Optimal number and location of measurement instruments in distributed systems for Harmonic State Estimation

The recent studies about the attribution of the responsibilities for harmonic distortion in electric distribution systems show that the most promising approaches are based on measurement data acquired simultaneously in different nodes of the network. Therefore, distributed measurement systems are needed. For economic reasons, only a small part of network nodes can be equipped with measurement instruments, whereas the information about the overall status of the electric system has to be evaluated by means of suitable Harmonic State Estimation (HSE) techniques.
This paper deals with the problem of choosing the optimal number and position of measurement devices, in order to ensure that the HSE algorithms can provide, at the lowest possible cost, results having a prefixed level of accuracy.
An optimization algorithm based on the techniques of the dynamic programming is proposed and the way to take into account the uncertainty introduced by all the elements of the measurement system is discussed.
Simulation results on a benchmark distribution network show the validity of the proposed approach.