Research On Parameters And Topology Identification And Network Reconfiguration Of Distribution System

Accurate parameter estimation,topology identification and reasonable network reconfiguration are three essential issues to realize optimal dispatch and operation for intelligent distribution network.Among them,correct network parameters and accurate network topology are the basis of distribution network reconstruction,which is one of the important functions of distribution automation.Parameter estimation,topology identification and network reconstruction of distribution network are studied in this thesis and some results have been achieved.The contents and results of this thesis are as follows:Firstly,this thesis proposes a parameter identification method for the distribution line and distribution transformer based on advanced meter infrastructure(AMI).Nonlinear equations involved line impedance parameters of low-voltage distribution network and distribution transformer parameters are established by the method based on multi-period measurement information provided by AMI measurement system.Least-square method is then applied to solve the nonlinear equations to realize the identification of distribution network lines and transformer parameters based on AMI measurement information.The method proposed in this thesis improves the accuracy of distribution network parameters identification effectively.Secondly,this thesis proposes a topology identification method for low-voltage distribution network based on AMI measurement information.Firstly,the voltage of the coupled node and the branch current of each load under the same distribution transformer are obtained based on the multi-period measurement information provided by AMI.Correlation analysis of the voltage of the coupled node and the branch current obtained is then conducted,thus the feeder of the load can be confirmed.The upstream and downstream relationship of each load in the feeder can be further determined according to voltage distribution of the coupled node of the load.Finally,load with wrong connection relationship is revised based on correlation analysis results and voltage distribution of coupled nodes of the load.Since then the verification and correction of low-voltage distribution network topology is completed.The method proposed in this thesis improves the accuracy of distribution network topology identification effectively.Lastly,this thesis proposes a dynamic reconstruction method for distribution network considering distributed power supply and time-varying loads.Firstly,a comprehensive evaluation index based on system voltage offset and load balance is established based on operating characteristics of the system.A time division method based on the index is established thus the distribution network can be reconstructed separately.Then an improved Mayeda spanning tree algorithm is proposed based on branch exchange method.The branch exchange method is used to select the branch used for switching quickly from the candidate set,instead of traditional Mayeda spanning tree algorithm which select the switching branch randomly.That is to say,an optimization technique is applied to eliminate the tree which is worse than the current tree in traditional Mayeda spanning tree algorithm,so that the search process of the optimal tree is significantly accelerated and the computational efficiency of the algorithm is improved.The method proposed in this thesis promotes the process of online implementation of distribution network reconfiguration.The research and achievements of this thesis promotes the practical process of distribution network scheduling automation effectively.