铝合金电阻点焊工艺设计智能化

采用集成基于事例推理，模糊推理，模糊神经网络(FNN)等多种人工智能推理技术，建立了铝合金电阻点焊工艺参数设计系统。选取导电率、屈服强度等材质的物理参数为电阻点焊工艺参数求解模块入口参数，更加符合铝合金电阻点焊过程的物理本质；引入规范强度参数，使铝合金电阻点焊焊接参数的选择更加灵活；在此基础上建立FNN铝合金电阻点焊工艺焊接求解模型，提高系统求解的智能性及其学习能力。在铝合金电阻点焊过程数值模拟的基础上，以铝合金电阻点焊焊接参数数值模拟作为已有的铝合金电阻点焊工艺参数库的补充，丰富了神经网络的训练样本，增强了系统的泛化能力。系统应用实例表明可以满足铝合金电阻点焊工艺参数设计要求。

Intelligent process design of resistance spot welding of aluminum alloys

A process parameters design system of resistance spot welding of aluminum alloys integrating case based reasoning, fuzzy inference and fuzzy neutral network (FNN) was developed. Material physical parameters such as electrical conductivity, yield strength etc. were used as inputs of the system. This more suitable to the physical mechanism of resistance spot welding process. It showed more flexibility to design process by using process intensity as one of the parameters. A FNN was built to improve the intelligence of solution to process design and its learning ability. The results of numerical simulation of resistance spot welding aluminum alloys were included in database as a complementary for the FNN training. It enriched the training samples of FNN and intensified generalization ability of the system. The applications approved that the system was able to meet the practical demand of the process design.