基于生成对抗网络的窃电检测方法

Power Theft Detection Method Based on Generative Adversarial Networks

针对智能电网背景下的窃电数据不平衡问题,提出了一种基于生成对抗网络(GAN)的窃电检测方法。生成器负责生成近似于真实的虚假样本去欺骗判别器,判别器负责区分真实样本和生成样本,通过二者的对抗训练来提高生成器的生成能力和判别器的判别能力,进而提高窃电检测的正确率与抗干扰能力。算例结果验证了所提算法的准确性和优越性。

To address the imbalance problem of power theft data in the context of the smart grid,a method of power theft detection based on generative adversarial networks(GAN)is proposed.The generator is responsible for generating false samples to deceive the discriminator,and the discriminator is responsible for distinguishing the real samples from the generated samples.The generation ability of the generator and the discrimination ability of the discriminator can be imporved through the adversarial training of the two,thus the accuracy and anti-interference ability of power theft detection can be improved.The accuracy and superiority of the proposed algorithm are validated by the example results.