Research On Voltage Characteristics And Assistant Decision-Making For Voltage Governance Of MV Feeder

The voltage quality of power distribution network is directly related to economic efficiency of power grid company and safe operation of subscriber equipment.Facing increasingly complex distribution network structure,various causes of voltage problems and so many choices of governance measures,power grid company generally rely on subjective experience to select governance measures for feeders with voltage problems,lacking of governance pertinence and investment effectiveness.Therefor,this thesis carries out the research on voltage characteristics and assistant decision-making for voltage governance of medium voltage feeder.Firstly,an evaluation method of voltage characteristics along the medium voltage feeder considering incoming voltage in transformer area is designed.Voltage curve deviation index of feeder can be obtained by combining the voltage probability distribution information of all transformers under the same feeder with their relative importance.Based on this,the feeders with poor voltage characteristics can be screened out,which can provide the research object for the following text.Secondly,a typical classification method of medium voltage feeder is proposed.A clustering index system is constructed by considering the internal influencing factors of voltage deviation.The feeders with voltage problems are sorted by Fuzzy C-mean clustering algorithm so as to provide learning samples for subsequent optimization model.On the basis of the previous part,a qualitative selection model of voltage governance measures for medium voltage feeder based on Generalized Regression Neural Network is designed.The corresponding relationship between the properties of clustering indexes and various governance measures is analyzed to determine which measures are suitable for each type of feeders.The feeder clustering results with targeted measures are used as learning samples to train GRNN model,which can qualitatively put forward suggestions on the selection of voltage governance measures for medium voltage feeders.Then,a multi-objective optimal allocation method of voltage governance measures is proposed.The three objective functions of line loss rate,voltage quality and equipment capacity are processed as a single objective function by weighted,and different optimization variables are set for various governance measures.The improved genetic algorithm is used to solve the model.from the perspective of voltage quality improvement,line loss rate,investment cost and efficiency,a comprehensive evaluation of several targeted measures is made.Thus,the best governance scheme can be determined quantitatively.Finally,combining the above research contents,a set of system is designed,including three modules: voltage characteristic evaluation of medium voltage feeder,targeted governance measures selection and optimal governance scheme design.According to the difficulty of obtaining the required basic data,the system can qualitatively give suggestions on the selection of governance measures or quantitatively output the best governance scheme of medium voltage feeders whose results of voltage characteristic evaluation are poor.