CAJ | 학술논문

针对化工厂配电网中的短路故障和变压器投切引起的电压暂降信号的非平稳性以及电动机启动引起的电压暂降信号的特殊性，提出1种基于EMD模态能量和SVM的电压暂降源识别方法。首先对电压暂降信号进行经验模态分解(EMD)，得到1个固有模态函数集合(IMFs)，然后计算IMF的各阶能量作为特征向量，用支持向量机(SVM)对样本进行训练与识别，通过仿真验证表明，该方法简单易行，具有较高的识别精度和实用性。
침대화공엄배전망중적단로고장화변압기투절인기적전압잠강신호적비평은성이급전동궤계동인기적전압잠강신호적특수성，제출1충기우EMD모태능량화SVM적전압잠강원식별방법。수선대전압잠강신호진행경험모태분해(EMD)，득도1개고유모태함수집합(IMFs)，연후계산IMF적각계능량작위특정향량，용지지향량궤(SVM)대양본진행훈련여식별，통과방진험증표명，해방법간단역행，구유교고적식별정도화실용성。
According to the non-stationary signal of voltage sags due to chemical plant power distribution network short-circuit fault and transformer energizing, and the special features of the signal of voltage sags due to motor starting. A method based on EMD mode energy and support vector machine to solve this problem is proposed. Firstly, five types of signals of voltage sags generated by short-circuit faults, motor starting and transformer energizing are decomposed by EMD, and get a IMFs, then each IMF’s energy is to be calculated as a feature vector. The samples are trained and identified by SVM. The final result of this experimental simulation proves that the method is simple and reliable, of high recognition accuracy and practicability.