多流股换热器设计及网络优化同步综合

发布时间：2018-11-10 06:43

【摘要】：近年来,多流股换热器以其结构紧凑、高效低耗等特点,成为过程强化研究的热门课题。然而,尽管它早已广泛应用于低温工业领域,但有关其设备及网络的研究仍然较少,且现有研究工作也并未完善。基于该现状,以多流股换热器为对象,展开设备设计与网络优化同步综合工作。首先,以其独特换热形式为着手点,建立多流股同步换热热耦合模型,并通过流股能量匹配,揭示其换热原理,以及它与常规双流股换热的拓扑关系。同时,引入状态空间(State Space, SS)超级结构,并将多流股换热器热耦合模型作为过程操作算子,纳入到过程操作(Process Operator, PO)网络中,从而建立多流股换热器网络综合方法。其次,考虑到多流股换热器工业应用背景,将环境因素与保温材料项纳入到多流股换热器设备设计中。通过系统预分析,提出平板式模型假设,构建多流股换热器集总模型；并结合SS超级网络,构造出得以量化环境影响与保温材料封包效果的系统预分析框架。接着,从流股层面对多流股换热器模型进行改进,详细量化环境因素对过程流股的影响,及保温材料封包效应,摆脱集总模型的笼统计算。提出多流股换热器流道分布规则,从而得到更为详细合理的设备设计策略,并将其作为新的过程操作算子。同时,改进SS超级结构,对过程单元有序编号,去除循环回路冗余连接,将此无序网络改造成更适用于换热问题的有序超级结构,进而建立起环境作用下的多流股换热器设备设计与网络优化同步综合框架。最后,对等温相变过程进行分析,以相变率及绝对焓替代温度,提出识别过程流股状态的新判据。通过适当数学约束,建立起对等温相变处理的新方法,并结合改进的有序超级结构,提出包含等温相变过程的多流股换热器设计与网络综合策略。此外,在模型求解策略上,以连续变量替代二元变量,通过互补性约束等处理手段,将混合整数非线性规划转化为单纯的非线性问题,避免混合整数与强非线性的复杂耦合,有效降低模型复杂度以及求解难度。同时,使用目前最为完善的确定性算法CONOPT,作为非线性问题的模型求解器,结合随机扰动赋初值和多次迭代反复求解的策略,经局部最优解筛选,捕捉全局最优网络。最终,通过实际算例对所提设计策略与综合方法可行性与优越性进行验证。
[Abstract]:In recent years, multi-flow heat exchanger with its compact structure, high efficiency and low consumption has become a hot topic in process strengthening research. However, although it has been widely used in the field of cryogenic industry for a long time, the research on its equipment and network is still few, and the existing research work is not perfect. Based on this situation, the synthesis of equipment design and network optimization is carried out with multi-stream heat exchanger as the object. Firstly, taking its unique heat transfer form as the starting point, the coupling model of simultaneous heat transfer of multi-flow strands is established, and the principle of heat transfer and the topological relationship between it and the conventional two-stream heat transfer are revealed through the energy matching of the flow streams. At the same time, the state space (State Space, SS) superstructure is introduced, and the heat coupling model of the multi-flow heat exchanger is taken as the process operator, which is incorporated into the process operation (Process Operator, PO) network, and the synthesis method of the multi-stream heat exchanger network is established. Secondly, considering the industrial application background of multi-flow heat exchanger, the environmental factors and heat preservation materials are brought into the design of multi-flow heat exchanger equipment. Through system pre-analysis, a flat plate model hypothesis is put forward, and a lumped model of multi-flow heat exchanger is constructed, and a system pre-analysis framework for quantifying environmental impact and heat preservation material sealing effect is constructed by combining with SS supernetwork. Then, the multi-flow heat exchanger model is improved from the flow level, and the influence of environmental factors on the process flow stock is quantified in detail, and the heat preservation material envelope effect is also analyzed to get rid of the general calculation of the lumped model. The flow channel distribution rules of multi-flow heat exchangers are proposed, and a more detailed and reasonable design strategy is obtained, which is regarded as a new process operator. At the same time, the superstructure of SS is improved, the process unit is ordered numbered, the redundant connection of the loop is removed, and the disorder network is transformed into an ordered superstructure which is more suitable for heat transfer problems. Furthermore, an integrated framework for the design of multi-stream heat exchanger equipment and network optimization is established. Finally, the isothermal phase transition process is analyzed, and a new criterion for identifying the flow state of the process is proposed by replacing the temperature with the transformation rate and absolute enthalpy. A new method for the treatment of isothermal phase transition is established by proper mathematical constraints. Combined with an improved ordered superstructure, a comprehensive strategy for the design and network of multi-flow heat exchangers with isothermal phase transition processes is proposed. In addition, in the strategy of solving the model, the mixed integer nonlinear programming is transformed into a simple nonlinear problem by using continuous variables instead of binary variables, and by means of complementary constraints, so as to avoid the complex coupling between mixed integers and strong nonlinearity. The complexity of the model and the difficulty of solving the model are reduced effectively. At the same time, the most perfect deterministic algorithm, CONOPT, is used as the model solver for nonlinear problems. Combined with the strategy of random perturbation with initial value and repeated iterative solution, the global optimal network is captured by the selection of local optimal solution. Finally, the feasibility and superiority of the proposed design strategy and synthesis method are verified by practical examples.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ051.5

【相似文献】