机构地区: 东北大学材料与冶金学院
出 处: 《材料科学与工艺》 2007年第3期316-318,共3页
摘 要: 为了减少实验量,降低成本,利用人工神经网络的原理,选取温度、颗粒尺寸、压坯密度为输入量,产物的孔隙度为输出量,建立了反映自蔓延高温合成反应参数与产物孔隙特性内在关系的模型.研究表明,该模型可以对选定工艺条件下产物的孔隙度进行良好的预测,预测结果在合理的误差范围内.说明建立的反应参数与孔隙特性的关系模型是可靠的,可以通过此模型优化反应参数. In order to decrease experimental work and cost, artificial neural network theory was used to build a model reflecting the relationship between process parameters of self-propagation high-temperature synthesis and porosity characteristic of products. In this model, temperature, particle size and green density were as input parameters and porosity of reacted products was as output parameter. The porosity of reacted products was well predicted at selected process parameters by this model. The predicted error lied at rational range, which indicated that the built model was reliable and could be used to optimize process parameters.
领 域: [一般工业技术]