摘要
利用人工神经网络的自学习以及非线性逼近能力对材料元素与硬度的相关性进行拟合和预测；并用遗传算法的强寻优能力对喷涂材料成分进行优化．优化结果与非学习Ｎｉ基喷涂材料配方相比较表明，神经网络能对Ｎｉ基喷涂材料的性能进行较好地拟合和预测，而遗传算法则能在不同的样本区间对材料进行优化，二者的有机结合可进一步提高材料优化与设计的有效性．
Based on available samples of nickel based spraying and coating materials, self learning algorithm and non linear approaching feature of neural network are used to fit and predict the correlation between the elements and hardness of the materials and, meanwhile, powerful genetic algorithm is also applied to optimize the composition of this kind of material. The optimization results indicate that this optimization method is superior to the non learning algorithm because the performance of Ni based spraying and coating material can better fitted and predicted by using neural network, moreover, the materials in different samples can well be optimized with genetic algorithm. Thus, the optimization of materials and the effectiveness of the design can be greatly improved by organic combination of these two methods.
出处
《甘肃工业大学学报》
1998年第1期11-15,共5页
Journal of Gansu University of Technology
关键词
喷涂材料
优化
神经网络
遗传算法
镍基合金
spraying and coating material, optimization, neural network, genetic algorithm, composition
