With the advancement of deep learning (DL) technology, DL-based intrusion detection models have emerged as a focal point of research within the domain of cybersecurity. This paper provides an overview of the datasets frequently utilized in the research. This article presents an overview of the widely utilized datasets in the research, establishing a basis for future investigation and analysis. The text subsequently summarizes the prevalent data preprocessing methods and feature engineering techniques utilized in intrusion detection. Following this, it provides a review of seven deep learning-based intrusion detection models, namely, deep autoencoders, deep belief networks, deep neural networks, convolutional neural networks, recurrent neural networks, generative adversarial networks, and transformers. Each model is examined from various dimensions, highlighting their unique architectures and applications within the context of cybersecurity. Furthermore, this paper broadens its scope to include intrusion detection techniques facilitated by the following two large-scale predictive models: the BERT series and the GPT series. These models, leveraging the power of transformers and attention mechanisms, have demonstrated remarkable capabilities in understanding and processing sequential data. In light of these findings, this paper concludes with a prospective outlook on future research directions. Four key areas have been identified for further research. By addressing these issues and advancing research in the aforementioned areas, this paper envisions a future in which DL-based intrusion detection systems are not only more accurate and efficient but also better aligned with the dynamic and evolving landscape of cybersecurity threats.
With the continuous increase in the number of Retired Lithium-Ion Batteries (RLBs), accurately estimating their Maximum Remaining Capacity (MRC) has become a key challenge for rapid sorting and secondary utilization. Conventional detection methods often suffer from low efficiency and limited scalability for large-scale applications. To address these issues, this paper presents a rapid MRC estimation method using a hybrid Convolutional Neural Network (CNN), Conv Block Attention Module (CBAM), and Long Short-Term Memory (LSTM) architecture. The proposed approach extracts key voltage and capacity features from only the initial 30 min charging phase, integrating both factory and laboratory data. Specifically, the CNN captures local temporal patterns, the LSTM models long-term dependencies, and the CBAM adaptively emphasizes critical characteristics. Experimental results demonstrate that the proposed method significantly outperforms traditional approaches, achieving a testing R2 of 98.05% and a Mean Absolute Percentage Error (MAPE) of 1.60%. These results highlight the superior performance of the proposed framework, exhibiting strong potential for high-throughput battery sorting and large-scale health monitoring systems.
With the continuous increase in the number of retired lithium-ion batteries, accurately and quickly estimating their MRC has become a key challenge for the rapid sorting and secondary utilization of retired lithium-ion batteries. Conventional detection methods often suffer from low efficiency, prolonged detection cycles, and limited scalability for large-scale applications. To address these issues, this paper presents a fast MRC estimation method for retired lithium-ion batteries using a hybrid Convolutional Neural Network (CNN)-Conv Block Attention Module (CBAM)-Long Short-Term Memory (LSTM) architecture (CNN-CBAM-LSTM). The proposed approach integrates both factory-scale test data and laboratory experimental data to extract key voltage and capacity features from the initial 30-minute charging phase. Specifically, the CNN captures local temporal patterns, the LSTM models long-term dependencies in the time-series data, and the CBAM enhances feature representation by emphasizing critical characteristics. Experimental results demonstrate that the proposed method achieves MRC estimation within 30 minutes, significantly outperforming traditional approaches in terms of accuracy. The R² value increased to 99.42%, while the MAPE decreased to 1.55%. These results highlight the superior performance of the proposed method, which not only holds strong potential for rapid battery sorting and cascaded utilization but also exhibits broad applicability in large-scale battery health monitoring systems.