Fuzzy robust model predictive fault tolerant control in wind turbine based on ST-SMO and PMIO

Area of Science:

• Renewable Energy Systems
• Control Engineering
• Mechanical Engineering

Background:

• Wind turbine maintenance often relies on post-failure diagnosis, leading to operational inefficiencies.
• Sustained operation during degraded conditions is critical for energy production.

Purpose of the Study:

• To develop a fault-tolerant model predictive control (MPC) framework for megawatt-class wind turbines.
• To improve fault detection, isolation, and compensation capabilities, especially under complex fault scenarios.

Main Methods:

• Implementation of a dual observer system: a proportional multiple integral (PMI) observer and a super-twisting sliding mode observer (ST-SMO).
• Coordination of observers using Takagi-Sugeno (T-S) fuzzy logic based on the ν-gap metric.
• Adaptive penalty terms in the MPC layer utilizing a linear parameter-varying (LPV) model for deviation compensation.

Main Results:

• 18% improvement in maximum power tracking accuracy under sensor faults.
• 50% suppression of drivetrain torsional torque fluctuation during cross-load switching.
• 19.63% reduction in tower bending moment damage equivalent load (DEL) under multi-fault conditions.

Conclusions:

• The proposed framework synergistically optimizes fault tolerance, fatigue load mitigation, and active power maximization.
• Innovative integration of dual observers, T-S fuzzy logic, and LPV switching addresses multi-fault coupling and cross-load transitions effectively.
• Demonstrates superior performance compared to single-observer fault-tolerant control (FTC) methods and conventional MPC.