Table 1 Summary for the related works.
Ref. no | Methodology | Objective | Dataset/scenario | Reported metrics (accuracy, RMSE, etc.) | Key findings |
|---|---|---|---|---|---|
Hybrid AI classification (HAICM) | Improves scheduling efficiency | EVs in V2G with 5G | Accuracy = 93%, Scheduling Latency ↓ | Efficiently identifies target EVs with cross-validation | |
NS-EC-SG + hybrid AI | Predict charging behavior | 5G Smart Grid, edge computing | Prediction Accuracy = 95%, Cost ↓ 12% | Better prediction vs. SOTA | |
Cloud + AI integration | Dynamic charging optimization, fleet mgmt. | EV networks | Reliability ↑ 15%, Maintenance Cost ↓ | Enhanced efficiency & fleet performance | |
AI + blockchain (AEBIS) | Secure V2G energy management | Smart grid simulation | R2 = 0.938R^2 = 0.938, Power Fluctuation ↓ | Secure and consistent energy delivery | |
LSTM vs. ML methods (DNN, KNN, RF, SVM, DT) | Compare ML for EV load prediction | Smart city EV dataset | LSTM: RMSE ↓ 20%, Billing Cost ↓ 8% | LSTM outperforms other ML models |
