Multi-Step Solar Irradiance Forecasting and Domain Adaptation of Deep Neural Networks

Harnessing the power of hybrid deep learning algorithm for the estimation of global horizontal irradiance

Accurately and precisely estimating global horizontal irradiance (GHI) poses significant challenges due to the unpredictable nature of climate parameters and geographical limitations. To address this challenge, this study proposes a forecasting framework using an integrated model of the convolutional neural network (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU). The proposed model uses a dataset of four different districts in Rajasthan, each with unique solar irradiance patterns. Firstly, the data was preprocessed and then trained with the optimized parameters of the standalone and hybrid models and compared. It can be observed that the proposed hybrid model (CNN-LSTM-GRU) consistently outperformed all other models regarding Mean absolute error (MAE) and Root mean squared error (RMSE). The experimental results demonstrate that the proposed method forecasts accurate GHI with a RMSE of 0.00731, 0.00730, 0.00775, 0.00810 and MAE of 0.00516, 0.00524, 0.00552, 0.00592 for Barmer, Jaisalmer, Jodhpur and Bikaner respectively. This indicates that the model is better at minimizing prediction errors and providing more accurate GHI estimates. Additionally, the proposed model achieved a higher coefficient of determination (R (Ghimire et al., 2019)), suggesting that it best fits the dataset. A higher R2 value signifies that the proposed model could explain a significant portion of the variance in the GHI dataset, further emphasizing its predictive capabilities. In conclusion, this work demonstrates the effectiveness of the hybrid algorithm in improving adaptability and enhancing prediction accuracy for GHI estimation.

Attention-Based Models for Multivariate Time Series Forecasting: Multi-step Solar Irradiation Prediction

Bangladesh's subtropical climate with an abundance of sunlight throughout the greater portion of the year results in increased effectiveness of solar panels. Solar irradiance forecasting is an essential aspect of grid-connected photovoltaic systems to efficiently manage solar power's variation and uncertainty and to assist in balancing power supply and demand. This is why it is essential to forecast solar irradiation accurately. Many meteorological factors influence solar irradiation, which has a high degree of fluctuation and uncertainty. Predicting solar irradiance multiple steps ahead makes it difficult for forecasting models to capture long-term sequential relationships. Attention-based models are widely used in the field of Natural Language Processing for their ability to learn long-term dependencies within sequential data. In this paper, our aim is to present an attention-based model framework for multivariate time series forecasting. Using data from two different locations in Bangladesh with a resolution of 30 min, the Attention-based encoder-decoder, Transformer, and Temporal Fusion Transformer (TFT) models are trained and tested to predict over 24 steps ahead and compared with other forecasting models. According to our findings, adding the attention mechanism significantly increased prediction accuracy and TFT has shown to be more precise than the rest of the algorithms in terms of accuracy and robustness. The obtained mean square error (MSE), the mean absolute error (MAE), and the coefficient of determination (R2) values for TFT are 0.151, 0.212, and 0.815, respectively. In comparison to the benchmark and sequential models (including the Naive, MLP, and Encoder-Decoder models), TFT has a reduction in the MSE and MAE of 8.4-47.9% and 6.1-22.3%, respectively, while R2 is raised by 2.13-26.16%. The ability to incorporate long-distance dependency increases the predictive power of attention models.

Causal Factor Disentanglement for Few-Shot Domain Adaptation in Video Prediction

An important challenge in machine learning is performing with accuracy when few training samples are available from the target distribution. If a large number of training samples from a related distribution are available, transfer learning can be used to improve the performance. This paper investigates how to do transfer learning more effectively if the source and target distributions are related through a Sparse Mechanism Shift for the application of next-frame prediction. We create Sparse Mechanism Shift-TempoRal Intervened Sequences (SMS-TRIS), a benchmark to evaluate transfer learning for next-frame prediction derived from the TRIS datasets. We then propose to exploit the Sparse Mechanism Shift property of the distribution shift by disentangling the model parameters with regard to the true causal mechanisms underlying the data. We use the Causal Identifiability from TempoRal Intervened Sequences (CITRIS) model to achieve this disentanglement via causal representation learning. We show that encouraging disentanglement with the CITRIS extensions can improve performance, but their effectiveness varies depending on the dataset and backbone used. We find that it is effective only when encouraging disentanglement actually succeeds in increasing disentanglement. We also show that an alternative method designed for domain adaptation does not help, indicating the challenging nature of the SMS-TRIS benchmark.

