In this paper,we put forward a method,which produce samples effectively,and reduce workload greatly.BP neural network that used to assess air environmental quality was set up,batch processing train type,order processing train type and early stop train type were applied to train the established BP network,the properties of three train types were compared,theoretic analyse was been done as well,based on all of which,we conclude that early stop train type is the high-speedest and accuratest train method in air environmental quality assessment.%文中提出的样本生成方式幅度地减少了样本输入的工作量。建立了用于空气环境质量评价的BP神经网络模型，分别采用批处理训练方式、顺序处理训练方式和提前终止法训练方式对网络进行训练，并对三种训练方式的性能进行了对比和理论分析，从而得出提前终止法在空气环境质量评价中的快速性和准确性。
展开▼
