Letter Detecting the One-Shot Dummy Attack on the Power Industrial Control Processes With an Unsupervised Data-Driven Approach

Dear Editor, Dummy attack (DA), a deep stealthy but impactful data integrity on power industrial control processes, is recently recognized as hiding the corrupted measurements in normal measurements. In this letter, targeting more practical case, we aim to detect one-shot DA, with purpose of revealing DA once it launched. Specifically, first formulate an optimization problem generate DAs. Then, unsupervised data-driven approach based modified local outlier factor (MLOF) proposed them. To improve detection performance, are preprocessed gamma transformation and patterns extracted from historical integrated into MLOF algorithm. Finally, extensive experiments conducted evaluate performance real-world load data.