Lauri Kutt

TL;DR: This research uses machine learning and deep learning-based Recurrent Neural Networks (RNN) algorithms for the day-ahead load forecasting of an Estonian household and results indicate that the RNN based algorithm gives better forecasting based on lower Root Mean Square Error (RMSE) value.

TL;DR: In this paper, the authors presented a case example of simulations of the total harmonic load currents when the electric vehicle charging is added to the residential area distribution network workday harmonic current load profiles.

Abstract: Electric vehicle chargers are mostly power converters with power factor correction and waveform shaping features included. Different vehicle makes and models will likely have different charger control and topology. While the charger charging currents have generally low AC waveform distortions, including variety of chargers at the same time can provide even further lower harmonic currents. This is due to harmonic cancellation brought by different harmonic patterns of different chargers. The present paper analyzes the sine distortions and the sum of the harmonic currents of multiple harmonic sources of such type. The inputs for the analysis are the measurements of the commercial EV's harmonic charging current patterns. It will be shown that remarkable harmonic cancellation will be likely if different vehicle types are included in the pool of vehicles charging at the same time. The results presented here for the EV chargers would be valid also for other similar mains interface converters.

TL;DR: In this article, an experimental investigation of two PD defects/sources at different locations on a medium voltage (MV) cable section is presented, where a high frequency current transformer is used for single end PD measurements.

TL;DR: In this paper, the effects of the AC supply voltage variations to the load current harmonic patterns and summation of harmonic currents when the chargers are used at different AC voltage levels are discussed.