期刊导航

论文摘要

三角旋回算法及其在水电厂日前市场优化调度中的应用

The Triangle Gyration Algorithm and Its Application in Short term Scheduling of Hydro Plant

作者:左幸(四川大学 能源发展研究中心,四川 成都 610065);马光文(四川大学 能源发展研究中心,四川 成都 610065);涂扬举(四川大学 能源发展研究中心,四川 成都 610065)

Author:(Center of Energy Development Research,Sichuan Univ., Chengdu 610065,China);(Center of Energy Development Research,Sichuan Univ., Chengdu 610065,China);(Center of Energy Development Research,Sichuan Univ., Chengdu 610065,China)

收稿日期:2006-01-06          年卷(期)页码:2007,39(3):62-66

期刊名称:工程科学与技术

Journal Name:Advanced Engineering Sciences

关键字:水电系统;三角旋回算法;优化调度

Key words:hydro power system; triangle gyration algorithm; TGA; optimization scheduling

基金项目:国家自然科学基金资助项目(50539140)

中文摘要

为了解决水电复杂巨系统的优化问题,借鉴遗传算法和量子算法,提出一种新的全局优化方法——三角旋回算法(Triangle Gyration Algorithm,TGA),其具有结构简单、鲁棒性强和快速收敛的特点。算法的寻优过程采用历史最优目标函数值进行指导,利用三角变换进行迭代使其能够快速收敛到全局最优。用一个典型的算例对三角旋回算法进行了性能分析,并且将该算法应用在水电站日前现货市场优化调度中,通过与其他几种常见算法结果进行比较,该算法的优化结果日收益比动态规划增加8.58%,同时通过计算过程可以看出,该算法具有灾变机制,能够防止其过早陷入局部最优;算法结构简单,克服了随机搜索的盲目性,算法的迭代机理和选值与传统优化算法具有本质的差别;目前该算法研究处于初级阶段,具有巨大的可塑性。

英文摘要

In order to solve the optimal problems of complex and gigantic hydro power system, a new global optimization method, named Triangle Gyration Algorithm (TGA) was proposed. It embodies the characteristics of simple structure, good robust and fast convergence. The individual of algorithm in every generation is guided by the best objective function value and transformed by the trigonometric function, which ensure the whole progress converge quickly at the global optimal points. Compared to the usual optimization methods with a numerical example and short term scheduling of a hydro plant, TGA manifests the better performance in running time and optimal results and it provides a profit 8.58% more than DP does. The algorithm has the catastrophe mechanism capable of preventing algorithm being trapped in local optimization, simple structure, and overcomes the blindness of random searching. Iteration mechanism and value selection of TGA manifests essential difference from other traditional optimal methods.

关闭

Copyright © 2020四川大学期刊社 版权所有.

地址:成都市一环路南一段24号

邮编:610065