论文部分内容阅读
把可拓学中定量化描述解决矛盾问题过程的工具——关联函数——引入到电弧炉可拓专家系统中,用以解决电炉出钢前钢水成分不合格的矛盾问题.电弧炉终点目标神经网络预报模型作为可拓专家系统评价机构的基础.终点预报模型用混合遗传算法训练,达到了较高的预报精度.仿真结果表明电炉可拓专家系统可以指导技术人员进行更合理的炼钢操作.
The quantitative analysis of extenics in the process of solving the problem of contradictory tools - correlation function - introduced to the EAF extension expert system to solve the problem of unqualified molten steel before tapping the furnace. The network forecasting model serves as the basis for extenics expert system evaluation.The end-point forecasting model is trained by hybrid genetic algorithm, which achieves a high forecasting accuracy.The simulation results show that electric stove extender expert system can guide technicians to conduct more reasonable steelmaking operations.