Design And Implementation Of Short-term Passenger Rail Predictive Analysis System

Railway transportation has become the preferred means for citizens to go out.It is playing the important role in urban public transport.It can not only improve efficiency of the traveler significantly,but also relieve the ground-based traffic congestion.However,with the continuous expansion of the scale of urban rail transit network construction in Shanghai,passengers'travel path is becoming more diverse,and the spatial distribution of passenger flow is also becoming more prominent.Therefore,when government provided the convenience for the public,the difficulty of the relevant departments which can manage large passenger flow was increased.At present,to prevent and predict like‘the 2014 the Bund Shanghai trampled'and other large passenger events occur is a matter of great importance to the relevant departments.To solve above problems,this paper has taken some Research Institute of engineering big data innovation project as background,combined with the regional flow monitoring and early warning system for the actual project,based on the problem of traditional forecasting methods,I proposed the method that is a seasonal time series SARIMA model of passenger flow forecasting^[10].I design and implement a short-term traffic forecasting system.The system is able to predict a 15-minute traffic conditions,you can see the tide of history and passenger migrations,alarm parameters can be adjusted and the setting of early warning system can adapt the forecast demand for different time periods.After system testing and on-line operation for 6 months showed that the system is feasible and effective.The main contents and work as follows:?1?Through the analysis of historical data which was collected by the AFC system of Shanghai rail transit system and the OIDD signaling data,the basic statistical characteristics are obtained:the short-term passenger flow forecast algorithm is studied,and the daily passenger flow is a week,the overall performance is nonlinear and non-stationary.^[9]?2?The system thresholds and monitoring alarm setting:warning level can be related to values,the number of thresholds and other regions to adjust.Once the abnormal situation occurs,the first time to notify the administrator,and targeted early warning measures to prevent the occurrence of danger.?3?Visual display of the passenger flow forecast system:the number of people in various regions is displayed directly on the map by real-time statistics.Through the use of ECharts,the report shows the trend of statistical analysis,so as to achieve the effect of visualization.Finally,the system performance has been analyzed,which proves that the system has practical application value.