基于多目标进化算法的独立型风光互补系统的能量优化
发布时间：2018-09-03 13:50
【摘要】：可再生能源是世界各国未来经济与能源利用发展的重要方向。微电网不仅能并网运行,而且也可以作为独立电网运行,在部分山区和海岛等偏远孤立地带,微电网不仅能够解决当地供电问题,而且可以减少大量供电设施建设的成本,改善偏远地区的生产生活水平。其中风力、光伏及储能电池组成的风光互补系统是众多新能源示范项目的主要构成形式,本文以其为研究对象,主要研究工作有:(1)多目标差分进化算法的改进研究。为了避免算法陷入局部最优,增加其收敛速度,更好地模拟生物自然进化过程,在原有的基于分解的多目标进化算法(MOEA/D)的基础上,将混沌初值理论、参数自适应方法和线性加权和的方法引入该算法,通过测试函数的测试初步达到了改进的效果。(2)独立型风光互补系统中分布式电源经济性和可靠性容量配置的研究。本部分分别分析系统的光伏电源、风力发电机和蓄电池的工作特性,建立系统的经济性和可靠性目标,进行IMOEA/D的微源容量配置优化的研究,并与其他常用多目标进化算法(SPEA、NSGA-Ⅱ、MOPSO、NNIA)的优化结果进行比较,得出符合工程需要的微网系统分布式电源容量优化配置组合。(3)独立型风光互补系统分布式电源能量控制的MATLAB/Simulink仿真研究。在最大限度利用可再生能源的原则下,先后做出了光伏电源和风力发电机的变步长爬山法最大功率的跟踪(MPPT)仿真模型以及蓄电池的恒压充放电的能量控制模型,提高微网系统的可再生能源的利用率和稳定性。将本文提出的IMOEA/D应用到实际工程中,在可靠性近似相等的情况下,IMOEA/D获得经济成本均小于其它算法获得的结果;而对于本文提出的变步长爬山法跟踪结果,其跟踪时间和跟踪过程中产生的波动都比传统的定步长的要小,可以提高跟踪效率和稳定性;最后对蓄电池的控制技术可以是其输出电压稳定在理想电压范围内,并且可以实现充放电模式的转换和控制。
[Abstract]:Renewable energy is an important direction of economic and energy utilization in the world. Microgrid not only can be connected to the grid, but also can be operated as an independent grid. In some remote and isolated areas such as mountainous areas and islands, microgrid can not only solve the local power supply problem, but also reduce the construction cost of a large number of power supply facilities. Improve production and living standards in remote areas. Wind, photovoltaic and solar energy storage cells are the main components of many new energy demonstration projects. In this paper, the main research work is as follows: (1) the improvement of multi-objective differential evolution algorithm. In order to avoid the algorithm falling into local optimum, increase its convergence speed, and better simulate the natural evolution process of biology, the chaotic initial value theory is based on the original decomposition-based multiobjective evolutionary algorithm (MOEA/D). The parameter adaptive method and the linear weighted sum method are introduced, and the improved results are obtained by testing the test function. (2) the research on the economy and reliability capacity configuration of distributed power supply in the independent wind and wind complementary system. In this part, the characteristics of photovoltaic power supply, wind turbine and battery are analyzed respectively, and the economic and reliability targets of the system are established, and the optimization of IMOEA/D micro-source capacity configuration is studied. The results are compared with those of other commonly used multi-objective evolutionary algorithms (SPEA,NSGA- 鈪，
本文编号：2220152
[Abstract]:Renewable energy is an important direction of economic and energy utilization in the world. Microgrid not only can be connected to the grid, but also can be operated as an independent grid. In some remote and isolated areas such as mountainous areas and islands, microgrid can not only solve the local power supply problem, but also reduce the construction cost of a large number of power supply facilities. Improve production and living standards in remote areas. Wind, photovoltaic and solar energy storage cells are the main components of many new energy demonstration projects. In this paper, the main research work is as follows: (1) the improvement of multi-objective differential evolution algorithm. In order to avoid the algorithm falling into local optimum, increase its convergence speed, and better simulate the natural evolution process of biology, the chaotic initial value theory is based on the original decomposition-based multiobjective evolutionary algorithm (MOEA/D). The parameter adaptive method and the linear weighted sum method are introduced, and the improved results are obtained by testing the test function. (2) the research on the economy and reliability capacity configuration of distributed power supply in the independent wind and wind complementary system. In this part, the characteristics of photovoltaic power supply, wind turbine and battery are analyzed respectively, and the economic and reliability targets of the system are established, and the optimization of IMOEA/D micro-source capacity configuration is studied. The results are compared with those of other commonly used multi-objective evolutionary algorithms (SPEA,NSGA- 鈪，
本文编号：2220152
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2220152.html