A systematic review on predicting PV system parameters using machine learning

Due to the growing demand, assessing performance has become obligatory for photovoltaic (PV) energy harvesting systems. Performance assessment involves estimating different PV system parameters. Traditional ways, such as calculating solar radiation using satellite data and the IV characteristics approach as assessment methods, are no longer reliable enough to provide a reasonable projection of PV system parameters. Estimating system parameters using machine learning (ML) approaches has become a reliable and popular method because of its speed and accuracy. This paper systematically reviewed ML-based PV parameter estimation studies published in the last three years (2020 - 2022). Studies were analyzed using several criteria, including ML algorithm, outcome, experimental setup, sample data size, and error metric. The analysis revealed several interesting factors. The neural network was the most popular ML method (32.55%), followed by random vector functional link (13.95%) and support vector machine (9.30%). Dataset was sourced from hardware tests and computer-based simulations: 66% of the studies used data from only computer simulation, 18% used data from only hardware setup, and the 16% experiments used data from both hardware and simulations to evaluate different system parameters. The top three most commonly used error metrics were root mean square error (29.1%), mean absolute error (17.5%), and coefficient of determination (15.9%). Our systematic review will help researchers assess ML algorithms' projection in PV system parameters estimation. Consequently, scopes shall be created to establish more robust governmental frameworks, expand private financing in the PV industry, and optimize PV system parameters.

Solar Irradiance Forecasting Using a Data-Driven Algorithm and Contextual Optimisation

Solar forecasting plays a key part in the renewable energy transition. Major challenges, related to load balancing and grid stability, emerge when a high percentage of energy is provided by renewables. These can be tackled by new energy management strategies guided by power forecasts. This paper presents a data-driven and contextual optimisation forecasting (DCF) algorithm for solar irradiance that was comprehensively validated using short- and long-term predictions, in three US cities: Denver, Boston, and Seattle. Moreover, step-by-step implementation guidelines to follow and reproduce the results were proposed. Initially, a comparative study of two machine learning (ML) algorithms, the support vector machine (SVM) and Facebook Prophet (FBP) for solar prediction was conducted. The short-term SVM outperformed the FBP model for the 1- and 2- hour prediction, achieving a coefficient of determination (R2) of 91.2% in Boston. However, FBP displayed sustained performance for increasing the forecast horizon and yielded better results for 3-hour and long-term forecasts. The algorithms were optimised by further contextual model adjustments which resulted in substantially improved performance. Thus, DCF utilised SVM for short-term and FBP for long-term predictions and optimised their performance using contextual information. DCF achieved consistent performance for the three cities and for long- and short-term predictions, with an average R2 of 85%.

Voltage Regulation For Residential Prosumers Using a Set of Scalable Power Storage

Among the electrical problems observed from the solar irradiation variability, the electrical energy quality and the energetic dispatch guarantee stand out. The great revolution in batteries technologies has fostered its usage with the installation of photovoltaic system (PVS). This work presents a proposition for voltage regulation for residential prosumers using a set of scalable power batteries in passive mode, operating as a consumer device. The mitigation strategy makes decisions acting directly on the demand, for a storage bank, and the power of the storage element is selected in consequence of the results obtained from the power flow calculation step combined with the prediction of the solar radiation calculated by a recurrent neural network Long Short-Term Memory (LSTM) type. The results from the solar radiation predictions are used as subsidies to estimate, the state of the power grid, solving the power flow and evidencing the values of the electrical voltages 1-min enabling the entry of the storage device. In this stage, the OpenDSS (Open distribution system simulator) software is used, to perform the complete modeling of the power grid where the study will be developed, as well as simulating the effect of the overvoltages mitigation system. The clear sky day stored 9111 Wh/day of electricity to mitigate overvoltages at the supply point; when compared to other days, the clear sky day needed to store less electricity. On days of high variability, the energy stored to regulate overvoltages was 84% more compared to a clear day. In order to maintain a constant state of charge (SoC), it is necessary that the capacity of the battery bank be increased to meet the condition of maximum accumulated energy. Regarding the total loading of the storage system, the days of low variability consumed approximately 12% of the available capacity of the battery, considering the SoC of 70% of the capacity of each power level.