This project aims to push the boundary for the fault-tolerant control technology of offshore wind turbines (OWTs), which is of significant importance considering the fact that OWTs are usually under long-duration operations subject to harsh environments in remote areas. An innovative hybrid fault-tolerant control system that takes advantage of both passive & active fault-tolerant methods will be developed via deep reinforcement learning to achieve rapid performance recovery & optimal fault compensation for OWTs. The resulting methods will be data-driven and model-free, overcoming the limitations of mainstream fault-tolerant control approaches, e.g., reliance on accurate analytical models and lack of adapting abilities to different types of faults. The outcomes will reduce operation & maintenance costs and extend the lifespan of OWTs, promoting offshore wind’s development. This project will also allow the PIs to promote the cooperation between Warwick and Hull in developing innovative machine learning and artificial intelligence technologies for offshore wind, leading to future research projects and funding in this promising area.
