Explore the vast range of titles available.

Predicting the time series energy consumption data of manufacturing processes can optimize energy management efficiency and reduce maintenance costs for enterprises. Using deep learning algorithms to establish prediction models for sensor data is an effective approach; however, the performance of these models is significantly influenced by the quantity and quality of the training data. In real production environments, the amount of time series data that can be collected during the manufacturing process is limited, which can lead to a decline in model performance. In this paper, we use an improved TimeGAN model for the augmentation of energy consumption data, which incorporates a multi-head self-attention mechanism layer into the recovery model to enhance prediction accuracy. A hybrid CNN-GRU model is used to predict the energy consumption data from the operational processes of manufacturing equipment. After data augmentation, the prediction model exhibits significant reductions in RMSE and MAE along with an increase in the R2 value. The prediction accuracy of the model is maximized when the amount of generated synthetic data is approximately twice that of the original data.

The rapid development of photovoltaic (PV) energy and its growing penetration in power systems have made accurate and robust PV power forecasting a critical challenge. However, the strong stochasticity, nonlinearity, and heterogeneity of PV output, driven by complex environmental conditions, hinder the performance of conventiona l forecasting methods. To address these issues, this study proposes a novel hybrid framework that integrates TimeGAN-based data augmentation, extended LSTM (xLSTM), and Transformer networks for probabilistic and accurate PV power prediction. First, TimeGAN is employed to synthesize realistic PV time series data, effectively capturing temporal correlations while preserving the irradiance–temperature dependency, thus mitigating the limitations of scarce or imbalanced historical datasets. Second, a hybrid xLSTM-Transformer architecture is developed, where the matrix memory-enhanced xLSTM module focuses on local feature extraction, and the Transformer module models long-range dependencies via self-attention mechanisms. Finally, the proposed model is validated using real-world operation data from the State Grid of China. Experimental results demonstrate that the proposed framework significantly improves prediction accuracy under realistic operating conditions. Compared with conventional LSTM and Transformer baselines, the proposed TimeGAN–xLSTM–Transformer model achieves a reduction of approximately 48.1% in RMSE and 44.1% in MAE, highlighting its superior capability in capturing both short-term fluctuations and long-term temporal dependencies of photovoltaic power generation. This research contributes to advancing intelligent PV forecasting technologies by leveraging data generation, temporal modeling, and attention mechanisms, offering theoretical and practical support for the reliable integration of PV in renewable-dominated power systems.

To address the challenges of scarce unmanned aerial vehicle (UAV) track samples, severe class imbalance, and high motion similarity between UAVs and birds in low-altitude radar recognition, this paper proposes a trajectory classification method integrating a TCN-Transformer model with a physics-constrained TimeGAN (PC-TimeGAN) data augmentation framework. Specifically, the PC-TimeGAN generates high-quality, kinematically compliant UAV trajectories to alleviate data scarcity and class imbalance. A multi-scale TCN-Transformer is then constructed to comprehensively extract features, utilizing multi-kernel dilated convolutions for local temporal correlations and self-attention mechanisms for global temporal dependencies, thereby improving the discrimination between UAV and bird trajectories with similar motion patterns. Furthermore, a joint loss function combining Focal Loss and Triplet Loss is employed to optimize the decision boundaries and feature space, enhancing model robustness and generalization. Experiments on a measured dataset demonstrate that, under the 15-dimensional input setting, the proposed method achieves a UAV recall of 80.00%, an FAR of 3.15%, a precision of 64.00%, and an F1-score of 0.7111. Compared to baseline methods (e.g., SVM, LSTM, GRU, Transformer, and 1D-CNN), the proposed approach significantly improves UAV recall under limited trajectory information while keeping the false-alarm rate of misclassifying birds as UAVs low. Ultimately, this method markedly enhances the comprehensive performance of rapid track-level target classification for low-altitude surveillance radars.

