely and trouble-free operated every week which is related to the vital interests of airlines and passengers' travel experience.However,because a series of unreliable factors,the civil aviation passenger information system at home and abroad frequently broke down.Therefore,how to accurately evaluate the reliability of the system and maintain the normal operation of the system,which is a very urgent task.This paper attempts to analyze a large number of the operation and maintenance monitored data produced by the civil aviation passenger information system,through the keyword extraction and semantic parsing of regular expressions,to extract the reliability failure data and build the model of software reliability of neural networks,which could be accuratly predict the invalid data,to reduce the potential uncertainties in reliability factors,thus to improve the Accuracy of prediction.The main works is as follows:First of all,through analysis and research the extraction method about reliability data,designed a reliability data extraction method to civil aviation passenger information system data based on regular expressions,meanwhile realized this extraction method.Secondly,established two kinds of reliability model about civil aviation passenger information system,respectively are the BP neural network model and Elman neural network model.Furthermore,utilize the established reliability model to predict failure reliability data produced by civil aviation passenger information system,and verified the feasibility of the neural network reliability model in the prediction of the civil aviation passenger information system reliability.It can solve the applicability problem of the model.Thirdly,according to the experimental verification about the two models proposed above,it is found that the data in the multi-user software system have the characteristics of strong randomness and dynamics,Therefore,proposed a gray neural network model to predict the software reliability based combined gray model and Elman neural network.Finally,according to the evaluation criteria of software reliability model,compared the results of the three prediction models.The results show that the software reliability model based on gray Elman neural network has a better prediction accuracy and applicability than traditional models.To a greater extent,this model could improve the accuracy of the software reliability prediction.