随着科技的发展与时代的进步，能源消耗加剧，人们的环保意识逐渐减弱，城市污染日益严重。在综合国力飞速前进的同时，许多环境问题接踵而至。其中最具代表性的就是城市雾霾问题，雾霾是一种危害人体健康的物质，它会危害人体呼吸道从而导致多种呼吸道疾病。因此，治霾防霾就显得尤为重要，这是一个任重而道远的过程，当下并没有较为有效的方法来彻底解决雾霾污染问题。因此本文提出了BP神经网络和以深度学习为基础的长短时记忆网络来预测雾霾天气情况，并且对模型做出了优化，使其能够更加有效准确地预测。通过引入各项具有时间序列特性的数据，例如大气污染物、不同空间地理上检测出的影响因子要素和气象因素等，经过LSTM模型的运算整合，形成波形图直观的显示出未来一段时间内的雾霾天气情况。
With the development of science and technology and the progress of the times, energy consumption has intensified, people’s awareness of environmental protection has gradually weakened, and urban pollution has become increasingly serious. While the overall national strength is advancing rapidly, many environmental problems have followed one after another. The most representative one is the problem of urban smog. Smog is a substance that is harmful to human health. It will harm the human respiratory tract and cause a variety of respiratory diseases. Therefore, the treatment of smog and the prevention of smog are particularly important. This is a process with a long way to go, and there is no effective way to completely solve the problem of smog pollution. Therefore, this paper proposes a BP neural network and a long and short-term memory network based on deep learning to predict haze weather conditions, and optimizes the model to make it more effective and accurate. Through the introduction of various data with time series characteristics, such as atmospheric pollutants, influencing factors detected in different spatial geography, and meteorological factors, through the operation and integration of the LSTM model, a waveform chart is formed to intuitively show the future period of time Haze weather conditions in China.
