Abstract:Photovoltaic (PV) power interval prediction can provide a variation range of prediction results, which is of great significance to promoting the optimization of power grid dispatching and maintaining the stability of the power system. However, the existing PV interval prediction fails to accurately reflect error fluctuations in their construction process due to dependency on data distribution assumptions and unreasonable estimation of prediction error standard deviations (PEStds). To address these issues, we propose an adaptive interval prediction method based on multi-objective optimization. Firstly, K-means clustering is used to segment similar days. Subsequently, Convolutional Neural Network-Gated Recurrent Unit-Attention (CNN-GRU-Attention) is utilized for point prediction. Furthermore, Grey Relation Analysis (GRA) is used to screen dates similar to the target date under the same weather type, enabling a more precise estimation of prediction error fluctuations. Ultimately, the estimated values are multiplied by non-fixed parameters to construct intervals, which avoids the dependency on data distribution assumptions. Among them, the optimal values for non-fixed parameters are obtained through multi-objective optimization considering interval coverage probability, interval width, and deviation. To verify the effectiveness of the proposed model, we conduct comparative experiments on two datasets with different resolutions. The results demonstrate that the proposed model offers more flexible and higher-quality intervals. Not only does this research improve the accuracy of PV power generation interval prediction, but it also helps to promote the development of smart grid technology and improve the adaptive ability of power systems facing dynamic environments and complex data, which has an important impact on the future energy management and power market.

Photovoltaic Power Generation Prediction Using Data Clustering and Parameter Optimization

Photovoltaic Cluster Forecasting Based on Spatial Correlation-Informed Deep Learning

Nonparametric Probabilistic Forecasting of Regional Photovoltaic Power Based on Spatial Clustering and Combining Quantile Regression

Weather Condition Clustering for Improvement of Photovoltaic Power Plant Generation Forecasting Accuracy

Short-term power prediction for photovoltaic power plants using a hybrid improved Kmeans-GRA-Elman model based on multivariate meteorological factors and historical power datasets

Photovoltaic Solar Power Prediction Using iPSO-Based Data Clustering and AdaLSTM Network

Prediction of photovoltaic power output based on similar day analysis using RBF neural network with adaptive black widow optimization algorithm and K-means clustering

Forecasting a Short-Term Photovoltaic Power Model Based on Improved Snake Optimization, Convolutional Neural Network, and Bidirectional Long Short-Term Memory Network

Clustering and LSTM-based Robust Day-ahead Hourly Forecasting of Photovoltaic Power

An adaptive photovoltaic power interval prediction based on multi-objective optimization

Research on regional photovoltaic (PV) power output forecasting method based on weather pattern clustering

A Combined Method of Two-model Based on Forecasting Meteorological Data for Photovoltaic Power Generation Forecasting

Enhanced solar photovoltaic power prediction using diverse machine learning algorithms with hyperparameter optimization

Photovoltaic data clustering method, photovoltaic power prediction method, storage medium and device

Forecasting Method of Photovoltaic Output Power Based on k-Means/KPCA/PSO-BP

Research on Probability Distribution of Short-Term Photovoltaic Output Forecast Error Based on Numerical Characteristic Clustering

Forecasting and uncertainty analysis of day-ahead photovoltaic power using a novel forecasting method

A Novel Photovoltaic Power Output Forecasting Method Based on Weather Type Clustering and Wavelet Support Vector Machines Regression

Interval Prediction of Photovoltaic Power Using Improved NARX Network and Density Peak Clustering Based on Kernel Mahalanobis Distance

Photovoltaic power generation prediction based on k-means clustering analysis and GRO-CNN-LSTM Attention

Short-term photovoltaic power forecasting using parameter-optimized variational mode decomposition and attention-based neural network