Temperature Prediction Model Based On Time Series For Airport Runway

In recent years, with the rapidly development of the Civil Aviation, travel by air has become the important way in daily life. However, the number of aviation accide-nt is growing at the same time. Especially under the weather of snow and ice, the r-unway need to be close and lead to the plan delay directly. Therefore, it need to acc-urately predict the runway iced by technology in order to avoid or reduce the closing time under bad weather. It is to say that the change for runway temperature in few hours or few minutes have important value for whether to temporarily shut down the runway, when to close and take equipment onto runway.The temperature data in this topic is based on the values of the measurement sys-tem of wireless sensor network in Dalian airport, which uses the microelectronics sen-sor node, transmission network for wireless data and system software for server point information, it can collect real-time data of runway pavement meteorological. In this paper, first we get the system data, and analysis the encrypted data, then process the data between December2010and April2011and stored in the Excel and Access data base in different time interval, also established the visualization interface under VC++6.0. Go over current situation of data mining, and meticulous analysis on the time serie-s of data mining, then according to the modular design method, analysis and design t-he whole system. Set up each ARMA forecast model with48times temperature fall down to zero. Compared the forecast data and the actual data from the temperature se-nsor, then summarized how to choose the standard model in different cases. By choo-sing the different time interval and different number of data that the mean error can be achieved within1℃, and the biggest deviation on single value also within2℃. So it can be said that this system achieved the original expectation, and meet the basic require of temperature prediction.Through comparison with the meteorological data from Dalian airport in2010. Th-e accuracy rate within two hours is more than75%when the temperature down to t-he zero and the error is less than0.5℃, and if the prediction error is less than1℃, the accuracy at about85%. So it avoid the grave threat of the bad weather conditions witch lead to the runway freezes or water, and it reduce the impact of takeoff and1 -anding, that’s to say this topic has important application value in the airport flight sa-fety area.