Research On Optimal Scheduling Of Integrated Energy System Including Electricity Hydrogen Conversion Based On Deep Learning Scenario Generation

The problem of global environmental pollution is getting worse,and the structure of energy supply and demand needs to be adjusted urgently.Rational use of renewable energy represented by wind power and photovoltaics is an important way to improve the level of green and sustainable development of the power grid,but the randomness of wind and solar energy brings new challenges to the stable operation of the power system.As a green energy,hydrogen energy has become the research focus of countries all over the world,and it is the key to solving energy problems and realizing energy conservation and emission reduction under the "double carbon" goal.The integrated energy system couples a variety of heterogeneous energy sources and is bound to become one of the mainstream energy supply methods in the future.In this context,it is of great theoretical significance and practical value to construct an integrated energy system containing electricity-hydrogen conversion,to study how to describe the uncertainty of wind and solar output,to rationally utilize hydrogen energy,and how to improve the absorption level of renewable energy and reduce the carbon emissions of the system.The main research work of this paper is as follows:(1)Construct the whole structure of the integrated energy system containing electricity-hydrogen conversion.Analyze the coupling relationships of heterogeneous energy subnetworks in the integrated energy system,and their operating characteristics and mechanisms are deeply analyzed.Considering the electricity-hydrogen conversion,the mathematical models of electrolyzer and hydrogen fuel cell are given.The mathematical models of electrical power subsystem,thermal subsystem,natural gas subsystem and various coupling equipment are established.(2)Aiming at the uncertainty of renewable energy,a renewable energy output scenario generation method based on deep learning is proposed;an improved VAE-DCGAN model is established,and a deep convolutional generative adversarial network and variational autoencoder are used for unsupervised self-learning of historical data,use gradient penalty items to strengthen Lipschitz constraints,and rely on data-driven to obtain and generate renewable energy output scenarios;simulation results verify the effectiveness of the model.(3)An optimal scheduling method for integrated energy system containing electricity hydrogen conversion based on deep learning scenario method is proposed;construct an optimal scheduling model with the goal of minimizing the overall system cost,use the synchronous back reduction method to obtain typical scenarios of wind and solar output,and use the improved inverse function one-dimensional approximation method to linearize the model;the effectiveness of the proposed method is verified by the simulation of the integrated energy system consisting of IEEE39-node power system,Belgium 20-node natural gas system and Belgium 6-node thermal system.(4)In order to strengthen the initiative of each energy supplier to actively improve the consumption level of renewable energy and reduce the carbon emissions of the system,the role of the market was considered and carbon trading and green certificate trading mechanisms were introduced to construct an optimization scheduling model of the integrated energy system containing electricity-hydrogen conversion considering carbon-green certificate trading;taking the lowest system comprehensive cost as the objective function,the method of scenario analysis is used for day-ahead optimal scheduling,and the effectiveness of the proposed method is verified by an example.In summary,a renewable energy output scenario generation method based on deep learning is proposed,which can stably generate samples that conform to the historical real output law;an optimal scheduling method of integrated energy system containing electricity-hydrogen conversion based on deep learning scenario method is proposed,this method can effectively improve the level of renewable energy consumption in the system;an optimal scheduling model of the integrated energy system containing electricity-hydrogen conversion considering carbon-green certificate trading is constructed,which can effectively improve the initiative of energy supply entities to actively reduce carbon emissions,and further improve the consumption level of renewable energy and reduce the comprehensive cost of the system.