Improvement of observability in poorly measured distribution networks

The paper addresses the problem of observability in poorly measured distribution networks (DNs). In the lack of redundancy and statistical information about the measurements an innovative approach is proposed: it consists in minimizing the sum of the squares of the differences between measured and estimated values of the quantities provided by the measurement equipment present in the system. The proposed technique is based on the Interior Point method. The benefits of the approach in terms of observability improvement are illustrated by simulations involving realistic MV distribution system models. The results presented in the paper represent a part of the INGRID 2 project, which is the product of the collaboration among Politécnico di Milano, SIEMENS SpA and Universita degli Studi di Milano and represents a tool developed to answer the above mention needs of the Distribution System Operator (DSO).