Aug 2026· International journal of pattern recognition and artificial intelligence· 0 citations
TL;DR
Experimental results on various publicly available PV power datasets show that BiMS-DGCN consistently outperforms several state-of-the-art forecasting methods, thereby improving the accuracy of multi-site photovoltaic power forecasting.
Abstract
Accurate multi-site photovoltaic (PV) power forecasting plays a pivotal role in enhancing smart grid operation and facilitating large-scale renewable energy integration. However, existing forecasting approaches often struggle to simultaneously capture the dynamic spatial interactions among geographically distributed PV plants and the diverse temporal dependencies embedded in PV power generation data. To overcome these limitations, this paper proposes a Bidirectional Multi-Scale Dynamic Graph Convolutional Network (BiMSDGCN). The proposed framework integrates two key components. First, a Dynamic Graph Convolutional Network (DGCN) is designed to adaptively learn a time-dependent adjacency matrix to capture evolving spatial correlations among PV sites. Second, a Bidirectional Multi-Scale Attention (BiMS) is developed to model both cross-scale and intra-scale dependencies, enabling more comprehensive spatiotemporal feature extraction and information fusion. Experimental results on various publicly available PV power datasets show that BiMS-DGCN consistently outperforms several state-of-the-art forecasting methods. Ablation studies further verify that both DGCN and BiMS play critical roles in modeling complex spatiotemporal dependencies, thereby improving the accuracy of multi-site photovoltaic power forecasting.
Accurately predicting the power generation of distributed photovoltaic (PV) power plants is extremely challenging due to their heterogeneity and complex temporal variations. To address these issues, this paper proposes a clustering-based prediction framework that combines K-means clustering with our proposed model. First, a multidimensional feature representation is constructed to characterize the statistical properties, dynamic fluctuations, and meteorological correlations of PV power plants. This feature representation is then used to classify the power plants into homogeneous clusters. For each cluster, a hybrid deep learning model is developed to capture local temporal patterns and long-term dependencies, with a focus on key time steps. Finally, the outputs of all clusters are aggregated to obtain the regional power generation prediction results. Experimental results demonstrate that the proposed method outperforms the baseline model on multiple error metrics, proving its effectiveness in predicting the power generation of multiple PV power plants.
Jian Tang, Lu-Lu Lin· European Conference on Elect...· 0 citations
A hybrid spatiotemporal prediction framework integrating an improved Long Short-Term Memory (LSTM) network with dynamic graph embedding for deep feature mining and coordinated forecasting and offers methodological support for intelligent electromagnetic energy management and resilient smart power systems is proposed.
S. Wan, J. Tan, Teng Luo et al.· Advanced Electromagnetics· 0 citations
Experimental results show that MSCNN-ResLSTM achieves higher forecasting accuracy and greater stability across the full 24-step prediction horizon, consistently outperforming all competing baselines while effectively suppressing recursive error propagation.
Yu-Hang Zhang, Yi-Ting Zhao, Yu-Jing Meng et al.· De Computis· 0 citations
Existing studies still have limitations in characterizing the complex temporal patterns of photovoltaic time series. In particular, current forecasting models often struggle to capture local abrupt fluctuations and nonlinear relationships among variables, while hyperparameter optimization remains challenging. To overcome these limitations, this paper proposes a CNN-iTransformer short-term photovoltaic power forecasting method based on multiscale decomposition and adaptive optimization. First, the original photovoltaic power time series is decomposed into trend, seasonal, and residual components using seasonal-trend decomposition based on Loess (STL). The residual component is then further decomposed through variational mode decomposition (VMD) to fully extract the latent multiscale temporal information embedded in the sequence. Second, a convolutional neural network (CNN) is employed to extract local fluctuation features, while the iTransformer is utilized to model the nonlinear relationships and long-term temporal dependencies between multiple meteorological variables and photovoltaic power. Finally, the Phototropic Growth Algorithm (PGA) is introduced to optimize key hyperparameters of the forecasting model, thereby improving its generalization capability and forecasting accuracy under multi-seasonal scenarios. Simulation experiments are conducted using real-world data from a photovoltaic power station in Alice Springs, Australia. The experimental results show that, under the same PGA optimization conditions, the proposed model reduces the root mean square error (RMSE) and mean absolute error (MAE) by approximately 19.12% and 17.43%, respectively, compared with the baseline iTransformer. Meanwhile, its performance is better than that of several mainstream deep learning models, which verifies the generalization ability and forecasting accuracy of the proposed method under complex seasonal variations.
Accurate short-term photovoltaic (PV) power forecasting is essential for secure grid operation, economic dispatch, and efficient utilization of renewable energy. However, PV output exhibits strong nonlinearity and nonstationarity due to rapidly varying meteorological conditions, cloud movements, and system uncertainties. Conventional statistical models and single-source deep learning approaches often fail to fully exploit the rich multi-source information available in modern PV plants, such as historical power, on-site meteorological measurements, and clear-sky or numerical weather prediction (NWP) features. In this paper, we propose MSF-TransPV, a Multi-Source Fusion Transformer framework for short-term PV power forecasting. The model adopts a multi-branch temporal encoder that separately processes historical PV output, meteorological variables, and optional clear-sky/NWP-derived features, mapping them into a shared latent space. A cross-variable multi-head attention module is then introduced to explicitly capture the dependencies between PV dynamics and atmospheric conditions, enabling fine-grained interaction across different feature sources. On top of the fused representation, a temporal Transformer encoder models long-range temporal dependencies and feeds a lightweight decoder that supports both point forecasting and probabilistic forecasting via quantile regression. Experimental results on real-world PV datasets demonstrate that MSF-TransPV consistently outperforms persistence, statistical baselines, recurrent neural networks, and vanilla Transformer models in terms of RMSE, MAE, and normalized error metrics, while also providing reliable prediction intervals. These results indicate that explicit multi-source fusion and cross-variable attention are highly effective for PV forecasting under complex and rapidly changing weather conditions.
Xiao-Mei Wang, Pei-Xuan Xu, Xiao-Hui Wang· European Conference on Elect...· 0 citations
Accurate prediction of electric vehicle (EV) charging demand is critical for planning charging infrastructure and allocating resources efficiently. However, existing methods often fail to capture multi-scale spatial dependencies and struggle to model both long-range temporal dependencies and short-term fluctuations. To address these limitations, a deep learning framework, STGFormer, is developed for citywide EV charging demand forecasting. First, a temporal dilated convolution module (TDConv) is proposed to extract multi-scale local temporal patterns. Second, an adaptive spatial dilated graph attention module (ADGAT) is proposed to mine multi-hop spatial correlations between geographically adjacent and functionally similar regions. Third, a hybrid xLSTM-Transformer encoder captures global temporal dependencies while preserving local continuity. The performance of STGFormer is evaluated on a real-world dataset from Shenzhen. Extensive experiments demonstrate that STGFormer consistently outperforms fifteen representative baseline models. It achieves average improvements of 8.58% in RMSE, 13.95% in RAE, and 14.64% in MAE.
Ye Li, Zhuoran Ye, Bo Li et al.· Electronics· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
