Building low-carbon energy management method based on reinforcement learning

The invention discloses a building low-carbon energy management method based on reinforcement learning, and the method comprises the following design steps: taking a building integrated energy system as an intelligent agent in reinforcement learning, building a simulation environment of building energy management, and determining system operation constraints in the environment, including a building energy balance equation, a building energy balance equation, a building energy balance equation and a building energy balance equation; calculating an operation constraint equation of the energy device; under the condition of considering new energy consumption and building carbon emission, an energy management problem of minimizing building low-carbon operation cost is modeled as a partially observable Markov decision process, and corresponding environment states, actions and reward functions are designed; and a near-end strategy optimization algorithm is utilized to train optimal actions of various energy conversi