Multi-objective Optimization Research And Application On Advanced Energy Utilization System

There are many multi-objective optimization problems in science and engineering areas, and the multi-objective optimization is becoming a research focus as well as difficult problem in the area of engineering. For the multi-objective optimization problem, it is difficult to find a satisfactory solution because that the objects mutually restricted and influenced with each other, and optimizing one of them must at the expense of the sacrifices of others. The multi-objective optimization technology with important scientific significance and application value can balanced both objects and find the tradeoff solution.Coupling with the growth of energy consumption, pollution problems have become more and more serious. Then the conventional energy utilization system has been unable to meet the current requirements of the advanced energy system, which is highly efficient, nopollution, economic and safety, and has become an important factor of the constraint on the sustainable development of human society. For this reason, governments around the world are looking for sustainable and advanced energy utilization systems to meet the growing need of energy and reduce the environment pollution. The performance evaluation system of advanced energy utilization systems included technical object, economic object, environmental object and social object, etc. And the importance of them were different from each other. Therefore, the optimization of this advanced energy utilization systems was a typical multi-objective optimization problem.In this thesis, a mathematical model of multi-objective optimization for the hybrid solid oxide fuell cell and gas turbine power generation system was established on the basis of studying the method for solving the multi-objective optimization problems. In this thesis, the multi-objective optimization for the hybrid solid oxide fuell cell and gas turbine power generation system were performed, using the traditional multi-objective optimization method and genetic algorithms respectively, in order to find the best work point of the power generation system.The main contents are as follows:①Studying the traditional method for solving multi-objective optimization problems and analysing the shortcomings of them.②On account of the shortcomings and limitations of the traditional strategies for solving multi-objective optimization problems, this thesis studies a multi-objective optimization evolutionary algorithms -- genetic algorithm, and its improved model -- NSGA-Ⅱ. And borrowed the ideas from the concept of dynamic penalty function, this thesis building a penalty function which suitable for solving constrained multi-objective optimization problems. Finally, an example strongly illustrated the effectiveness of the NSGA-Ⅱ.③In this thesis, the basic concept and theories of solid oxide fuel cells are reviewed and analyzed. On this basis, the mathematical model of multi-objective optimization for the solid oxide fuell cell power generation system was established. And then, the multi-objective optimization problem was solved. By optimizing, the SOFC power generation system can achieve an actual potential of 0.737(V), a system electrical efficiency of 53.65% and a net output work of 1 608.1×103(W). The optimization results shown that the best work point of this power generation system can be obtained through the multi-objective optimization system for SOFC.④According to the characteristics that the temperature of exhaust gas of SOFC was very higher, this paper also presented an improved program, the mathematical model of multi-objective optimization for the hybrid solid oxide fuell cell and gas turbine power generation system, and analyzed the effects of the main parameters such as the fuel utilization rate, the cell temperature and the fuel flow rate on the objective functions. In this thesis, the multi-objective optimization for the hybrid solid oxide fuell cell and gas turbine power generation system were performed, using the traditional multi-objective optimization method and genetic algorithms respectively, and the results of both methods were compared.The optimization results shown that the integration into a more complex GT cycle of a pressurised SOFC can result in a further improvement of the system electrical efficiency firstly. And it also shown that comparing with the traditional method, NSGA-Ⅱalgorithm was more suitable for multi-objective optimization problem, and can efficiently obtain the uniformly distributed pareto set.