机构地区: 西北工业大学
出 处: 《西安交通大学学报》 2003年第3期249-251,330,共4页
摘 要: 将人工神经网络的典型模型———误差后传播 (BP)算法用于改性的碳 /碳复合材料氧化烧蚀率的研究 ,建立了碳 /碳复合材料改性添加剂组成 氧化烧蚀率BP网络模型 .研究结果表明 ,所建模型较好地反映了添加剂含量与试样氧化烧蚀率间的内在规律 ,预测的氧化烧蚀率与实验值间的误差小于 0 3 2 % .将模型筛选出的最优添加剂配方用于基体改性 ,试样的氧化烧蚀率下降 49 3 % ,说明将人工神经网络用于基体改性是可行和有效的 . An artificial neural network (ANN) model for research of the burning rate of modified carbon/carbon composites (C/C composites) is developed by using back-propagation (BP) algorithm. The relationship between the modified additives and burning rate is analyzed on the basis of the model. The results show that the relative error between the expected value and the outputs of the network is lower than 0.3%. By the aid of the ANN model, an optimized combination of these additives is obtained. The burning rate of this kind of samples decreases by 49.3%, which proves the method effective and feasible. The model could reveal the inner regularity between the additive contents and burning rate of C/C composites.