Research On Airport Noise Prediction For Multi-runway Airport Flight Delay Recovery

In recent years, with the rapid growth of civil aviation passenger and freight transport volume, many domestic airports increase the capacity of the airport by increasing the number of runway and the expanding of the scale of the airport. The increase in throughput also causes in the increase of the uncertainty factors in the use of the runway and the aggravation of the airport noise impact. With the increasing number of multi-runway airport, the combinations of flights, runways and flight procedures are more diverse and noise distribution is different, so that the noise prediction methods of multi-runway airports are different and more complex. Current prediction methods are mainly for the prediction of single runway single flight event. However, the airport noise has the characteristics of high sound level, wide range affected and instability. For the "short time high noise" phenomenon in the flight delay recovery time, the prediction method has a large error, which is difficult to put into use. Therefore, it is necessary to study the prediction of multi-runway airports noise in the flight delay recovery time.Current prediction methods are mainly for the prediction of single runway single flight event. For the aggravation of the airport noise impact because of current airport's construction, expansion, the runway's increase and the intensive flights' take off in the flight delay recovery time, the concept of "noise equivalent" is presented in this paper. In order to get flights and runways equivalently matched and reduce the uncertainties of noise prediction in the flight delay recovery time, an improved hybrid clustering algorithm with constraints is proposed to construct a clustering model based on airport noise equivalent. Besides, in order to predict the average noise energy in a certain period of time and provide basis and reference for the competent department, an airport noise equivalent prediction model is constructed by using the BP neural network and the NPD curves interpolation method. Experimental results show that the proposed prediction model can improve the prediction effect significantly and make the effective prediction of multi-runway airport noise in the flight delay recovery time.The noise distribution around the airport is mainly controlled by the track, and the noise is mainly affected by aircraft type and flight parameters. In order to make it more convenient and efficient to predict and evaluate the impact of multi-runway airport noise in the flight delay recovery time from the overall situation, the data of the center track of each cluster and the representative models of each cluster are input into INM(Integrated Noise Models) to calculate the noise value and create the noise database, which is based on different track clustering and aircraft type clustering. A multi-runway airport noise prediction model based on Bayesian classification is proposed by Bayesian classification algorithm. Once the data of flight number, aircraft type, track, destination and departure point are input into the model, the noise prediction results can be quickly obtained. Experimental results show that the proposed model can not only predict the noise impact on the normal time of flight release, but also can be used to predict the noise value of the sensitive points around the airport during the time of flight delay recovery within a certain error range.