耦合有机朗肯循环的换热网络全局优化

Global Optimization of Heat Exchanger Networks Integrated with Organic Rankine Cycle

针对换热网络与有机朗肯循环耦合集成优化问题,建立了严谨的混合整数非线性规划(MINLP)模型。同时,通过添加与物流匹配的关键约束,剔除掉重复性的网络结构,开发了一种新型的枚举算法,将复杂的MINLP模型分解为混合整数线性规划(MILP)和非线性规划(NLP)两个子模型。迭代运行MILP模型,枚举出所有可行的网络结构;再利用全局求解器BARON优化每一个网络结构所对应的NLP模型,求得固定结构的年度总费用;最后对比所有网络结构的年度总费用,筛选出全局最优的设计方案。案例分析结果表明,该算法仅需16 s就能收敛至全局最优解,与文献相比,年度总费用降低33.4%,且所提出的网络结构约束能使重复性的网络结构数量减少81.25%,从而提高算法的优化求解效率。

A rigorous mixed integer nonlinear programming(MINLP)model is developed to solve the integrated optimization problem of a heat exchange network integrated with the organic Rankine cycle.At the same time,a new enumeration algorithm is developed to decompose the complex MINLP model into mixed integer linear programming(MILP)and nonlinear programming(NLP)sub-models by adding the key constraints of stream matching and eliminating the repetitive network structure.Within this algorithmic framework,the first step involves iteratively solving the MILP model to enumerate all feasible network structures.Subsequently,for each network structure,a global solver(BARON)is employed to optimize the NLP model and determine the total annual cost(TAC)of that specific structure.Finally,through a comparison of the TACs of all network structures,the globally optimal design solution is determined.The case study demonstrates that,the proposed algorithm can converge to the global optimal solution in only 16 s,and the proposed network structure constraints can reduce the number of repetitive network structures by 81.25%,thus improving the optimization efficiency of the algorithm.