中文摘要：

针对采用实验法测定电厂动力配煤的发热量和着火温度存在操作繁琐和信息滞后较大等不足,建立Elman神经网络预测模型。该网络模型在学习过程中确定混煤的发热量和着火温度与单煤的水分、灰分、挥发分之间的非线性映射关系。模型利用单煤的水分、灰分和挥发分含量直接预测混煤的发热量和着火温度,预测结果误差较小。利用置信区间分析法对预测模型的预测效果进行检验。研究结果表明：预测模型具有较高的可靠性和置信度。

英文摘要：

In order to overcome the disadvantages of tedious operation and serious information lag in testing calorific value and ignition temperature of blended coal in laboratory,the Elman neural network prediction model was established.In the learning process of the neural network,the nonlinear mapping relationships between calorific value and ignition temperature of blended coal and moisture content,ash,volatile matter of single coal were determined.The calorific value and ignition temperature of blended coal were predicted by the model,using the information of moisture content,ash,volatile matter of single coals.Prediction results of the predication model were verified by using confidence interval method.The results show that the model has high reliability and confidence.