Intelligent energy cyber physical systems (iECPS) for reliable smart grid against energy theft and false data injection

The Internet of Things (IoT)-based advanced metering infrastructure (AMI) provides real-time information from smart meters to both grid operators and customers. The meter data is collected and stored digitally and is transmitted through wireless networks. With the remarkably increasing usage of smart grid systems over conventional utility grid networks, it becomes imperative to ensure that such a system is secured from potential threats. Energy theft by meter tampering and false data injection (FDI) attacks are the most common forms of attacks that attempt to compromise the security of a smart grid system. In this paper, we introduce a new machine learning-based multi-model prediction system known as intelligent energy cyber physical systems (iECPS) for smart energy theft detection and verification. Furthermore, the system has more robustness since a verification from the users themselves (through e-mail) is considered as final validation to decide further course of action. In addition, to successfully detect and prevent FDI attack during transmission of data from sub-meters or meters to smart grids, a watermarking and de-watermarking system is also proposed. It involves encryption and decryption of meter data by means of a PN-sequence available at both ends. The proposed scheme detects electricity theft performed by tapping on the user’s smart meter with high accuracy. Also, the watermarking scheme discerns FDI attacks while maintaining data consistency, with a significant similarity factor and low channel SNR. The proposed system has accuracy of 99.77%, precision 0.998, recall 1.0, F1 score 0.999 and FPR 0.999 which has better results than similar previous works.