A FDIA Detection Method Based on Similar Month and Discrete Wasserstein Distance in Smart Girds

False Data Injection Attacks (FDIA) have emerged as a critical threat to smart grid security by injecting sophisticated false data, which can lead to severe operational disruptions. Existing statistical analysis-based detection methods face a trade-off between detection accuracy and false positive rate, as they fail to fully capture the spatiotemporal features of power system measurement data. To address this, we propose a FDIA detection method by uti-lizing the properties of Similar Month, Joint Transformation and Discrete Wasserstein Distance (sMJT-dWD). First, the month similarity is calculated through a spatiotemporal rank correlation analysis of multi-node load sequences, which helps eliminate the influence of spatiotemporal variations on power grid dynamics extraction. Next, a joint transformation algorithm enhances the discriminative features of measurement variations by reconstructing the spatial distributions of these variations. Finally, to quantify distributional deviations, we apply discrete wasserstein distance to guarantee robustness of FDIA detection. Extensive experiments on the IEEE 14-bus system validate the effectiveness of the proposed (sMJT-dWD) method. The results indicate that the proposed method attains a 99.99% detection rate for 10% and 5% FDIA, and 99.9% for 1 % FDIA, with a false positive rate below 0.1 %.