基于带权有向图的非清晰复杂精馏序列的合成方法

在精馏系统中非清晰分离能够克服清晰分离所固有的返混效应,从而提高系统能效。基于数据结构理论,提出了非清晰精馏序列的合成与优化方法。对于N组元混合物的分离,合成的精馏序列由N-1个精馏塔构成,且其中的非清晰分离任务含有所允许的任意多个中间分配组分。为了合成上述分离序列,利用带权有向图建立了一种新的精馏分离序列模型,并基于"vector"动态数组,提出了一个同时拥有数组和链表优点的带权有向图的储存结构。序列合成过程定义了一系列的相关操作,且选择广度优先的策略以提高合成效率。算例证明了该方法在精馏序列合成和优化的高效性。算例结果表明该算例的最优分离序列可以有效减少设备投资和能耗。     

考虑中间换热器的能量集成精馏序列合成

针对考虑中间换热器（IHE）的精馏序列合成问题,提出基于随机优化策略的能量集成非清晰精馏序列（IHE-HIDSs）合成方法。通过对精馏序列分离任务合并处引入二元0/1变量表示是否存在IHE,以精馏序列的年总成本（TAC）为优化目标,建立了该合成问题的隐式混合整数非线性规划模型（MINLP）,通过模拟退火和粒子群优化（SA-PSO）混合随机优化算法进行求解。为验证在精馏序列合成中同时考虑IHE的必要性以及所提出合成方法的有效性,对五组分醇类混合物和五组分烷烃类混合物两个算例的精馏序列合成问题进行了研究。结果表明,相比同时考虑热耦合和能量集成的精馏序列,IHE-HIDS具有更低的TAC。此外,所提出的方法可以在合理的计算时间内以高概率获得多个分离序列方案。     

采用分块物料分配图研究多原料非清晰精馏序列合成

提出1种分块物料分配图以表达多原料非清晰精馏过程,在此基础上提出1种多原料非清晰精馏序列的合成方法.合成过程中,用经验法决定分离顺序,各物流量及相关操作以分块物料分配图表示.通过对只带旁路和清晰分离塔的多原料非清晰精馏序列合成示例的处理,说明了本文方法的具体应用.     

改进的物料分配图用于非清晰精馏过程合成

本文提出一种改进的物料分配图(material allocation diagfam)表达分离过程。并在此基础上提出一种非清晰精馏过程合成方法。示例表明:本方法可很好地处理单一进料、只具有旁路的非清晰精馏过程合成问题,有时甚至可直接得最优解。     

应用遗传算法实现精馏分离序列优化综合

阐述了采用遗传算法求解具有组合爆炸特征的精馏分离序列综合问题。首先在数据结构上将精馏分离序列抽象为二叉树结构编码,然后应用图论方法建立起有效的交叉和变异算子,最后对10组分精馏分离序列综合问题实例进行求解。结果表明,遗传算法对最优解命中率达80％,平均空间搜索率仅为0．0318。因此遗传算法能够成功解算大规模精馏分离序列优化综合问题。     

设计演化算法实现精馏分离序列优化综合

采用演化算法求解具有组合爆炸特征的精馏分离序列优化综合问题。由于精馏分离序列与二叉树之问具有同构性．在数据结构上精馏分离序列可以抽象为二又树,直接采用二叉树结构编码方案可以同时表达个体的基因型和表现形。借鉴生物界父本可以经过有性繁殖或无性繁殖得到子代的现象。对遗传算子进行设计。应冒图论方法建立有效的二叉树结构演化重组机制,从而形成基于双亲或单亲父本的交叉算子。实例表明：双亲遗传操作其空间搜索率较低,而单亲遗传操作其最优解命中率较高。因此。演化算法能够成功解算大规模精馏分离序列优化综合问题。     

基于模糊逻辑推理的精馏分离序列优化综合

模糊逻辑推理能够反映自然语言要素信息传递特性. 通过抽象有效重组参数提炼if-then模糊逻辑推理规则、根据精馏分离序列综合问题数字特征创建Sugeno模糊逻辑推理模型,系统化地提出了基于模糊逻辑推理的精馏分离序列优化综合方法.     

基于神经网络混合整数线性规划的精馏分离序列优化综合

针对精馏分离序列综合问题,运用图论方法研究网状超级结构。结合相对费用函数建立了混合整数线性规划模型。开发了面向大规模精馏分离序列综合问题的自动求解程序,实现了10组分精馏分离序列综合问题的成功解算,探索了神经网络混合整数线性规划的思想框架。     

非清晰精馏序列的两步合成法

１引言精馏序列合成是过程合成的一个重要分支。多组分精馏可分为两大类，第一类：将单一原料分离成若干纯组分；第二类：将一个或几个多组分原料分离为若干个多组分产物。２０多年来，第一类精馏序列合成问题受到了极大重视，并提出了不少研究方法［１］。对第二类问题的...     

