Photovoltaic coordination autonomous method based on multi-agent coordination control strategy and reinforcement learning

• Main Authors: LU HAIJUN, XU XIAO, LI WEIZHUO, XU YUNBING, LUO WEIQI, ZHOU DAMOU, GAO HUI
• Format: Patent
• Language: Chinese; English
• Published: 09.01.2024
• Subjects: CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTINGELECTRIC POWER; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER; ELECTRICITY; GENERATION; SYSTEMS FOR STORING ELECTRIC ENERGY

The invention discloses a photovoltaic coordination autonomous method based on a multi-agent coordination control strategy and reinforcement learning. The method comprises the following steps: acquiring various data of a photovoltaic inverter and energy supply and demand data of energy consumption equipment of each node; designing an initial multi-stage agent grid-connected coordination control strategy based on a layered distributed architecture; training a reinforcement learning model based on the historically collected photovoltaic data and energy supply and demand data; optimizing an initial multi-stage agent grid-connected coordination control strategy based on a reinforcement learning model; and determining the photovoltaic output of each current node and the demand adjustment amount of energy consumption equipment based on the optimal multi-stage agent grid-connected coordination control strategy. The method has the remarkable effects that the effective consumption of photovoltaic energy is realized, t