Building air conditioner load scene generation method, device and equipment based on conditional time sequence GAN

• Main Authors: ZHU XIAO, LI SIQI, NI MINJUE, YANG TING, WANG YUYING, FAN WENYAO, LU DANHONG
• Format: Patent
• Language: Chinese; English
• Published: 09.08.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention provides a building air conditioner load scene generation method, device and equipment based on a conditional time sequence GAN, and the method comprises the steps: obtaining a historical real data set, including static features and dynamic features, of air conditioner load energy consumption of a building application scene, carrying out the preprocessing, and taking data of a set proportion as a training sample set; constructing a conditional time sequence GAN model; obtaining a global optimization loss function of the conditional time sequence GAN model; training the constructed conditional time sequence GAN model through unsupervised adversarial training and supervised training to obtain a trained conditional time sequence GAN model; inputting scene generation characteristic conditions into the trained condition time sequence GAN model, and taking an output data set as an air conditioner load data scene set; the method can better capture the step-by-step dynamic probability distribution of th