Trade-off between Number of Meters and Bad Data for Hamiltonian Cycle Based Distribution State Estimation

Abstract

There are many issues involved in state estimation such as how to determine the location of the measurement and obtain good state estimation results This paper focuses on the trade-off between the number of meters and bad data that proposes a state estimation method based on Hamiltonian loop estimation which can estimate value of the unknown state of the whole network only by meter data in low error The advantage of this method is simple fast and accurate that because the pseudo-measurement relationship with high error is not included Also this method can recognize the influence state of the bus by the specific bad data In order to verify the effectiveness of the method three different types of three-phase distribution systems with four different types of measurement cases considered by simulation hypothesis data acquisition method bad data elimination method and the estimation error tolerance This simulation use DIgSILENT to build the bus system model DPL (DIgSILENT programming language) to generate the modeling parameters in MATLAB and MATLAB to establish the mathematical model and to estimate In the part of state estimation there are weighted least square method and Hamiltonian loop estimation method In the part of Bad data detection and identification there are chi-square verification maximum residual normalized residual method and maximum whitened residual method In different condition with the number of meters for those identified bad data will be replaced by related value and simulation shows that the result meets the original specifications and standards

Date of Award	2019
Original language	English
Supervisor	Hong-Tzer Yang (Supervisor)