精馏分离序列综合邻域结构的研究

Neighborhood Structure of Distillation Separation Sequence Synthesis

为了有效解决精馏分离序列优化综合问题,研究邻域(超级)结构是成功实现寻优算法的前提。由于精馏分离序列与二叉树之间具有同构性,在数据结构上精馏分离序列可以抽象为二叉树,进而采用图论方法对其进行研究。本文运用组合数学理论深入研究了精馏分离序列综合问题,简明分析了有序剖分问题的计算复杂性;通过二叉树相邻切分点变换机制,实现了对精馏分离序列的随机搜索;提炼出后序遍历相邻变换的等价规则,从而构造出高效演化邻域结构。

In order to solve the distillation separation sequence synthesis problem effectively, super or neighborhood structure must be researched, and it is the foundation of realizing optimization algorithm. Because there is analogous structure between the distillation separation sequence and the binary tree, distillation separation sequence can be described as binary tree data structure, and then adopt the method of graph theory to study it. Distillation separation sequence synthesis problem has been further researched with combinatorial mathematics theory, and the computing complexity of sequence decomposition problem has been concisely analyzed; random search for distillation separation sequence has been implemented; what's more, through the change mechanism of neighboring split position in binary tree, the equivalent rules of neighboring change based on post order traversal has been abstracted, consequently, a kind of high effective evolutional neighborhood (super) structure has been constructed.