基于自适应并行禁忌搜索的精馏分离序列优化综合

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.采用图论方法建立了二叉树相邻切分点变换机制,进而形成有效的邻域结构.将自适应机制和并行技术引入禁忌搜索算法中,根据记忆频率信息自适应调整禁忌长度与候选集规模,通过搜索任务分配安排形成多任务并行处理.数值实验表明自适应并行禁忌搜索算法能够成功解算大规模精馏分离序列优化综合问题.     

A systematic procedure for synthesis of intensified nonsharp distillation systems with fewer columns

Nonsharp separation sequences can intrinsically improve thermodynamic efficiency than that of sharp sequences for a multicomponent distillation. There are much more nonsharp sequences than that of sharp sequences. In this paper, the synthesis of new intensified distillation systems from nonsharp separation sequences is presented. Such intensified distillation systems have fewer columns and have the potential to reduce both energy and capital costs than conventional configurations. First, the representation of any nonsharp sequence with the simple column configuration is found to be the necessary starting point. Then, five strategies are introduced which present different mechanisms to change the structures of the distillation configurations. Third, a systematic procedure is formulated which can achieve the intensified distillation systems with fewer columns systematically. Finally, the systematic method has been applied to different examples of nonsharp separation sequences. It is demonstrated that the method can be used to generate the new intensified distillation systems from any nonsharp separation sequences. The method is easy to use and can explicitly derive all of the intensified distillation systems for the nonsharp separation sequences for a multicomponent distillation.     

The synthesis of distillation-based separation systems

The last dozen years have produced almost 50 English language articles on the synthesis of separation systems [1]. Most work has considered separating a single relatively ideal mixture into sharply split, usually pure-component products using systems of single-feed, two-product distillation columns. Of late, the work has considered heat-integrating these columns to reduce energy consumption. Present research will very likely produce valuable results for nonsharp separations.Earlier approaches developed both algorithmic- and heuristic-based tree search algorithms for discovering the better systems of non-heat-integrated columns. New results permit the design of heat-integrated systems.The paper highlights the above developments.     

Synthesis of nonsharp distillation sequences via genetic programming

This paper addresses the application of Genetic Programming (GP) to the synthesis of multicomponent product nonsharp distillation sequences. Combined with the domain knowledge of chemical engineering, some evolutionary factors are improved, and a set of special encoding method and solving strategy is proposed to deal with this kind of problem. The system structural variable is optimized by GP and the continuous variable is optimized by the simulated annealing algorithm simultaneously. Because GP has an automatic searching function, the optimal solution can be found including distillation, splitting, blending and bypassing operations automatically without any superstructures of nonsharp distillation sequences. Three illustrative examples are presented to demonstrate the effective computational strategies.     

Stochastic GP synthesis of heat integrated nonsharp distillation sequences

Genetic programming (GP) is used to solve the synthesis problem of heat integration nonsharp distillation sequences and the optimization objective of it is to seek the optimal heat integration nonsharp flow based on minimizing the annually total cost. Meanwhile, the major technological parameters for these important equipments are given. Combining with the domain knowledge of chemical engineering, some evolutionary factors are improved, and a set of special encoding method and heat integration strategy is proposed to deal with this kind of problem. The system structural variable is optimized by GP and the continuous variable is optimized by the complex algorithm simultaneously. Because GP has the automatically searching function, the optimal heat integration solution can be found automatically without any superstructures of nonsharp distillation sequences. Three illustrating examples are presented to demonstrate the effective computational strategies.     

A novel stochastic optimization method to efficiently synthesize large-scale nonsharp distillation systems

This study presents a novel stochastic optimization method for the efficient synthesis of large-scale nonsharp distillation systems, where heat integration and thermal coupling can be involved simultaneously. A new binary tree encoding method was developed to represent distillation sequences with no limits on the number of middle components in nonsharp splits to ensure a complete solution space. Thermally coupled structures were defined by 0–1 binary variables. Evolutionary rules were developed to generate neighboring distillation configurations randomly. Finally, an optimization framework was proposed, where simulated annealing (SA) algorithm was used to optimize distillation configurations; for a certain distillation configuration randomly generated by SA, its continuous variables were optimized using particle swarm optimization algorithm. Four cases—including the synthesis of six- and seven-component nonsharp heat integrated and thermally coupled distillation sequences—were studied to demonstrate that the proposed method was efficient and could obtain optimal and valuable suboptimal solutions with high probabilities.     

TCS-S热偶精馏过程的优化设计

建立了TCS－S热偶精馏过程优化设计的数学模型，给出非精晰分割精馏塔最小回流比的求解方法，并给出简捷计算和严格计算的求解过程，编制了计算机程序。用苯－甲苯－乙苯物系为例进行考核计算，与普通塔的两个分离序列进行比较，节能效果十分显著。