Abstract: A model generating apparatus and method are provided. The apparatus receives a plurality of sample images. The apparatus generates a plurality of adversarial samples corresponding to the sample images. The apparatus inputs the sample images and the adversarial samples respectively to a first encoder and a second encoder in a self-supervised neural network to generate a plurality of first feature extractions and a plurality of second feature extractions. The apparatus calculates a similarity of each of the first feature extractions and the second feature extractions to train the self-supervised neural network. The apparatus generates a task model based on the first encoder and a plurality of labeled data.

Abstract: A power control method optimized against both a shortfall in supply and an oversupply, comprising acquiring scheduling information of a plurality of power supply lines, the scheduling information including electric power data, status data, and operation data of the power supply lines; inputting the electric power data into a first training model to obtain a recommended schedule of turning on the plurality of power supply lines; modifying the recommended schedule according to the status data and the operation data of each of the plurality of power supply lines to obtain an optimal power supply line; turning on the loads connected to the optimal power supply line. The method recommends power supply lines according to the equipment being used in actuality and historically. A power control system and a power control device, are also provided.