Towards energy efficiency: A comprehensive review of deep learning-based photovoltaic power forecasting strategies

Abstract

Time series forecasting still awaits a transformative breakthrough like that happened in computer vision and natural language processing. The absence of extensive, domain-independent benchmark datasets and standardized performance measurement units poses a significant challenge for it, especially for photovoltaic forecasting applications. Additionally, since it is often time domain-driven, a plethora of highly unique and domain-specific datasets were produced. The lack of uniformity among published models, developed under diverse settings for varying forecasting horizons, and assessed using non-standardized metrics, remains a significant obstacle to the progress of the field as a whole. To address these issues, a systematic review of the state-of-the-art literature on prediction tasks is presented, collected from the Web of Science and Scopus databases, published in 2022 and 2023, and filtered using keywords such as "photovoltaic," "deep learning," "forecasting," and "time series." Finally, 36 case studies were selected. Before comparing, a state-of-the-art demonstration of key elements in the topic was presented, such as model type, hyperparameters, and evaluation metrics. Then, the 36 articles were compared in terms of statistical analysis, including top publishing countries, data sources, variables, input, and output horizon, followed by an overall model comparison demonstrating every proposed model categorized into model type (artificial neural network units, recurrent units, convolutional units, and transformer units). Due to the mostly utilization of specific private datasets measured at the targeted location, having universal error metrics is crucial for clear global benchmarking. Root Mean Squared Error and Mean Absolute Error were the most utilized metrics, although they specifically demonstrate the accuracy relative to their respective sites. However, 33% utilized universal metrics, such as Mean Absolute Percentage Error, Normalized Root Mean Squared Error, and the Coefficient of Determination. Finally, trends, challenges, and future research were highlighted for the relevant topic to spotlight and bypass the current challenges.

Keywords: Convolutional neural network; Deep learning; Long-short-term memory; Photovoltaic power forecasting; Time series; Transformer.

Figure 1

Trend of Worldwide total installed PV panels capacity. (source: IEA PVPS 2023).

Figure 2

Techniques adapted in the publications of PV power forecasting over time.

Figure 3

Network connection of related publication in 2023.

Figure 4

Systematic review methodology.

Figure 5

Data representation.

Figure 6

Connected neural network with two inputs and one bias.

Figure 7

Schematic diagram of Recurrent cells.

Figure 8

Convolutional schematic.

Figure 9

Attention mechanism schematic.

Figure 10

Transformer architecture.

Figure 11

Feed forward neural network and gradient backpropagation schematics.

Figure 12

PV forecasting models recent published articles (2022 and 2023).

Figure 13

The distribution of the selected 36 case studies around the world.

Figure 14

Sources of datasets utilized in selected case studies.

Figure 15

Variables types of recent published papers, (count).

Figure 16

Number of variables utilized in recent published papers.

Figure 17

Data resolutions of recent published papers.

Figure 18

Input sequences length of recent published papers.

Figure 19

Output types of recent published papers.

Figure 20

Forecasting horizons of recent published papers.