Neural networks for online learning of non-stationary data streams: a review and application for smart grids flexibility
- PDF / 3,158,467 Bytes
- 44 Pages / 439.37 x 666.142 pts Page_size
- 13 Downloads / 134 Views
Neural networks for online learning of non‑stationary data streams: a review and application for smart grids flexibility improvement Zeineb Hammami1 · Moamar Sayed‑Mouchaweh2 · Wiem Mouelhi1 · Lamjed Ben Said1
© Springer Nature B.V. 2020
Abstract Learning efficient predictive models in dynamic environments requires taking into account the continuous changing nature of phenomena generating the data streams, known in machine learning as “concept drift”. Such changes may affect models’ effectiveness over time, requiring permanent updates of parameters and structure to maintain performance. Several supervised machine learning methods have been developed to be adapted to learn in dynamic and non-stationary environments. One of the most well-known and efficient learning methods is neural networks. This paper focuses on the different neural networks developed to build learning models able to adapt to concept drifts on streaming data. Their performance will be studied and compared using meaningful criteria. Their limits to address the challenges related to the problem of the improvement of electrical grid flexibility in presence of distributed Wind–PV renewable energy resources within the context of energy transition will be highlighted. Finally, the study provides a self-adaptive scheme based on the use of neural networks to overcome these limitations and tackle these challenges. Keywords Concept drift · Adaptive neural networks · Dynamic and non-stationary environments · Electrical grid flexibility · Renewable energy resources
* Zeineb Hammami [email protected] Moamar Sayed‑Mouchaweh moamar.sayed‑mouchaweh@imt‑lille‑douai.fr Wiem Mouelhi [email protected] Lamjed Ben Said [email protected] 1
SMART Lab, Computer Science Department, University of Tunis, ISG, Tunis, Tunisia
2
IMT Lille Douai, 59500 Douai, France
13
Vol.:(0123456789)
Z. Hammami et al.
1 Introduction The dynamic nature and time-varying behavior of actual environments create serious challenges for learning models. Thus, changes may deteriorate the accuracy of model predictions, which requires permanent adaptation strategies. Changes can take several forms, the most common being the shift of the underlying function generating the data, a phenomenon known in machine learning as “concept drift” (Lu et al. 2014). Conceptual drifts can occur on stream data across many application areas. Among these applications, the energy sector can be referred to since, due to digitalization, huge amounts of data are generated in stream fashion on energy consumption, transmission, distribution, and production. Learning models built on the use of the data generated enable to optimize electricity generation and consumption and improve the stability and resilience of the grid. Here, the accuracy of model forecasting can be significantly affected by the evolutionary and non-stationary behavior of power production and consumption, due to changes in weather conditions and consumers profiles. In this paper, we focus on the application that improves the
Data Loading...
