Font Size: a A A

The Forecast Of Unsafe Events For Civil Aviation

Posted on:2020-08-10Degree:MasterType:Thesis
Country:ChinaCandidate:N GuoFull Text:PDF
GTID:2381330596494449Subject:Safety science and engineering
Abstract/Summary:PDF Full Text Request
Under the background of the rapid development of the civil aviation industry and the rapid growth of flight volume,the uncertain factors affecting the development of civil aviation are increasing,and civil aviation safety faces enormous challenges.Forecasting the monthly data of civil aviation unsafe events in China timely and accurately is important to set up the safety performance targets,which are suitable for civil aviation of China.It is of great significance to promote safety performance management of civil aviation.Firstly,research the fluctuation characteristics of civil aviation unsafe events in China.Use the seasonality adjustment method and HP filtering method to analyze the data feature of trend,seasonality,volatility and distribution characteristics.Then,on the base of 2007-2016 unsafe events monthly data,use the ARIMA model and GRNN model to build the ARIMA-GRNN combined forecasting model.The combined model is used to predict monthly unsafe events for civil aviation of China.The rationality of the model was tested by the monthly data of unsafe events in 2017-2018.The prediction accuracy of the constructed model is compared with the single prediction models and the other combined forecasting models.The average prediction accuracy of the model is 11.72% higher than the ARIMA model,and is 6.77% higher than the GRNN model.To compared with the ARIMA+ LS-SVM combination model and the ARIMA+LS-SVM+BP combination model,the ARIMA-GRNN combined forecasting model increase by 8.49% and 5.85%.Finally,the ARIMA-GRNN combined forecasting model is used to predict the number and the number of unsafe events per ten thousand aircrafts of China's civil aviation from January 2019 to December 2020.The predicted results of the number of unsafe events are: 958?895?967?1254?1483?884?1299?1086?1879?1542?936?791?1096?870?963?1019?1283?885?937?1694?2017?1802?1145?968.And the predicted results of the number of unsafe events per ten thousand aircrafts are: 25.16?25.43?25.87?30.00?38.95?26.49?32.12?26.51?45.44?38.40?24.69?23.32?30.29?21.83?22.84?24.66?30.40?19.60?21.32?38.14?48.17?41.26?27.24?22.17.
Keywords/Search Tags:civil aviation, unsafe events, monthly forecast, ARIMA, GRNN
PDF Full Text Request
Related items