Academic project
Expert System for Wind Turbine Diagnostics
A simple but useful portfolio case: it turns turbine measurements into explainable diagnostics.
Goal
The system helps maintenance technicians interpret critical turbine variables and suggest the next step using rule-based reasoning.
How it works
- An event contains a set of turbine measurements.
- The rules engine compares the values against the expected ranges.
- The system returns a diagnosis and a justification.
Variables
11
wind, rpm, vibration, temperature, pressure, and consumption
Diagnoses
8+
rules focused on common operational failures
Output
Explainable
every result includes a reason and a suggested action
Format
Static
content is versioned in Git and published without a backend
Main rules
- Very low RPM with abnormal wind: possible sensor or actuator failure.
- Hydraulic blade pressure out of range: possible leak or air in the oil.
- Torque out of pattern: possible calibration error.
- High vibration with normal oil temperature: check the yaw system.
- High oil temperature with normal vibration: temperature adjustment.
- Lubricating oil pressure out of range: possible leak or pump issue.
- High vibration, temperature, and noise: possible bearing failure.
- Abnormal electrical consumption: possible mechanical error.
- Power inconsistent with wind: possible sensor error.
Diagnosis example
Diagnosis: Sensor or actuator failure
Reason: Very low RPM with abnormal wind.
Suggested action: check the speed sensor and control system.
Flowchart
Closing note
This project works well in a portfolio because it shows rule-based reasoning, explainable diagnosis, and an industrial maintenance focus.