论文部分内容阅读
提出非数值的佛山惠景城站点空气质量预报模型,将BP神经网络引入到预报模型中,以降雨量、风速、风向、温度、湿度和云量等气象参数和前1 d污染物浓度为模型输入参数,建立了结构为7-7-1的非季节预报模型和夏季预报模型。结果表明,夏季模型无论在模型检验还是在实际预报精度方面都略优于非季节模型。夏季模型的级别预报评分基本在90分以上,综合评分比非季节模型高10%。对夏季模型进行了参数敏感性分析,结果表明具有较好的稳定性。
A non-numerical model of the air quality of Foshan Hui King City site was proposed. The BP neural network was introduced into the forecasting model. The parameters of rainfall, wind speed, wind direction, temperature, humidity and cloud cover, , Established a non-seasonal forecast model with a structure of 7-7-1 and a summer forecast model. The results show that the summer model is slightly better than the non-seasonal model in terms of model test and actual forecast accuracy. The level forecast of the summer model is basically above 90 points, and the comprehensive score is 10% higher than the non-seasonal model. The parameter sensitivity analysis of the summer model shows that the model has good stability.