Description

Problem

Reliability assessment of renewable energy systems typically relies on manually constructed reliability models and assumed failure rates. However, modern renewable energy systems continuously generate operational data that can be used to automatically identify failures and estimate reliability parameters. Controlled laboratory experiments provide a unique opportunity to collect labeled fault data under reproducible conditions, enabling the development and validation of data-driven reliability models.

Goal

Develop a framework for automatically extracting reliability models from experimental data collected from photovoltaic, wind energy, or hybrid renewable energy systems. The framework should detect faults from time-series measurements, estimate reliability parameters such as failure and repair rates, construct models such as Fault Trees or Markov Chains, and evaluate system reliability and availability through simulation.

Requirements

• Basic knowledge of reliability engineering
• Python programming
• Time-series data analysis
• Probability and statistics
• Basic knowledge of renewable energy systems