| Distributed power generators are the major components of microgrid.The intermittent and volatility of distributed power generators and the access of a large number of power electronic devices have brought great challenges to the power quality of microgrid.Aiming at the shortcomings of the existing power quality assessment methods,such as high subjectivity,poor accuracy of comprehensive assessment,neglecting the operation characteristics of microgrid and poor follow-up control effect of evaluation,the thesis proposes the estimation method of available capacity of battery energy storage system in microgrid based on active lithium content.The deep belief network model for the power quality assessment in microgrid based on battery energy storage is established.In the improved 14-bus microgrid,the battery energy storage system is used for controlling power quality.All of the above were verified by simulation.Firstly,the power quality problems in the microgrid are introduced.The economic and practical significance of power quality assessment and control are analyzed.The current status of the methods for power quality assessment and the respective advantages and disadvantages of them are summarized.The similarities and differences between the power quality control of traditional method and the energy storage-based method in microgrid are introduced and compared.Secondly,the thesis analyzes the different problems of power quality brought by different distributed generations in microgrid,compares the advantages and disadvantages of different energy storage technologies in microgrid,and analyzes the positive effect of battery energy storage system on power quality assessment and control in microgrid.Thirdly,the electrochemical model of LiFePO4 battery is established.The equivalent circuit model is established based on the analysis of the working principle of LiFePO4 battery.For the shortcomings of the existing SOC estimation methods for Lithium battery energy storage system,this thesis proposes the estimation method of SOC for lithium battery based on the content of active lithium in electrode particles,and establishes a SOC prediction model.The effectiveness of the SOC estimation method is demonstrated by comparison with the results of the equivalent circuit model.Then,the thesis proposes a method of deep learning to assess the power quality indicators,builds the deep belief network and a power quality assessment model at the deep belief network.The SOC prediction result of the battery energy storage system is used as another indicator of the power quality assessment in microgrid.The power quality assessment model of microgrid based on deep belief network is established.The simulation verification is completed in the MATLAB/Simulink environment.Finally,power quality of different assessment levels is hierarchically controlled.The control of the battery energy storage system in the microgrid is realized by the virtual synchronous generator method.The 14-bus model of the microgrid in the MATLAB/Simulink environment is built.The thesis analyzes the power quality of the microgrid system under disturbance when the battery is under power quality hierarchical control or not.The positive effect of battery energy storage system in power quality hierarchical control of microgrid has been verified. |
