【摘要】 以大量煤质分析数据为基础，建立了利用煤工业分析数据（ 包括水份、灰份、挥发份及热值） 计算元素分析数据的ＢＰ神经网络预测模型，并将该模型与现有经验公式进行了比较，结果表明神经网络模型有很好的推广能力。可以满足工业应用的要求更多还原

【Abstract】 There existsintrinsicrelation between the primary elements such as carbon , hydrogen ,oxygen and nitrogen and so on ,and the proximate analysis data including mois ture content,ash content,volatile content,fixed carbon and heatvalue ofcoal- However, thisrelationis much complicated and difficultto understandfully- Based on alarge quantity of coaldata ,a BPneuralnetwork modelis putforward inthis paperto predictthe ultimate analysisofa coalfrom its proximate analysis data- Comparison with the existing empirical modelsindicatesthat neuralnetwork techniqueis quite effectivein drawingtheintrinsic and intricate relationship betweenthe ultimate analysis ofa coaland its proximate analysis data-更多还原