换热器网络设备面积与清洗时序同步优化

Simultaneous optimization of areas and cleaning schedule for heat exchanger networks

表面结垢会严重影响换热器的传热效率,定期清洗是解决该问题的主要方式。针对以往换热器网络清洗时序优化方法中用于决策的整型变量较多而难以求解的问题,提出以换热器清洗的最大允许污垢热阻为优化变量,取代表示换热器是否清洗的二进制变量,将混合整数非线性规划问题转化成非线性规划问题,能够有效地减小问题规模,降低求解难度。优化过程中兼顾换热器网络的设计型与操作型问题,采用遗传/模拟退火算法同步优化换热器的面积与清洗时序。将该方法用于一个实例,所得年度总费用与文献基本一致,验证了该方法的有效性。

Fouling over the heat transfer surface of equipments always causes reduction to the overall heat transfer coefficient. Regular cleanings become an important means of saving energy. In the existing researches about the optimal cleaning schedule in heat exchanger networks, too many binary variables were involved, which raised the difficulties to solve the problem. This paper took the maximum allowable fouling resistance of the heat unit as the optimization variable, instead of the binary variables to represent requirement of cleaning the heat exchangers. The proposed method converted the mixed integer nonlinear programming （MINLP） problem into a nonlinear programming （NLP） problem, effectively reducing the size of the problem and the difficulty to solve the problem. With consideration of both design and operation problems of the heat exchanger network, genetic/simulated annealing algorithm （GA/SA） was adopted to optimize areas and cleaning schedule of the heat exchangers simultaneously. An example was studied to illustrate the effectiveness of the method.