Enhancing crop water productivity is crucial for regional water resource management and agricultural sustainability, particularly in arid regions. However, evaluating the spatial heterogeneity and temporal dynamics of crop water productivity in face of data limitations poses a challenge. In this study, we propose a framework that integrates remote sensing data, time series generative adversarial network (TimeGAN), dynamic Bayesian network (DBN), and optimization model to assess crop water productivity and optimize crop planting structure under limited water resources allocation in the Qira oasis. The results demonstrate that the combination of TimeGAN and DBN better improves the accuracy of the model for the dynamic prediction, particularly for short-term predictions with 4 years as the optimal timescale (R 2  > 0.8). Based on the spatial distribution of crop suitability analysis, wheat and corn are most suitable for cultivation in the central and eastern parts of Qira oasis while cotton is unsuitable for planting in the western region. The walnuts and Chinese dates are mainly unsuitable in the southeastern part of the oasis. Maximizing crop water productivity while ensuring food security has led to increased acreage for cotton, Chinese dates and walnuts. Under the combined action of the five optimization objectives, the average increase of crop water productivity is 14.97%, and the average increase of ecological benefit is 3.61%, which is much higher than the growth rate of irrigation water consumption of cultivated land. It will produce a planting structure that relatively reduced irrigation water requirement of cultivated land and improved crop water productivity. This proposed framework can serve as an effective reference tool for decision-makers when determining future cropping plans.

Generating air-conditioning system load data is crucial for tasks such as power grid scheduling and intelligent energy management. Air-conditioning load data exhibit strong non-stationarity. Their load curves are influenced by seasonal variations and highly correlated with outdoor meteorological conditions, indoor activity patterns, and equipment operational strategies. These characteristics lead to pronounced periodicity, sudden shifts, and diverse data patterns. Existing load generation models tend to produce averaged distributions, which often leads to the loss of specific temporal patterns inherent in air-conditioning loads. Moreover, as purely data-driven models, they lack explicit physical constraints, resulting in generated data with limited physical interpretability. To address these issues, this paper proposes a hybrid generation framework that integrates the DTW clustering algorithm, a physically-constrained TimeGAN model, and an LSTM-based model selection mechanism. Specifically, DTW clustering is first employed to achieve structured data partitioning, thereby enhancing the model’s ability to recognize and model diverse temporal patterns. Subsequently, to overcome the dependency on detailed building parameters and extensive sensor networks, a parameter-free physical constraint mechanism based on intrinsic temperature-load correlations is incorporated into the TimeGAN supervised loss. This design ensures thermodynamic consistency even in sensor-scarce environments where only basic operational data is available. Finally, to address adaptability challenges in long-term sequence generation, an LSTM-based selection mechanism is designed to evaluate and select from clustered submodels dynamically. This approach facilitates adaptive temporal fusion within the generation strategy. Extensive experiments on air-conditioning load datasets from Southeast China demonstrate that the framework achieves a local similarity score of 0.98, outperforming the state-of-the-art model and the original TimeGAN by 11.4% and 13.3%, respectively.

This paper presents a novel framework that integrates Conditional Generative Adversarial Networks (CGANs) and TimeGAN to generate synthetic Building-Integrated Photovoltaic (BIPV) power data, addressing the challenge of data scarcity in this domain. By incorporating time-related attributes as conditioning information, our method ensures the preservation of chronological order and enhances data fidelity. A tailored learning scheme is implemented to capture the unique characteristics of solar power generation, particularly during sunrise and sunset. Comprehensive evaluations demonstrate the framework’s effectiveness in generating high-quality synthetic data, evidenced by a 79.58% improvement in the discriminative score and a 13.46% improvement in the predictive score compared to TimeGAN. Moreover, integrating the synthetic data into forecasting models resulted in up to 23.56% improvement in mean absolute error (MAE) for BIPV power generation predictions. These results highlight the potential of our framework to enhance prediction accuracy and optimize data utilization in renewable energy applications.

Accurate forecasting of mining-induced surface deformation is critical for coal-mine safety assessment and hazard mitigation. InSAR deformation time series are often short, temporally sparse, and strongly nonlinear. These characteristics can make purely data-driven predictors unreliable in small-sample settings. To address this issue, we propose a generation–prediction–interpretation framework that combines generative augmentation with physics-informed forecasting. We first develop a TCN-TimeGAN model to synthesize high-fidelity deformation sequences and expand the training set. Recurrent modules in the generator and discriminator are replaced with causal TCN residual blocks, and a temporal self-attention layer is further stacked on top of the TCN backbone to adaptively reweight informative time steps. We then construct a physics-informed Kolmogorov–Arnold Network, termed PI-KAN. Subsidence-consistency and smoothness priors are embedded in the learning objective to promote physically plausible predictions while retaining spline-based interpretability. Experiments on SBAS-InSAR deformation series from the Guqiao coal mine show that the framework achieves an RMSE of 0.825 mm and an R2 of 0.968. It outperforms TGAN-KAN, CNN-BiGRU, and BiGRU under the same evaluation protocol. Visualizations of the learned spline-based edge functions further reveal stronger nonlinear responses for lagged inputs closer to the forecast horizon, providing interpretable evidence of short-term temporal sensitivity under sparse observations.

