Thermodynamic analysis and evolutionary algorithm based on multi-objective optimization performance of actual power generating thermal cycles
Thermodynamic analysis and evolutionary algorithm based on multi-objective optimization performance of actual power generating thermal cycles
复制标题
DOI：
10.1016/j.applthermaleng.2016.01.122
复制
发表时间：
2016-04
影响因子：
6.4
通讯作者：
M. Ahmadi;M. Ahmadi;F. Pourfayaz
中科院分区：
文献类型：
--
作者：
M. Ahmadi;M. Ahmadi;F. Pourfayaz
本研究的主要目的是找出不可逆发电热循环的最佳评估原则。这些方法，可以通过以前的工作中发现，生态性能系数，火用性能系数和最大可用功。不可逆卡诺功率循环被定义为系统。热力学分析中包括外部和内部不可逆性。在本文中，两个场景中定义的优化进展。每一种设想方案的产出都分别进行了研究。在第一种情况下，以最大化生态性能系数（ECOP）、火用性能标准（E P C）和最大可用功（M A W）为目标，采用多目标优化算法。此外，在整个第二场景中，通过多目标优化方法同时最大化包括第一定律效率（η）、有效能性能标准（E P C）和最大可用功（M A W）的三个目标函数。将多目标进化算法与非支配排序遗传算法相结合，提出了一种新的求解方法。决策是通过三个众所周知的方法，包括LINAMP和TOPSIS和FUZZY。最后，对上述系统的输出进行了误差分析。
The key objective of this research is to find out the best assessment principles for irreversible power generating thermal cycles. These approaches, which can be found through previous works, are the ecological coefficient of performance, exergetic performance coefficient and maximum available work. Irreversible Carnot power cycle is defined as system. External and internal irreversibilities are encompassed in the thermodynamic analysis. In this paper, two scenarios are defined in the optimization progression. The outputs of each scenarios are studied individually. Throughout the first scenario, with the aim of maximize the ecological coefficient of performance (ECOP), the exergetic performance criteria (E P C) and maximum available work (M A W), multi-objective optimization algorithms is engaged. Furthermore, throughout the second scenario, three objective functions comprising the first law efficiency (η), the exergetic performance criteria (E P C) and maximum available work (M A W) are maximized at the same time via multi objective optimization approaches. The multi objective evolutionary approaches (MOEAs) coupled with non-dominated sorting genetic algorithm (NSGA-II) approach is applied in the present paper. Decision making is performed via three well-known approaches comprising LINAMP and TOPSIS and FUZZY. Finally, error analysis of the outputs are accomplished for the aforementioned system.