应用逻辑回归法辨识风电外送系统是否产生次同步振荡的研究  

Research of the Application of Logistic Regression Method to Identify Whether the Wind Power Transmission System Generates Subsynchronous Oscillation

作　　者：党存禄[1] 刘姗姗 李旭鹏 DANG Cunlu;LIU Shanshan;LI Xupeng(Lanzhou University of Technology,Lanzhou 730050)

机构地区：[1]兰州理工大学,兰州730050

出　　处：《计算机与数字工程》2024年第9期2614-2619,2625,共7页Computer & Digital Engineering

摘　　要：风电外送系统运行环境复杂,电力电子设备、风速大小、串补度大小都可能诱发风电外送系统产生次同步振荡(Sub-synchronous oscillation,SSO)。因此提出一种利用机器学习的方法进行辨识风电外送系统是否产生次同步振荡。首先在pscad仿真系统中创建风电外送系统模型,再通过所创建的风电外送系统模型创建数据集,使用python语言编写预测系统,将所创建的数据集代入预测系统进行训练。论文创建一组新的测试数据代入已经训练好的系统得到测试结果。将结果与pscad中的结果进行对比,验证预测系统的可行性。该预测系统运行环境简单,运行速度较快,并且可以同时预测多组数据。The operating environment of the wind power transmission system is complex,and the power electronic equipment,wind speed size,and series complement degree may induce the power system to generate sub-synchronous oscillation(SSO).In this paper,a method using machine learning is proposed to predict whether a wind power transmission system will generate subsynchronous oscillation.First of all,the wind power transmission system model is created in the pscad simulation system,and then the data set is created by the created wind power transmission system model,the prediction system is written in python language,and the created data set is substituted into the prediction system for training.This paper creates a new set of test data to substitute into the trained system to get test results.It compares the results with those in pscad to verify the feasibility of the predictive system.The forecasting system operates in a simple environment,runs faster,and can predict multiple sets of data at the same time.

关 键 词：机器学习 次同步振荡 预测系统 逻辑回归 

分 类 号：TM614[电气工程—电力系统及自动化]