一种用于机场周边区域噪声指标预测的代理模型

A Surrogate Model for Noise Index Prediction in Areas Around Airports

摘要

航空噪声已成为制约民航可持续发展、危害周边居民身心健康的关键环境问题，传统噪声预测模型计算量大、场景适配性不足。本文构建基于 BP 神经网络的机场噪声代理预测模型，选取 B738 机型为研究对象，以飞机高度、飞行速度、航迹夹角、测点水平距离为输入，瞬时 A 声级为输出开展建模。依托机场凌晨时段航迹与噪声监测数据集完成样本预处理，设置 3×2×1 隐藏层神经元结构，经 40 轮迭代训练筛选最优模型。结果表明最优模型验证集均方误差为 7.91，预测相对偏差低于 10% 样本占比 95.43%，预测噪声分布与实测曲线吻合度高。基于瞬时 A 声级叠加计算计权等效连续感觉噪声级L_WECPN、年均昼夜等效声级YL_DN并实现可视化。该模型大幅降低计算成本，可快速模拟航迹调整、航班量变化等多场景噪声影响，能够为机场规划、噪声管控与扰民事件溯源提供量化技术支撑。

Abstract

Aircraft noise has become a critical environmental issue restricting the sustainable development of civil aviation and threatening the physical and mental health of nearby residents. Traditional noise prediction models suffer from high computational costs and poor adaptability to real-world scenarios. This study develops a surrogate noise prediction model based on the BP neural network, using the B738 aircraft as the research subject. The model takes flight altitude, speed, flight path angle, and horizontal distance from the measurement point as inputs, and outputs instantaneous A-weighted sound levels. Sample preprocessing is conducted using flight track and noise monitoring data collected during early morning hours at the airport. A hidden layer structure of 3×2×1 is adopted, and the optimal model is selected through 40 rounds of iterative training. Results show that the mean squared error of the optimal model on the validation set is 7.91, with 95.43% of predictions deviating by less than 10%, and predicted noise distributions closely match measured curves. Based on the instantaneous A-weighted sound levels, weighted equivalent continuous perceived noise level L_WECPN and annual average day-night equivalent sound level YL_DN are calculated and visualized. This model significantly reduces computational cost and enables rapid simulation of noise impacts under various scenarios such as flight path adjustments and changes in traffic volume, providing quantitative technical support for airport planning, noise control, and source tracing of noise complaints.

关键词

机场噪声 / BP神经网络 / 代理模型 / 噪声预测 / 航空噪声指标

Key words

airport noise / BP neural network / surrogate model / noise prediction / aviation noise indicators

参考文献

[1] H. Feng, Y. Zhou, W. Zeng, C. Ding. Review on metrics and prediction methods of civil aviation noise. International Journal of Aeronautical and Space Sciences 24(2023)1199-1213.

[2] AM. Nassur, M. Lefèvre, B. Laumon, D. Léger, AS. Evrard. Aircraft noise exposure and subjective sleep quality: the results of the DEBATS study in France. Behav Sleep Med 17(4)(2019)502-513

[3] S. Steven, K. Frenis, S. Kalinovic, et al. Exacerbation of adverse cardiovascular effects of aircraft noise in an animal model of arterial hypertension. Redox biology 34(2020)101515.

[4] A. Saucy, B. Schäffer, L. Tangermann, D. Vienneau, JM. Wunderli, M. Röösli M. Does night-time aircraft noise trigger mortality? A case-crossover study on 24,886 cardiovascular deaths. Eur Heart J 42(8)(2021)835-843

[5] T. Münzel, FP. Schmidt, S. Steven, J. Herzog, A. Daiber, M. Sørensen. Environmental noise and the cardiovascular system. J Am Coll Cardiol 71(6)(2018)688-697

[6] M. Klatte, J. Spilski, J. Mayerl, U. Möhler, T. Lachmann, K. Bergström. Effects of aircraft noise on reading and quality of life in primary school children in Germany: results from the NORAH study. Environ Behav 49(4)(2017)390-424

[7] FL. Friedt, JP. Cohen. Perception vs. reality: the aviation noise complaint effect on home prices. Transp Res Part D Transp Environ 100(2021)103011

[8] PJ. Wolfe, JL. Kramer, SR. Barrett. Current and future noise impacts of the UK hub airport. J Air Transp Manag 58(2017)91-99