基于专家知识的字典序优化决策方法及工业应用

为有效实现有色冶金配料过程的实时优化，首先根据氧化铝配料过程特点，建立一种字典序区间目标规划模型；然后通过整理长期积累的专家经验知识，构建了具有优先级的分类知识库；接着基于专家知识库的分类结构，提出一种字典序专家推理策略，以实现区间约束条件下的多质量指标的优化控制．工业应用结果表明，所提出的方法能够很好地实现生料浆质量的优化控制，为其他有色冶金工业配料过程的控制提供了一种优化模式．

Expert-knowledge-based lexicographic order optimal decision method and its industrial application

In order to effectively solve the real-time optimization problem in the raw material blending process of nonferrous metal production, an interval lexicographic goal programming mode is proposed based on the characteristics of alumina blending process. Then, a kind of classificatory knowledge base with precedence is constructed by coordinating the expert knowledge accumulated in long-term production. Based on the knowledge base, a kind of lexicographic reasoning strategy is proposed to realize the optimal control of multiple quality indexes with interval constraints. The results of industrial application show that, the proposed method can realize the optimal control of slurry quality indexes, which provides a good optimization mode for other blending processes of nonferrous metal production.