Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Federated Anomaly Detection and Mitigation for EV Charging Forecasting Under Cyberattacks

Oluleke Babayomi, Dong-Seong Kim

TL;DR

The paper tackles cyber threats and data privacy in EV charging demand forecasting by proposing an anomaly-resilient federated LSTM framework that combines LSTM autoencoder-based anomaly detection with interpolation-based data mitigation and Federated Averaging for privacy-preserving collaboration. It validates the approach on Shenzhen EV data with simulated DDoS attacks, demonstrating a $15.2 ext{%}$ improvement in $R^2$ over centralized models, $47.9 ext{%}$ recovery of attack-induced loss, and a precision of $0.913$ with a false positive rate of $0.0121$. The integrated anomaly detection effectively identifies attacks while preserving temporal integrity, enabling trustworthy datasets for forecasting. The findings support a paradigm shift toward distributed intelligence in industrial IoT and smart grids, offering privacy-preserving, resilient, and scalable forecasting for distributed EV charging networks.

Abstract

Electric Vehicle (EV) charging infrastructure faces escalating cybersecurity threats that can severely compromise operational efficiency and grid stability. Existing forecasting techniques are limited by the lack of combined robust anomaly mitigation solutions and data privacy preservation. Therefore, this paper addresses these challenges by proposing a novel anomaly-resilient federated learning framework that simultaneously preserves data privacy, detects cyber-attacks, and maintains trustworthy demand prediction accuracy under adversarial conditions. The proposed framework integrates three key innovations: LSTM autoencoder-based distributed anomaly detection deployed at each federated client, interpolation-based anomalous data mitigation to preserve temporal continuity, and federated Long Short-Term Memory (LSTM) networks that enable collaborative learning without centralized data aggregation. The framework is validated on real-world EV charging infrastructure datasets combined with real-world DDoS attack datasets, providing robust validation of the proposed approach under realistic threat scenarios. Experimental results demonstrate that the federated approach achieves superior performance compared to centralized models, with 15.2% improvement in R2 accuracy while maintaining data locality. The integrated cyber-attack detection and mitigation system produces trustworthy datasets that enhance prediction reliability, recovering 47.9% of attack-induced performance degradation while maintaining exceptional precision (91.3%) and minimal false positive rates (1.21%). The proposed architecture enables enhanced EV infrastructure planning, privacy-preserving collaborative forecasting, cybersecurity resilience, and rapid recovery from malicious threats across distributed charging networks.

Federated Anomaly Detection and Mitigation for EV Charging Forecasting Under Cyberattacks

TL;DR

The paper tackles cyber threats and data privacy in EV charging demand forecasting by proposing an anomaly-resilient federated LSTM framework that combines LSTM autoencoder-based anomaly detection with interpolation-based data mitigation and Federated Averaging for privacy-preserving collaboration. It validates the approach on Shenzhen EV data with simulated DDoS attacks, demonstrating a improvement in over centralized models, recovery of attack-induced loss, and a precision of with a false positive rate of . The integrated anomaly detection effectively identifies attacks while preserving temporal integrity, enabling trustworthy datasets for forecasting. The findings support a paradigm shift toward distributed intelligence in industrial IoT and smart grids, offering privacy-preserving, resilient, and scalable forecasting for distributed EV charging networks.

Abstract

Electric Vehicle (EV) charging infrastructure faces escalating cybersecurity threats that can severely compromise operational efficiency and grid stability. Existing forecasting techniques are limited by the lack of combined robust anomaly mitigation solutions and data privacy preservation. Therefore, this paper addresses these challenges by proposing a novel anomaly-resilient federated learning framework that simultaneously preserves data privacy, detects cyber-attacks, and maintains trustworthy demand prediction accuracy under adversarial conditions. The proposed framework integrates three key innovations: LSTM autoencoder-based distributed anomaly detection deployed at each federated client, interpolation-based anomalous data mitigation to preserve temporal continuity, and federated Long Short-Term Memory (LSTM) networks that enable collaborative learning without centralized data aggregation. The framework is validated on real-world EV charging infrastructure datasets combined with real-world DDoS attack datasets, providing robust validation of the proposed approach under realistic threat scenarios. Experimental results demonstrate that the federated approach achieves superior performance compared to centralized models, with 15.2% improvement in R2 accuracy while maintaining data locality. The integrated cyber-attack detection and mitigation system produces trustworthy datasets that enhance prediction reliability, recovering 47.9% of attack-induced performance degradation while maintaining exceptional precision (91.3%) and minimal false positive rates (1.21%). The proposed architecture enables enhanced EV infrastructure planning, privacy-preserving collaborative forecasting, cybersecurity resilience, and rapid recovery from malicious threats across distributed charging networks.

Paper Structure

This paper contains 22 sections, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Data Description and Preparation
  4. Anomaly Injection and Detection Framework
  5. Forecasting Model Architecture
  6. Centralized LSTM Model
  7. Federated LSTM Model
  8. Discussion of Results
  9. Experimental Design and Evaluation Framework
  10. Impact of Simulated DDoS Attacks on Forecasting Performance
  11. Effectiveness of Anomaly Detection and Filtering Mechanisms
  12. Federated versus Centralized Architecture Performance
  13. Local Specialization versus Global Generalization
  14. Training Dynamics and Optimization
  15. Computational Efficiency
  16. ...and 7 more sections

Figures (3)

  • Figure 1: EV charging demand forecasting. (a) Conventional centralized learning approach. (b) Proposed trustworthy anomaly-resilient federated LSTM framework.
  • Figure 2: Performance of the proposed anomaly-resilient federated LSTM for Client 1.
  • Figure 3: Comparison of R$^2$ for both federated LSTM and centralized LSTM for filtered data.