论文部分内容阅读
建立了有分流情况下换热器网络综合的数学模型,该模型改进了文献中等温混合的不合理假设.有分流情况下的换热器网络综合问题本质上是一个混合整数非线性规划(MINLP)问题.这类问题的非凸非线性特性使得目标函数存在多个局部最优解.传统的基于梯度的搜索方法在处理这类问题时由于计算规模庞大且极易陷于局部最优解而不再适用.因此对遗传算法进行多流股换热器网络综合的方法进行了研究,最后通过例题证明了方法的有效性.
A mathematic model of the heat exchanger network with diversion is established, which improves the irrational assumption of isothermal mixing in the literature.The heat exchanger network synthesis with diversion is essentially a mixed integer non-linear programming (MINLP ). The non-convex nonlinearity of these problems leads to the existence of multiple local optimal solutions to the objective function.Traditional gradient-based search methods in dealing with such problems due to the large-scale calculation and easily trapped in the local optimal solution So it is applicable to the genetic algorithm for multi-stream heat exchanger network synthesis method was studied, and finally proved the effectiveness of the method through examples.