ALDDPS:Abnormal Load Detection Based on Biased Random Key Genetic Algorithm and Denoising Self-encoder of Distributed Power Station

This paper presents a method for detecting abnormal load in distributed power stations using a denoising autoencoder algorithm optimized by a biased random key genetic algorithm. Our method uses electricity load data to train a denoising autoencoder model. Based on the model reconstructed data and the original data, the residuals are calculated and compared with the set thresholds to determine whether abnormal load situation occurs. Parameters of the denoising autoencoder are optimized using the biased random key genetic algorithm, while the threshold is dynamically generated. Experiments are conducted on the real dataset and the results demonstrate the performance of the proposed method in detecting abnormal load compared to other algorithms.