摘要
将传统专家系统和人工神经网络的优点相结合,构建混合型专家系统应用于材料设计,弥补了神经网络和专家系统各自的缺陷,建立了材料的成分和性能之间的关系,并通过实验验证了成分对性能的影响规律,实现了知识库自动获取知识,提高了对新材料的预测能力,为加快新材料的研究和开发提供了有力的工具.此研究结果在实践中有较高的应用价值,可以大大降低试验成本,提高新材料开发速度.
A hybrid expert system is used inmaterial design, which collecting virtue of expert system andthat of artificial neural network, making up the defect of applying expert system or artificial neural network singly, founding the relation between composition and capability of material, verifying infection rules of the material composition to capability, re-alizing auto-obtaining knowledge of knowledge base, improving the forecasted ability of new material, providing a powerful tool for researching and opening up the new material. The result will have a better applied value, and reduce the test cost greatly, increase exploited speed of new material.
出处
《河北工业大学学报》
CAS
2005年第1期61-64,共4页
Journal of Hebei University of Technology
关键词
人工智能
神经网络
混合型专家系统
材料设计
材料性能
artificial intelligence
neural network
hybrid expert system
material design
material capability.