Topology and Line Parameter Identification Using Kernel Density Estimation in Distribution Networks

Distribution network topology and line parameter identification are vital in developing a resilient grid, which can realize real-Time monitoring and supervision of the system status, assist the operator in identifying potential risks, and enable fast diagnosis and recovery ability to increase the overall flexibility and reliability. This paper proposes a data-driven method to simultaneously identify and correct the topology and line parameters in the non-PMU distribution networks. In the first stage, the method will provide a preliminary identification of network topology and a rough estimation of line parameters based on limited measurement data. The mutual information calculated based on kernel density estimation is introduced to help correct the topology based on voltage measurement data. Then, in the second stage, an iterative method based on Newton-Raphson is presented. The voltage drop model has also been adopted for further correction of the line parameter estimation. The proposed method is implemented on IEEE 33 and 123 distribution systems to validate the effectiveness of the proposed method under noise effect with only limited measurements. The simulation results show the proposed method can have an estimation accuracy of 98% and realize near real-Time monitoring of the distribution system even under noise interference.