⚡ APTRANSCO Smart Grid Unified Intelligence System

An advanced, multi-modal machine learning platform and real-time ingestion pipeline designed for state-level electrical power infrastructure. This system integrates IoT telemetry, drone imagery analytics, and weather sensor data to predict Asset Remaining Useful Life (RUL), execute dynamic line rating adjustments, and perform predictive network load balancing.

---

🛠️ Core Functional Architecture

1. Multi-Modal ML Engine

• Asset Health Regressor: Built on a RandomForestRegressor ensemble pipeline trained across multiple vectors.
• Feature Integration: Synthesizes internal thermal loads (temp_C, load_pct), structural health degradation indices (insulation_health), drone spatial anomalies (veg_distance_m, conductor_sag_cm), and environmental IoT metrics (ambient_temp_C).

2. Dynamic Line Rating (DLR) Logic

• Adaptive Ceilings: Automatically scales safe operational line thresholds between 76.0% and 96.0% capacity.
• Thermal Modeling: Computes ambient cooling coefficients using real-time local wind velocity measurements weighted against thermal air stresses.

3. Predictive Load Balancing Engine

• Automated Mitigation: Scans the infrastructure mesh for nodes dropping below critical RUL boundaries (< 30 days).
• Shedding Optimization: Isolates available system headroom and smoothly steps down overstressed elements, rerouting power directly into under-utilized grid layers without dropping generation baseloads.

---

💻 Tech Stack

• Language: Python 3.10+
• ML Framework: Scikit-Learn (Random Forest Ensemble)
• Data Processing: NumPy, Pandas
• UI Layer: Streamlit Web Framework / Interactive Control Room CLI Terminal

---

⚙️ Installation & Workspace Setup

1. Clone the Project

git clone https://github.com
cd aptransco-grid-intelligence

2. Configure Your Virtual Environment

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies

pip install numpy pandas scikit-learn streamlit

---

🚀 How to Run

Option A: Interactive Streamlit Dashboard Engine

To run the high-performance, responsive live tracking interface:

streamlit run app.py

Option B: Integrated Control Room Console Outflow

To start the standard telemetry streaming loop in a terminal session:

python grid_monitor.py

---

📁 System Parameters & Storage Outflow

Financial Allocation Ledger (INR ₹)

The platform evaluates financial risk exposure and tracks capital preservation values in real time:

Financial Cost Parameter	Allocated Value (INR)	Description
COST_REPLACEMENT_GEN	₹4,50,00,000	Total Generation Asset Unit Fail Capital
COST_REPLACEMENT_TX	₹1,40,00,000	Major Transmission Asset Fail Capital
COST_LINE_REPAIR	₹22,00,000	Local Distribution/Feeder Line Fix
REGULATORY_FINE	₹1,80,00,000	State Environmental & Safety Breach Fine
EMERGENCY_LABOR_RATE	₹16,000 / hr	Reactive Breakdown Emergency Crew Rate
PLANNED_LABOR_RATE	₹4,200 / hr	Proactive Preventive AI Dispatch Rate

Persistence Layer

On every pipeline cycle tick, the engine appends compressed telemetry and optimization logs directly into the centralized historical ledger file:

• Output Path: ap_grid_unified_intelligence.csv
• Log Columns: timestamp, asset_id, level, temp_C, load_pct, predicted_rul, optimized_load_pct, reactive_risk_inr, net_saved_capital_inr.

---

📄 License

This grid control system software is proprietary property. Licensed for authorized use inside APTRANSCO and APSPDCL Power Systems Operations Control Centres (SLDC).

---

Created by: Srinivasta