Distributed Multi-Phase Distribution Power Flow: Modeling, Solution Algorithm and Simulation Results

Distributed Multi-Phase Distribution Power Flow: Modeling, Solution Algorithm, and Simulation Results Michael R. Kleinberg Karen Miu, Ph.D. With the increasing presence of distributed intelligence throughout power distribution systems, the possibilities for distributed control and operation schemes are becoming progressively more attractive and feasible. Distributed operations will require tools to properly assess and predict the present and future status of the system in order to make proper control decisions. Multi-phase distribution power flow is a basic tool which calculates the operating state of the distribution system and is used to support all other applications. Therefore, this thesis will present a new method for calculating distribution power flow using physically remote distributed processors. The proposed power flow requires the distribution system to be partitioned with each partition distributed to a remote processor. Each processor then only requires detailed information about the portion of the network it will represent. Distributed analysis component models for multi-phase distribution systems have been developed to model the remaining network not explicitly retained in each partition. These models are embedded in a new distributed algorithm for multi-phase distribution power flow. Properties of the converged solution of this algorithm have been investigated and will be reported. A distributed processor test bed was designed to emulate the distribution of intelligent devices throughout power distribution networks and simulations were conducted to asses the proposed algorithm. Results have shown the proposed method