基于PSOGSA-LSTM模型的长江经济带空气质量指数预测研究

A Study on Forecast of Air Quality Index in the Yangtze River Economic Zone Based on PSOGSA-LSTM Model

本文以长江经济带100个地级及以上城市的空气质量指数(AQI)为研究对象,选取了六大污染物和11个影响空气质量指数的气象因子作为影响因素。针对空气质量相关数据的特性,将粒子群算法和万有引力算法结合的混合算法(PSOGSA)与长短期记忆(LSTM)神经网络进行组合,建立PSOGSA-LSTM组合预测模型,对模型的预测精度进行了三个方面的检验,并与传统的LSTM模型的预测结果进行比较,最后将其应用于长江经济带100个城市未来7天的空气质量指数预测。研究结果表明,PSOGSA-LSTM模型相比传统的LSTM模型具有更高的预测精度和较强的稳定性。

In this paper,the air quality index(AQI)of 100 cities at prefecture level and above in Yangtze River Economic Belt is taken as the research object.Six major pollutants and 11 meteorological factors affecting air quality Index are selected as the influencing factors.According to the characteristics of air quality data,the hybrid algorithm that combining of particle swarm algorithm(PSO)with gravitational search algorithm(PSOGSA)is combined with long short-term memory(LSTM)to established PSOGSALSTM model.The prediction accuracy is tested in three aspects and compared with the prediction results of the traditional LSTM model.Finally,it is applied to the prediction of the air quality index of 100 cities in the Yangtze River Economic Belt in the next 7 days.The results show that the PSOGSA-LSTM model has higher prediction accuracy and stronger stability than the traditional LSTM model.