Energy consumption data is being used for improving the energy efficiency and minimizing the cost. However, obtaining energy consumption data has two major challenges: (i) data collection is very expensive, time-consuming, and (ii) security and privacy concern of the users which can be revealed from the actual data. In this research, we have addressed these challenges by using generative adversarial networks for generating energy consumption profile. We have successfully generated synthetic data which is similar to the real energy consumption data. On the basis of the recent research conducted on TimeGAN, we have implemented a framework for synthetic energy consumption data generation that could be useful in research, data analysis and create business solutions. The framework is implemented using the real-world energy dataset, consisting of energy consumption data of the year 2020 for the Australian states of Victoria, New South Wales, South Australia, Queensland and Tasmania. The results of implementation is evaluated using various performance measures and the results are showcased using visualizations along with Principal Component Analysis (PCA) and t-distributed stochastic neighbor embedding (TSNE) plots. Overall, experimental results show that Synthetic data generated using the proposed implementation possess very similar characteristics to the real dataset with high comparison accuracy.

In surface deformation monitoring for mining areas, interferometric synthetic aperture radar (InSAR) technology has become a popular research topic due to its efficiency and high accuracy. However, transforming temporal monitoring data into surface deformation predictions remains challenging. In practical applications, InSAR data often face limitations like low acquisition frequency and insufficient data volume, leading to prediction models being prone to overfitting and having poor accuracy. Therefore, this paper proposes an improved temporal convolutional network (TCN) time-series generative adversarial network (GAN) with an attention mechanism, called the Attention–TCN–TimeGAN, to enhance InSAR surface deformation data for better prediction results. By combining the embedding, recovery, generator, and discriminator networks, we used the TCN to expand the receptive field and capture long-term temporal features. Additionally, we integrated the self-attention mechanism into the generator and discriminator to adapt to random vectors, achieving better data generation results. The loss function uses the Wasserstein distance to measure the original data distribution and adds a gradient penalty term with adaptive weights to achieve effective feature extraction from time-series data. Experimental results show that the data generated by our model more comprehensively cover the original data distribution. The prediction results at four test points showed the lowest mean absolute error and mean-squared error and the highest coefficient of determination (R2). These results demonstrate the effectiveness of our generative model in predicting small-sample InSAR time-series data, providing a new method for surface deformation monitoring.

Flood forecasting is integral to water resources management and flood prevention. Recently, deep learning has made substantial progress in flood forecasting. However, neural network-based models require extensive datasets to ensure reliable precision. The scarcity of sufficient observed time-series data on flood events poses a major challenge to precise flood forecasting. Therefore, a modified Time-series Generative Adversarial Network (TimeGAN) is proposed to data enhancement and handle flood forecasting in this study. The proposed model uses the Transformer and Wasserstein distance loss function in the sequence generator, termed TW-TimeGAN, avoids gradient vanishing and exploding explosion, improving the reliability of model for long-term forecast. Meanwhile, the integrate of feature-temporal dual-attention with Recurrent Neural Network (RNN) in the recovery function enhances the capacity of TW-TimeGAN in extracting features. This paper utilized observed data and synthetic data to construct a flood prediction model by employing Long Short Term Memory (LSTM) and Bidirectional LSTM (BiLSTM). Through a comparative analysis of results, TW-TimeGAN achieves the lowest Dynamic Time Warping (DTW) (0.2332), demonstrating that TW-TimeGAN could effectively enhance the learning of precipitation and streamflow features, enabling the generation of high-quality synthetic flood sequences that closely resemble real flood events. Combined with flood prediction models, especially BiLSTM, the average Root Mean Square Error (RMSE) (7.67) and average Mean Absolute Error (MAE) (3.88) of the prediction results are the smallest, and the overall Nash-Sutcliffe Efficiency (NSE) (0.903) is the largest. It can be concluded that TW-TimeGAN-BiLSTM has the best prediction performance and demonstrates greater applicability in flood prediction.