An Improved TCN Considering Data Augmentation in Enabling Load Classification

Accurate load forecasting is essential for ensuring safe, stable, and economical operation of energy internet. Temporal convolutional networks (TCNs) have demonstrated superior performance, when compared to recurrent neural network models, since their introduction in electrical load forecasting. However, the current TCN‐based models are unable to obtain a large receptive field and strong long‐time feature extraction capability owing to the specific kernel size of the 1D convolution structure. This paper proposes a temporal inception convolutional network based on multi‐head attention (TICN‐Att) for ultra‐short‐term load prediction. By introducing an inception structure into the TCN, the proposed model can extract multi‐dimensional information from the input features, through the multiple hidden convolutional kernels of different scales, without stacking layers depth‐wise. Simultaneously, by introducing a multi‐head attention mechanism, the TICN‐Att model has a long time‐dependent extraction capability, similar to that of the long short‐term memory network models. The generalization and validity of the model are tested using the global energy forecasting competition (GEFCOM2014) dataset, electrical load data of a city in Jiangsu (China), and PJM power system dataset. The experimental results demonstrated that the proposed model has the best prediction effect, compared to the other state‐of‐the‐art models.

When considering agricultural commodity transaction data, long sampling intervals or data sparsity may lead to small samples. Furthermore, training on small samples can lead to overfitting and makes it hard to capture the fine-grained fluctuations in the data. In this study, a multi-scale forecasting approach combined with a Generative Adversarial Network (GAN) and Temporal Convolutional Network (TCN) is proposed to address the problems related to small sample prediction. First, a Time-series Generative Adversarial Network (TimeGAN) is used to expand the multi-dimensional data and t-SNE is utilized to evaluate the similarity between the original and synthetic data. Second, a greedy algorithm is exploited to calculate the information gain, in order to obtain important features, based on XGBoost. Meanwhile, TCN residual blocks and dilated convolutions are used to tackle the issue of gradient disappearance. Finally, an attention mechanism is added to the TCN, which is beneficial in terms of improving the forecasting accuracy. Experiments are conducted on three products, garlic, ginger and chili. Taking garlic as an example, the RMSE of the proposed method was reduced by 1.7% and 1% when compared to the SVR and RF models, respectively. Its R2 accuracy was also improved (by 4.3% and 3.4%, respectively). Furthermore, TCN-attention and TCN were found to require less time compared to GRU and LSTM. The accuracy of the proposed method increased by about 5% when compared to that without TimeGAN in the ablation study. Moreover, compared with TCN, the Gated Recurrent Unit (GRU), and the Long Short-term Memory (LSTM) model in the multi-scale price forecasting task, the proposed method can better utilize small samples and high-dimensional data, leading to improved performance. Additionally, the proposed model is compared to the Transformer and TimesNet models in terms of its accuracy, deployment cost, and other metrics.

In recent years, the presence of malware has been growing exponentially, resulting in enormous demand for efficient malware classification methods. However, the existing machine learning-based classifiers have high false positive rates and cannot effectively classify malware variants, packers, and obfuscation. To address this shortcoming, this paper proposes an efficient deep learning-based method named AIFS-IDL (Atanassov Intuitionistic Fuzzy Sets-Integrated Deep Learning), which uses static features to classify malware. The proposed method first extracts six types of features from the disassembly and byte files and then fuses them to solve the single-feature problem in traditional classification methods. Next, Atanassov’s intuitionistic fuzzy set-based method is used to integrate the result of the three deep learning models, namely, GRU (Temporal Convolutional Network), TCN (Temporal Convolutional Network), and CNN (Convolutional Neural Networks), which improves the classification accuracy and generalizability of the classification model. The proposed method is verified by experiments and the results show that the proposed method can effectively improve the accuracy of malware classification compared to the existing methods. Experiments were carried out on the six types of features of malicious code and compared with traditional classification algorithms and ensemble learning algorithms. A variety of comparative experiments show that the classification accuracy rate of integrating multi-feature, multi-model aspects can reach 99.92%. The results show that, compared with other static classification methods, this method has better malware identification and classification ability.

The volatility of renewable energy generation impacts the safe and stable operation of power systems. Moreover, load uncertainty complicates renewable energy consumption. Therefore, accurately extracting load patterns using artificial intelligence (AI) technology is crucial. Load classification is an effective way to master load behaviour. However, issues in the collected load data, such as data class imbalance, significantly affect the accuracy of traditional load classification. To address this problem, this study proposes a novel classification method based on data augmentation and few‐shot learning, significantly enhancing the training efficiency of algorithm recognition. This addresses the challenge of real‐data recognition in power systems. First, time‐series load data are converted into images based on the Gramian angular field method to extract time‐series data features using a convolutional neural network. Subsequently, the data are augmented based on variational autoencoder generative adversarial network to generate samples with distributions similar to those of the original data. Finally, the augmented few‐shot data are classified using the embedding and relation modules of the relation network. A comparison of the experimental results reveals that the proposed method effectively improves power load classification accuracy, even with insufficient data.

