Prediction model training method based on big data and photovoltaic power generation performance detection method

• Main Authors: GUO JIANBO, ZHANG LEI, JIN ZHIBIN, LU LINGLING, ZHANG WEI, ZHANG QIANG, HE JIE, TAO YONGJING
• Format: Patent
• Language: Chinese; English
• Published: 24.11.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; ELECTRICITY; GENERATION; GENERATION OF ELECTRIC POWER BY CONVERSION OF INFRA-REDRADIATION, VISIBLE LIGHT OR ULTRAVIOLET LIGHT, E.G. USINGPHOTOVOLTAIC [PV] MODULES; PHYSICS

The embodiment of the invention discloses a prediction model training method based on big data and a photovoltaic power generation performance detection method. The training method comprises the following steps: carrying out model training on an LSTM common model based on a first training set, respectively training a plurality of LSTM independent models based on a plurality of second training sets, and judging, screening and determining an optimal model as a photovoltaic power generation power prediction model by utilizing a first loss function; and then, performing hyper-parameter training on the photovoltaic power generation power prediction model based on the test set, and performing prediction judgment by using a second loss function to obtain a photovoltaic power generation power prediction model with an optimal hyper-parameter. According to the detection method, power generation power prediction detection is carried out based on a photovoltaic power generation power prediction model obtained through tra