Event-Based Anomaly Detection Using a One-Class SVM for a Hybrid Electric Vehicle

Ji, Yonghyeok; Lee, Hyeongcheol

doi:10.1109/tvt.2022.3165526

Cited by 25 publications

(7 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The final detection results are obtained by using the trained OC-SVM classifier [25]. As shown in Figs.…”

Section: Experimental Results and Analysismentioning

confidence: 99%

“…That is, the background and the potential target are divided in adaptive threshold module; (5) According to the CFAR detection results, the clutter feature values F are separated. The features F of clutter data are used on the OC-SVM classifier learning and training [25]; (6) The designed anomaly detection system based on OC-SVM is used to extract outlier values (corresponding to the target) from the feature values F of SAR image.…”

Section: Detection Algorithm For Shipsmentioning

confidence: 99%

See 1 more Smart Citation

Curvature-based Feature Representation for Ship Detection in SAR Image

Chen,

Yang

2024

PIER Letters

View full text Add to dashboard Cite

This article aims to exploit Ricci tensor with certain geometric properties which are used for feature representation and ship detection in synthetic aperture radar (SAR) image. The proposed method is composed of the following key points. Firstly, Riemannian metrics on the Gamma manifold are constructed based on the family of Gamma density functions. Secondly, direct calculation gives the Ricci tensor of Gamma manifold, where the curvature tensor resorts to the torsion-free affine connection. Thirdly, a general scheme for Zermelo navigation problem on the Riemannian manifold is addressed, and the solution of the navigation problem is proposed. Fourthly, feature representation problems are formulated as certain forms of Finsler metric of Randers type, indicating a joint framework for extracting low-dimensional features with closed-form solutions. Comprehensive experiments on real SAR image data sets demonstrate the effectiveness of the proposed method against compared state-of-the-art detection approaches.

show abstract

“…The final detection results are obtained by using the trained OC-SVM classifier [25]. As shown in Figs.…”

Section: Experimental Results and Analysismentioning

confidence: 99%

Section: Detection Algorithm For Shipsmentioning

confidence: 99%

Curvature-based Feature Representation for Ship Detection in SAR Image

Chen,

Yang

2024

PIER Letters

View full text Add to dashboard Cite

show abstract

“…As a result, DL observers showed 84% to 95% of estimation accuracy between estimates and actual anomalies. Ji et al [12] presented an anomaly detection approach utilizing a oneclass Support Vector Machine (SVM) to validate control functions and streamline the analysis of test data from a hybrid electric vehicle (HEV). However, the results fail to highlight which signals cause anomalies for a comprehensive understanding.…”

Section: A Dl-based Anomaly Detectionmentioning

confidence: 99%

A Federated Learning Approach for Efficient Anomaly Detection in Electric Power Steering Systems

Kea,

Han,

Min

2024

IEEE Access

View full text Add to dashboard Cite

Electric power steering (EPS) has emerged as a valuable driver-assistance system. In an EPS system, an extensive amount of data collected from various sensors is analyzed to enhance the driving experience. Anomaly detection techniques have shown potential in ensuring the integrity of data patterns and detecting abnormalities to prevent adverse driving incidents, thus improving vehicle safety. However, traditional centralized anomaly detection methods require the collection of data from all EPS sensors, resulting in high communication network overhead. Herein, we propose an approach for anomaly detection using EPS data within a federated learning (FL) framework. Our approach exploits deep learning (DL) on time series data to achieve highly effective anomaly detection. This synergy of FL and DL enables all sensors in an EPS system to collaboratively train the anomaly detection model simultaneously. Our results demonstrate the feasibility of combining FL with DL anomaly detection in EPS systems, thus overcoming the limitations of the traditional centralized approach. The combined FL and DL model performs remarkably well, achieving an F1-Score of 99.01%, outperforming the 98.31%∼98.89% achieved by the centralized approach. It also exhibits high generalizability by incorporating insights from all sensors to comprehensively understand diverse driving scenarios. This results in a significant reduction in error rates, ranging from 20% to 33.25%, compared to centralized methods. Additionally, the proposed model exhibits significant advantages over traditional methods, including reduced training time and communication overhead, while maintaining comparable anomaly detection accuracy performance.INDEX TERMS Anomaly detection, deep learning, electric power steering (EPS), federated learning.

show abstract

“…The torque ripple generated by the engine at the side of driving plate can also have an adverse effect during the process of clutch engagement. Despite the short duration of the slipping phase, the nonlinear variation in torque will intensify the mismatching torque in the vehicle powertrain, leading to an obvious unpleasant driving sensation [20][21][22][23]. Hence, further improvement in the drivability and smoothness of HEVs during mode transition remains a challenge.…”

Section: Introductionmentioning

confidence: 99%

A Novel Coordinated Control Strategy for Parallel Hybrid Electric Vehicles during Clutch Slipping Process

Tian

Wang

et al. 2022

Applied Sciences

View full text Add to dashboard Cite

For parallel hybrid electric vehicles (HEVs), the clutch serves as a vital enabling actuator element during mode transitions. The expected drivability and smoothness of parallel HEVs are difficult to be achieve owing to the neglect of clutch-torque-induced disturbance and different response characteristics of power sources during clutch slipping. To address this issue, this paper proposes a novel control strategy to coordinate the engine and motor during the clutch slipping process. A sliding mode control strategy based on a group-preserving scheme was applied to control the motor. The vehicle dynamic equation was constructed by the sliding surface with the Lagrange function. The equation solutions obtained by introducing the Runge–Kutta method were used as motor control inputs. Meanwhile, an adaptive PI controller was designed to regulate engine speed for the reduction in the speed difference of the clutch. The hardware-in-the-loop simulations were conducted to validate the outstanding performance of the proposal strategy. The verification results indicate that the proposed strategy not only reduces the vehicle jerk and frictional losses effectively, but also improves vehicle driving comfort and reliability.

show abstract

Event-Based Anomaly Detection Using a One-Class SVM for a Hybrid Electric Vehicle

Cited by 25 publications

References 53 publications

Curvature-based Feature Representation for Ship Detection in SAR Image

Curvature-based Feature Representation for Ship Detection in SAR Image

A Federated Learning Approach for Efficient Anomaly Detection in Electric Power Steering Systems

A Novel Coordinated Control Strategy for Parallel Hybrid Electric Vehicles during Clutch Slipping Process

Contact Info

Product

Resources

About