Deep Learning Detection of Electricity Theft Cyber-attacks in Renewable Distributed Generation

Abstract

Performance evaluation of various deep learning algorithms such as DNN (deep feed forward neural network), RNN GRU and CNN for electricity cyber-attack detection is performed in this application. Now-a-days in advance countries solar plates are used to generate electricity and these users can sale excess energy to other needy users and they will be maintained two different meters which will record consumption and production details. While producing some malicious users may tamper smart meter to get more bill which can be collected from electricity renewable distributed energy. This attack may cause huge losses to agencies. To detect such attack we are employing deep learning models which can detect all possible alterations to predict theft. In all the models CNN is giving better detection accuracy. As extension we have added Hybrid Random Forest algorithm which will extract optimized features from CNN algorithm and then retrain itself to get better accuracy. Random Forest get trained on CNN filtered features and it has better quality of data so its prediction accuracy may get better. For example, if you get good quality of raw material then u will come up with better product development and same will be applied to Hybrid Random Forest.