| Electric buses have high charging frequency and large charging capacity,and the requirements for regional power supply capacity are constantly improving.At the same time,the large-scale electric vehicle load access will have a large impact on the grid and affect its stable operation.For this reason,in the background of the current "double carbon" policy,the establishment of solar-storage charging stations can not only solve the problem of high load demand,but also reduce the impact of load on the power grid and make full use of more renewable energy,which will become the future development trend of charging facility construction.Therefore,based on the solar-storage charging station system,the thesis proposes the electric buses load prediction method,capacity gaming configuration method of photovoltaics and energy storage and the optimal scheduling method of energy storage system,which provide new ideas for the overall planning and stable and economic operation of the solar-storage charging station system.First,the structure of the solar-storage charging station system is analyzed,and then the characteristics of each part of the photovoltaic power generation unit,energy storage system and charging station load are explained separately.The basic principle and output characteristics of photovoltaic power generation unit are introduced specifically,on which the actual output of photovoltaic power generation is combined to establish the photovoltaic output uncertainty model.Then the battery type,equivalent circuit model and battery charge/discharge characteristics curve of the energy storage system are described,and the battery capacity decay characteristics are further analyzed.Finally,the specific intermittent and fluctuating characteristics of the charging station load and the electric buses charging law were analyzed based on the actual charging station data.Secondly,the electric buses load curves characteristics are analyzed,and an electric bus charging load prediction method based on Bi-directional long short-term memory neural network(Bi-LSTM)is proposed according to the differences of electric buses charging load timing characteristics of different routes,which have intermittent,random and fluctuating characteristics.Improved spectral clustering with dynamic time warping similarity was used to cluster the load curves,ensuring that curves with the same temporal order and morphology were grouped,and a Bi-LSTM neural network model incorporating the Attention mechanism was built for each class of groups separately.Lay the foundation for optimal allocation and optimal control of solar-storage capacity.After that,the operation strategy of the solar-storage charging station is formulated based on the real-time tariff policy,and game theory is introduced to analyze the capacity configuration of the solar-storage charging station.Considering the role and status of photovoltaic and energy storage,the Stackelberg game model with photovoltaic as the dominant player and energy storage as the follower is established,and the objective function of the model is to maximize the sum of annual economic benefits of photovoltaic and energy storage.The model is solved using an improved artificial bee colony algorithm,and the accuracy and economy of the model are verified by example analysis.Finally,based on the load forecast of the solar-storage charging station and the determination of photovoltaic and energy storage capacity configuration,the current state of energy storage,the future output of photovoltaic and the charging load demand are considered,and the charging station daily operation plan is updated on a rolling basis.An optimal energy storage scheduling model with minimizing energy storage cost as the rolling optimization objective function and upper and lower SOC charging and discharging limits as the constraints is established.The simulation results show that this optimal control approach can eliminate load spikes,smooth out photovoltaic fluctuations,and improve daily revenue. |
PDF Full Text Request