GridLog-Cascade 2025

The Smart-Grid–Logistics Cyber-Physical Dataset (SGL-CPD 2025) is a real-world multivariate time-series dataset capturing the operational coupling between smart grid infrastructure, renewable energy integration, battery storage systems, cold-chain warehouse logistics, EV fleet charging networks, and industrial cybersecurity telemetry. The dataset reflects practical deployment environments where distribution grid operators, cold-chain logistics providers, and industrial IoT systems operate in an interconnected cyber–physical ecosystem. Data were collected from synchronized monitoring streams typically available in SCADA-based grid supervision systems, renewable generation controllers, warehouse management platforms, refrigeration control systems, EV charging stations, and cybersecurity monitoring logs. The dataset models realistic operational behavior under renewable variability, demand–supply imbalance, grid stress, automation load, and cyber intrusion events. The dataset contains 476,716 time-stamped records sampled at 5-minute intervals, ending at 2025-12-25 23:55:00 (inclusive). Each record represents a synchronized operational snapshot of grid, logistics, and cybersecurity states. The data are suitable for multi-task modeling, including cyberattack detection (multi-label classification), grid load forecasting (multi-horizon regression), and cascading grid failure severity prediction (multi-class classification). The class distribution reflects real operational imbalance, where normal behavior dominates, cyberattacks are rare, and severe cascading failures are very rare events. Feature Groups 1. Time and Calendar FeaturesTimestamp and temporal indicators used for time-series modeling: timestamp hour dayofweek month dayofyear year is_weekend hour_sin, hour_cos doy_sin, doy_cos 2. Renewable Generation and Grid Operation FeaturesThese features describe renewable integration, grid demand, stability, and stress conditions: solar_irradiance_wm2 pv_power_mw wind_speed_ms wind_power_mw renewable_supply_mw gen_fluctuation_index total_load_mw demand_supply_gap_mw grid_frequency_hz voltage_deviation_pct transformer_load_pct line_congestion_index 3. Energy Storage and Backup Support FeaturesBattery and backup generation characteristics representing grid resilience: battery_soc_pct battery_charge_kw battery_discharge_kw backup_gen_status backup_autonomy_hr 4. Cold-Chain Warehouse and Environmental FeaturesThese features capture refrigeration stability, environmental control, and warehouse energy usage: warehouse_temp_c refrigeration_load_kw humidity_pct cold_storage_uptime_pct warehouse_energy_kwh 5. Warehouse Automation and Throughput FeaturesOperational metrics from automated logistics systems: robot_utilization_pct conveyor_speed_mps inventory_throughput_items_hr order_processing_delay_min 6. EV Fleet Charging and Dispatch FeaturesElectric fleet operational and charging indicators: charging_station_load_kw fleet_soc_mean_pct charging_wait_min vehicle_dispatch_delay_min route_completion_rate_pct 7. Cybersecurity Telemetry FeaturesSecurity monitoring indicators derived from industrial control and IoT systems: scada_login_anom_cnt cmd_injection_alerts_cnt firmware_hash_mismatch network_latency_ms intrusion_severity_score 8. Cross-Domain Coupling FeaturesEngineered indicators quantifying interdependence between energy and logistics operations: energy_to_cooling_ratio charging_to_generation_ratio logistics_energy_criticality_score cascade_propagation_lag_min Target Variables Attack_Event (Multi-Label Classification) attack_primary_class (0–7) attack_bitmask attack_fdi attack_dos attack_cmd_inj attack_replay attack_meas_tamper attack_malware attack_load_alt Grid_Load_Forecast (Regression Targets) y_load_tplus_1h_mw y_load_tplus_6h_mw y_load_tplus_24h_mw Cascading_Grid_Failure (Multi-Class Severity Label) cascade_failure_severity (0–3)