提出了一种应用人工神经网络进行大气质量预测的方法,即采用多层前向网络的覆盖算法和时间序列进行短期的空气质量预测。针对空气污染的特点,选取了对空气质量有重要影响的气象因子的相关数据,并对其进行测试。实验结果表明:将该方法应用于空气质量预测,效果良好,学习速度快,识别率高,具有较强的实用价值,为实现大气质量预测提供了一种准确高效的方法。
A new method of prediction of air quality is proposed based on neural networks, wlfich is the covering algorithm of muki - layer neural networks and time series. According to the characteristics of the prediction of air quality,selects the data that greatly influence the air quality. The experimental result shows that the performance of prediction of air quality is favorable, the learning speed is fast and the rate of accurate is high,so it has a practical value . It provides a precise and efficient way for the prediction of air quality.