Core Engine: Analytics & Hardware

The private monorepo for the capacitor condition monitoring project. Contains all simulation files, edge C++ code, and ML training pipelines.

---

Repository Layout

• /hardware-esp32 — C++ code for the ESP32 to calculate ESR and Capacitance.
• /simulation-matlab — Simulink models (.slx & .mdl) and the MATLAB runner script (.m).
• /ml-engine — Jupyter notebooks and Python scripts for FFT processing and XGBoost inference.
• /data-pipeline — Scripts for generating synthetic physics-informed datasets.

---

Setup & Execution

1. Hardware (ESP32)

1. Open /hardware-esp32/esr_c_calc.ino in the Arduino IDE.
2. Flash to the ESP32. Ensure the baud rate is set to 115200.
3. The ESP32 will wait for Serial inputs formatted as V6,V12\n.

2. Simulation (MATLAB)

1. Open MATLAB and navigate to the /simulation-matlab folder.
2. Run run_simulation.m.
3. This script will automatically execute the Simulink model and export the dynamic vdc_data.csv and temp_reference.txt files to your local environment.

3. ML Engine & Watchdog

1. Install Python dependencies:

pip install pandas numpy scipy xgboost paho-mqtt pyserial

2. Run the main processing script:

jupyter notebook ml-engine/fft_rul_hi_xgb.ipynb