BP神经网络算法在空气质量预测中的应用——以江苏为例

Application of BP Neural Network in Air Quality Prediction——A Case Study of Jiangsu Province

空气污染治理已成为环保部门亟需解决的重要问题。基于此,笔者运用BP神经网络模型,从宏观层面对江苏省各监测点的空气污染数据进行预测,训练样本的平均总体相对误差为0.226,检验样本平均总体相对误差为0.193,略小于训练误差。显示预测结果具有较高可信度,预测效果整体较为理想,具有参考价值,研究进一步拓展了BP神经网络算法的应用范围。

Air pollution control has become an important issue that the environmental protection department needs to solve.In this paper,using the BP neural network prediction model,the air pollution data of the monitoring points in Jiangsu Province are predicted from the macro level.The average relative error of the training samples is 0.226,and the average relative error of the test samples is 0.193,which is slightly less than the training error.The prediction results are highly reliable,and the overall prediction effect is ideal and has reference values.Our research further expands the application of BP neural network algorithm.