Research On Energy Regulation Strategy Of Low-voltage Microgrid Based On Model Predictive Contro

The problems of energy shortage and environmental pollution have forced the rapid transformation and upgrading of global energy utilization.Under the strategy of implementing the energy internet in China,the interconnection of multiple micro networks is more economical and reliable compared to isolated single micro networks.Through energy interaction and coordination,multiple micro networks in low-voltage stations can promote the local consumption of new energy and alleviate the power supply pressure in low-voltage stations.Based on this,this paper proposes a two-level energy regulation method for connecting interconnected multi micro grid systems to low-voltage substations(low-voltage distribution networks).The main work contents are as follows:(1)Aiming at the energy coupling relationship between interconnected multi micro grid and low-voltage distribution network,a two-level energy operation framework is constructed;Analyze the operating characteristics of load,energy storage,and controllable distributed power generation in a microgrid system,and establish corresponding mathematical models.(2)Aiming at the problem that the proposed CNN-BILSTM load forecasting model has a large number of parameters,and manual parameter adjustment not only requires a large amount of work,but also easily leads to model overfitting or falling into local optimization.A whale optimization algorithm is used to automatically optimize the CNN-BILSTM combined model to obtain high-precision load forecasting data.(3)A short-term photovoltaic output prediction method based on LSSVM and VMD technology is proposed.VMD mainly decomposes photovoltaic power data with complex components to obtain more stable subsequences;The logarithmic decreasing adaptive weight and firefly optimization strategy are introduced to improve the SSA,and then the improved SSA algorithm is used to optimize the LSSVM model.Through breakthroughs in multiple key technologies,a more accurate photovoltaic output prediction model has been obtained.(4)A two-stage operation optimization control strategy is proposed based on the dual layer architecture of multi micro grid and low voltage distribution network.In the day ahead scheduling phase,both the upper and lower levels of the system take the lowest operating cost as the optimization goal,and use the goal cascade analysis method to iteratively solve the two-level model.The convergence of the solution algorithm is verified through IEEE33 numerical example simulation,and the optimal scheduling results of multiple microgrids are output.In the intra day scheduling phase,in order to better solve the problem of deviation between the actual situation and the day ahead scheduling plan,this paper proposes a rolling horizon optimization control strategy based on MPC.The example shows that the rolling horizon optimization control is implemented by MPC while the predicted output value is constantly revised,which largely eliminates the impact of system uncertainty,thus ensuring the effectiveness of the day ahead scheduling plan,Ensure the stable operation of the system.