【摘要】 用人工神经网络方法对高合金高强高韧钢的性能优化问题进行了研究．并结合模拟退火算法对ＢＰ算法进行了改进，使得网络结构的选取、动量项系数α及学习率η的确定更加合理，提高了网络的学习效率，改善了预测精度．研究结果可用于高强高韧钢的成分设计及热处理工艺的优化．更多还原

【Abstract】 The optimization of properties of high strength and high fracture toughness steels was investigated by artificial neural networks. The method of B P algorithm was combined with the simulating annealing algorithm,which makes the selection of networks structure and the determination of momentum parameter α and learning rate η more reasonable. The learning efficiency and the calculating precision was improved. The result proved availble for finding up an optimum heat treatment regime and for component part design for high strength high fracture toughness steel.更多还原