基于改进GP-GWO算法的电力负荷自动建模  

Automatic Power Load Modeling Based on Improved GP-GWO Algorithm

作　　者：杨晓萍[1] 吕图园 胡杨 高明昊 YANG Xiaoping;Lü Tuyuan;HU Yang;GAO Minghao(Xi’an University of Technology,Xi’an 710048,Shaanxi,China;State Grid Xianyang Power Supply Company,Xianyang 712000,Shaanxi,China)

机构地区：[1]西安理工大学,陕西西安710048 [2]国网咸阳供电公司,陕西咸阳712000

出　　处：《电网与清洁能源》2020年第3期60-65,共6页Power System and Clean Energy

基　　金：国家自然科学基金项目(91325201)。

摘　　要：负荷模型的建立是电力系统稳定性分析的重要环节,要求负荷模型既能准确反映负荷特征又要具有相对简单的模型结构。而目前的负荷模型往往是通过人工的方法来确定,具有很强的主观性。针对现有状况做出改进,提出将遗传算法和灰狼算法相结合的混合算法GP-GWO,实现电力负荷的自动建模。在使用遗传程序设计框架的基础上,首先用分层树结构表示复杂负荷函数模型。其次,由于灰狼算法具有较强的全局搜索能力和快速收敛能力,将改进GP算法与灰狼算法相结合,结合后的算法GP-GWO在优化模型参数、提高模型精度等性能方面有所提升。最后,从静态负荷模型和动态负荷模型中各选取一个算例,分别对其有功、无功进行建模、测试和评估,将GP-GWO算法结果数据与实验数据进行拟合对比,对比结果验证了该算法的可行性。Load modeling is an important part of power system stability analysis and it requires that the load model accurately reflect the load characteristics and have a relatively simple structure.The current load model is often determined by manual methods,which is highly subjective.To improve the existing conditions,this proposes a hybrid algorithm GP-GWO combining genetic algorithm and gray wolf algorithm to realize automatic modeling of power load.Based on the use of genetic programming framework,this paper,firstly,uses a hierarchical tree structure to represent a complex load function model.Secondly,due to the strong global search ability and fast convergence ability of the gray wolf algorithm,the improved GP algorithm is combined with the gray wolf algorithm.The combined algorithm GP-GWO has improved performance in optimizing model parameters and improving model accuracy.Finally,this paper selects one calculation example from the static load model and the dynamic load model respectively,and the active and reactive power of the example is tested and evaluated and modeled.And fitting comparison is made between the results of the GP-GWO algorithm with experimental data and the feasibility of the algorithm is thus verified.

关 键 词：负荷自动建模 遗传程序设计 灰狼优化算法 准确性 

分 类 号：TM714[电气工程—电力系统及自动化]