Tool wear is an inevitable and critical problem in manufacturing systems. Accurate prediction of tool wear under varying operation conditions is a hot and challenging problem. In this paper, temporal convolutional network (TCN) is employed to predict tool wear using raw sensor signals. Dilated causal convolution and residual connections are used in TCN to form a more flexible structure with stable training gradients for sequential modeling. In addition, auxiliary learning is used in this paper to enhance model performance under varying operation conditions. An auxiliary task of recognizing operating conditions is designed to support the tool wear prediction task. Experiment on a real milling wear dataset illustrates the superiority of TCN for tool wear prediction compared with traditional convolutional neural network (CNN) and long short-term memory (LSTM) network. Experimental results also demonstrate the effectiveness of the proposed auxiliary learning.

Enhancing deterministic prediction in unidirectional ocean waves using an Artificial Neural Network with exponential linear unit

Functional Electrical Stimulation(FES) is key for restoring motor function in patients with neurological injuries, yet its insufficient parameter accuracy and real-time regulation limit finger movement control in rehabilitation. This study proposes an adaptive closed-loop FES parameter control model based on Temporal Convolutional Network(TCN).Taking real-time ring finger motion(synchronously collected by a flex sensor and camera) as input, by optimizing the Temporal Convolutional Network(TCN) architecture—with progressive dilation factors(1/2/4/8)—and sliding window parameters(window size: 500 ms, window interval:10 ms), the accuracy of temporal feature extraction is enhanced through the integration of causal convolution and residual connections. outputting three core FES parameters: voltage level, pulse width, and pulse interval. The parameters drive the FES module to control finger movement in real time, with motion data continuously fed back for dynamic adaptive closed-loop adjustment. To verify its potential for clinical application, four healthy subjects were selected. By using elastic constraint bands and a 10°C low-temperature environment, their movement characteristics were modified to simulate the condition of patients with neurological injuries. The model's robustness under non-ideal movement conditions was confirmed.

SummaryThe usage of multi‐rotor unmanned aerial vehicles (UAVs) has rapidly increased in a variety of industries recently, which has caused a quick rise in the quantity of research works on the topic. Remaining battery life prediction, anomaly detection and instantaneous power consumption prediction are among the topics that attract the most attention of researchers. This article presents the development and utilization of a modified temporal convolutional network (TCN) model, a commonly employed approach for anomaly detection and instantaneous power consumption prediction in unmanned aerial vehicles (UAVs). The first modification to the TCN model was to use the Leaky ReLu activation function in place of the rectified linear unit (ReLu) activation function from the original TCN model. In the next step, instantaneous power consumption prediction for UAVs was performed by using the data obtained from both sensors and simple moving average (SMA) algorithm. As a result of the tests performed with the created simulation setup, it has been clearly shown that the proposed method gives better results compared to other deep learning models used for comparison with the lowest RMSE of 0.0496.

In the realm of power systems, short-term electric load forecasting is pivotal for ensuring supply–demand balance, optimizing generation planning, reducing operational costs, and maintaining grid stability. Short-term load curves are characteristically coarse, revealing high-frequency data upon decomposition that exhibit pronounced non-linearity and significant noise, complicating efforts to enhance forecasting precision. To address these challenges, this study introduces an innovative model. This model employs complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) to bifurcate the original load data into low- and high-frequency components. For the smoother low-frequency data, a temporal convolutional network (TCN) is utilized, whereas the high-frequency components, which encapsulate detailed load history information yet suffer from a lower fitting accuracy, are processed using an enhanced soft thresholding TCN (SF-TCN) optimized with the slime mould algorithm (SMA). Experimental tests of this methodology on load forecasts for the forthcoming 24 h across all seasons have demonstrated its superior forecasting accuracy compared to that of non-decomposed models, such as support vector regression (SVR), recurrent neural network (RNN), gated recurrent unit (GRU), long short-term memory (LSTM), convolutional neural network-LSTM (CNN-LSTM), TCN, Informer, and decomposed models, including CEEMDAN-TCN and CEEMDAN-TCN-SMA.

