Generate distribution circuit model using AMI data

This paper introduces a method of accurately estimating the impedance of branch circuit in distribution secondary network provided the distribution network topology is known from GIS. By leveraging a series of smart meter measurements at end nodes of the distribution network, mainly voltage, active power and reactive power at customer sites, it calculates the voltage of the most upstream node or source node in the radial network through bottom up approach. Considering the fact that the circuit parameters usually won´t change in a period of times, an optimisation model is built to minimize the source node voltage variance derived from different meter node measurements over a period of times, and thus obtain the best estimates for branch impedances. In case that direct measurements are available for the source node, the objective function of the proposed method is revised to minimise the differences of the estimated source node voltage from different smart meter measurements with actual voltage measurement in the source node. The proposed optimisation method is tested with a practical distribution network and the results achieved confirm its effectiveness of the proposed method. The results also show that direct measurement in the source node, the estimate accuracy is further improved than the case without this direct measurement.