Distribution state estimation based on voltage state variables: assessment of results and limitations

Abstract

Distribution networks are facing significant technical constraints that must be resolved at minimum cost: regulatory and customer pressures to improve reliability of supply, reduction of losses and the technical challenges related to the connection of an increasing amount of distributed generation. The use of new approaches is therefore being encouraged to develop cost-effective solutions often leading to the concept of Active Management. However, the lack of monitoring in present distribution networks is a barrier to the implementation of the corresponding automation functions. On the other hand a full monitoring of the distribution network is not economically realistic. Therefore, there is a need for intermediate solutions such as the one presented in this paper: EDF R&D has developed a distribution state estimator (DSE) based on voltage state variables. Such algorithm is expected to provide quasi-real time system supervision of voltages and power flows which are required as inputs of Distribution Management System (DMS) functions. Even if based on techniques applied to transmission networks, the requirements and performances of DSE are different and adapted to distribution needs. The paper discusses the limits of such approach and proposes thumbed rules for sensors placement as well as research perspectives to overcome these limits.

Principle and evaluation of a self-Adaptive Multi-Agent System for State Estimation of electrical distribution network

“…The work presented in [3] shows that the transmission networks techniques can be used for distribution systems. However, the performances are reduced and such a system is only able to estimate accurately some variables.…”

“…The error •ltering realized by Atena is a •rst step toward a robust and ef•cient Distribution System State Estimation, however it is far less ef•cient than classical approaches derived from transmission network state estimation such as the one presented in [3].…”

“…It is therefore necessary to !nd alternative ways to determine the state of a network. This introduces the concept of "Distribution State Estimation" [3]. The !rst ones who have published about this are Roytelman and Shahidehpour in their publication called "State estimation for electric power distribution system in quasi real-time conditions" [4].…”

“…This approach consists in minimizing the weighted square of errors between the model and measured values [4], [9], [3].…”

Choice of state estimation solution process for medium voltage distribution systems

“…By replacing the value for +1 and putting in the minimization problem [19] ∑ | =1 | (11) As the method consisted of one inner and one outer loop operation, the inner loop iteration is capped at a smaller number of iterations (which is five here) and allowed to terminate even if convergence is not achieved. The outer loop is the main control to decide when the algorithm is satisfying the convergence criteria and should terminate.…”