An explainable attention-based TCN heartbeats classification model for arrhythmia detection

Network intrusion detection plays an important role in the network security, but the increasingly complex network environment brings a serious challenge to intrusion detection. Although the existing efficient Convolutional Neural Network (CNN)-based network traffic intrusion detection models do not require manual design of the traffic features, but they do not make full use of the structured information of network traffic. In this paper, we propose a novel network intrusion detection model based on Temporal Convolutional Networks (TCN), it extracts the key features in the dataset through exploiting the characteristics of byte sequence in the network traffic packets. Compared with traditional recurrent neural networks (RNN), TCN shows a better performance in sequence modeling tasks and it can process the sequences in parallel for faster training. To solve the problem of poor detection accuracy caused by the "death" of some neurons on ReLU during the training stage, we use the ELU activation function in the TCN instead of ReLU. Finally, we compare our proposed TCN-based intrusion detection model with the state-of-the-art methods on the CTU public dataset, and the experimental results show that the use of TCN can obtain higher performance within less time consumption, in terms of higher average accuracy, higher average recall and higher average F1-measure.

Eye-tracking technology enables communication for individuals with muscle control difficulties, making it a valuable assistive tool. Traditional systems rely on electrooculography (EOG) or infrared devices, which are accurate but costly and invasive. While vision-based systems offer a more accessible alternative, they have not been extensively explored for eye-writing recognition. Additionally, the natural instability of eye movements and variations in writing styles result in inconsistent signal lengths, which reduces recognition accuracy and limits the practical use of eye-writing systems. To address these challenges, we propose a novel vision-based eye-writing recognition approach that utilizes a webcam-captured dataset. A key contribution of our approach is the introduction of a Discrete Fourier Transform (DFT)-based length normalization method that standardizes the length of each eye-writing sample while preserving essential spectral characteristics. This ensures uniformity in input lengths and improves both efficiency and robustness. Moreover, we integrate a hybrid deep learning model that combines 1D Convolutional Neural Networks (CNN) and Temporal Convolutional Networks (TCN) to jointly capture spatial and temporal features of eye-writing. To further improve model robustness, we incorporate data augmentation and initial-point normalization techniques. The proposed system was evaluated using our new webcam-captured Arabic numbers dataset and two existing benchmark datasets, with leave-one-subject-out (LOSO) cross-validation. The model achieved accuracies of 97.68% on the new dataset, 94.48% on the Japanese Katakana dataset, and 98.70% on the EOG-captured Arabic numbers dataset—outperforming existing systems. This work provides an efficient eye-writing recognition system, featuring robust preprocessing techniques, a hybrid deep learning model, and a new webcam-captured dataset.

Focusing on the problems of complex structure and low feature extraction efficiency that exist in some traditional neural network algorithms, an improved convolutional neural network, combined with a convolutional attention mechanism, is proposed to im- prove the network feature extraction efficiency and reduce the network complexity. First, the convolutional block attention module (CBAM) was introduced, which can strengthen the extraction of useful features from the data, thereby improving the network feature extraction efficiency. Second, the residual structure was optimized, and the original multi-layer convolution was simplified to a single-layer convolution, which reduces the complexity of the model and improves the diagnostic efficiency. Then, the Rectified Linear Unit (ReLU) activation function in the model was replaced with the better-performing Exponential Linear Unit (ELU) activation function. Finally, the overall structure of the ResNet model was optimized. The CBAM-ResNet network model was based on the traditional ResNet18 model and consists of a CBAM module and the optimized ResNet18 model. Finally, experiments were performed using sensor data collected from rolling bearing test benches. The experimental results show that the feature extraction capability and efficiency of the new model are better than those of several other classic deep network models and the original network model, and this reduces the time complexity and model size while also maintaining a high accuracy rate.

The consumer-side load probabilistic forecasting is meaningful and useful in the smart grid area and becoming more and more possible with the explosion of electricity consumption data collected by smart meters. Compared to aggregated load, forecasting on the consumer-side load prefers to be stochastic and uncertain, which places higher demands on the forecast performance of models. In this paper, a novel probabilistic forecasting method, quantile Huber guided Temporal convolutional network (TCN), is proposed to quantify the variability and uncertainty of consumer-side load forecasting. Specifically, TCN is a new deep neural network implemented for capturing the dependencies information within the load sequences. The quantile Huber function formulated weighted average multiple losses (WAML) function is used to generate probabilistic prediction intervals (PIs). Conditional distribution of deterministic forecasting residual is also considered to calibrate the probabilistic forecasting. The performance of the proposed forecasting method is validated by numerical experiments using real load datasets of residential and industrial consumers. The comparative results show the superiority of the proposed method over traditional methods.

Short-term cooling load prediction for central air conditioning systems with small sample based on permutation entropy and temporal convolutional network