EMS Core Features
Core Infrastructure & Device Connectivity
-
Modbus Protocol Adapter - Read/write registers via Modbus TCP/RTU; device-profile based.
-
Edge Core Cycle - Deterministic 1s (configurable) execution loop (scheduler).
-
Edge Configuration Sync - Edge pulls latest device mappings and controller parameters from cloud.
-
Command Dispatch - Bidirectional WebSocket: send setpoints from cloud/dashboard to edge.
-
CAN Bus Support - Driver for BMS communication
-
SunSpec Compilance - Auto-discovery and parsing of SunSpec-compilant inverters.
Device Abstraction & Virtualisation
-
ESS Abstraction - Standardised battery channels: SOC, power, limits, setpoints.
-
Grid/Site Meter - Unified import/export meter;
-
PV Inverter Abstraction - Active power, energy, voltage and current.
-
Load Meter Abstraction - Consumption meter, enables net/gross analysis.
-
Advanced SOC Estimation - Kalman filter (EKF) with voltage correction and improved accuracy.
-
Machine Learning SOC - LSTM/GRU model trained on site-specific BMS data; adaptive to ageing.
-
Reactive Power Capability - Inverter channels for Q setpoint, power factor, Volt-VAR curves.
-
Virtual Meter - Computer channels (sum of PV, wind, net load)
Control & Optimisation
Basic Controllers
-
Self-consumption (Balancing) - Zero grid target; charge from excess PV/wind, discharge to meet load.
-
Peak Shaving - Discharge battery when grid import exceeds threshold; reduce deman charges.
-
Manual Override - Operator can set fixed charge/discharge power via dashboard.
-
Load Following - Maintain predefined grid import/export profile; follow utility signal.
-
Fixed Power Factor - Maintain constant PF at grid connection point.
-
Volt-VAR Control - Autonomous reactive power response to voltage deviations.
-
Frequency-Watt - Active power curtailment during over-frequency events.
Predictive Controllers
-
Static Time-of-Use (ToU) - Schedule, charge during off-peak, discharge during on-peak.
-
Load & PV Forecasting - Persistence model; 24h horizon, 15min resolution.
-
ML-Based Forecasting - XGBoost/LightGBM trained on site data + weather API;
-
Optimal ToU Dispatch - Linear programming (LP) scheduler; minimises energy cost; respects battery degradation.
-
Dynamic Peak shaving - Forecast-aware threshold adjustment; avoid peaks proactively.
-
Demand Response - Accept external curtailment signals (OpenADR 2.0b).
Monitoring & Visualization
-
Live Single-Line Diagram - Real-time power flow; grid, solar, wind, battery, loads; updates via WebSocket.
-
Multi-Site Overview - Cards with key metrics (total solar, battery SOC, grid import, alarms)
-
Device Detail Page - Real-time channel values + mini-trends (last hour).
-
Historical Energy Charts - Consumed, generated, stored energy over user-selected range.
-
Alarm list & Timeline - Active/acknowledged/cleared alarms with severity and timestamp.
-
Energy mix Pie Chart - Share of solar, wind, grid, battery in supply.
Analytics & Reporting
-
Daily/Weekly/Monthly Energy Report - PDF/CSV; consumption, generation, battery throughput, grid exchange.
-
Peak Demand Report - Top 5 peak 15-min intervals per billing period.
-
Estimated Savings - Compare actual grid import with baseline (flat threshold).
-
Carbon Avoidance - Co2 emissions avoided (based on grid mix factors)
-
Power Quality Report - Voltage deviations, frequency, harmonics.
-
Battery Degradation Report - Throughput vs SOH estimate; remaining useful life.
-
Anomaly Detection - Unsupervised ML to identify unsual consumption/generation patterns.
-
Automated Email reports - Schedules delivery to stakeholders.
Management & Administration
-
Multi-Tenant Hierarchy - Organisation -> Site -> Edge -> Device; strict data isolation.
-
Role-based access control - Admin, Operator, Viewer roles per organisation.
-
Audit Log - Track user actions: configuration changes, manual commands.
-
Bulk Device Provisioning - CSV import; auto-assignment to sites.
Alarms & Events
-
Threshold Alarm Rules - User-definable high/low limits on any numeric channel.
-
Communication Loss Alarm - Edge or device disconnect detected; auto-raised.
-
Alarm Acknowledgement - User can acknowledge; status reflected in UI.
-
Email Notifications - Send alarm emails to configured receipients.
-
Alarm Escalation - If not acknowledged, notify higher-level user.
-
Predictive Alarms - ML-based early warning (e.g battery may fail in 7 days.)
Reliability & Scalability
-
Edge Heartbeat Monitoring - Cloud tracks last seen timestamp; alerts on stale.
-
TimescaleDB Compression - Reduce Storage costs for historical data.
-
Edge Local Storage - Cache telemetry locally if cloud unreachable; backfill on reconnect.
-
Raft-Based Edge HA - Active-passive cluster with automatic failover (no control gap).