No abstract

This paper presents the proof of concept for a new solution to the problem of recomposing missing information at the SCADA of energy/distribution management systems (EMS/DMS), through the use of offline trained autoencoders. These are neural networks with a special architecture, which allows them to store knowledge about a system in a nonlinear manifold characterized by their weights. Suitable algorithms may then recompose missing inputs (measurements). The paper shows that, trained with adequate information, autoencoders perform well in recomposing missing voltage and power values, and focuses on the particularly important application of inferring the topology of the network when information about switch status is absent. Examples with the IEEE RTS 24-bus network are presented to illustrate the concept and technique.

SUMMARYThis paper addresses the extension of the microgrid concept, following a massive integration of these active cells in power distribution networks, by adopting a coordinated management strategy together with distributed generation units directly connected to the medium voltage distribution network. In order to achieve this, a technical and commercial management scheme must be developed for coordinated control of a distribution system with multi-microgrids, which should take into account the specific technical capabilities and characteristics of each type of generating source. In particular, tools for coordinated voltage support and frequency control, as well as for state estimation have been developed for this type of network. Concerning voltage support, a new methodology exploiting an optimization tool based on a metaheuristic approach was developed. For state estimation, two approaches were considered: multi-microgrid state estimation and fuzzy state estimation. Regarding frequency control, the hierarchical structure of the multi-microgrid is exploited to deal with the transition to islanded operation and load following in islanded operation. All these tools have proved to be efficient in managing the multi-microgrid system in normal interconnected mode and, in case of the frequency control, in islanded operation.

This work proposes an innovative method based on autoencoders to perform state estimation in distribution grids, which has as main advantage the fact of being independent of the network parameters and topology. The method was tested in a real low voltage grid (incorporating smart grid features), under different scenarios of smart meter deployment. Simulations were performed in order to understand the necessary requirements for an accurate distribution grid state estimator and to evaluate the performance of a state estimator based on autoencoders.