Advanced Control Design for Load Reduction
The purpose of this task is the development of Multi Input Multi Output robust control strategies for WT, with particular reference to the problem of load reduction.
Advanced nonlinear control methodologies are intended to be used (particularly Sliding Mode control, Generalized Predictive control) in order to cope with the noticeable level of uncertainty present in the model and with the different conditions which the wind turbine can work into.
Extensive testing of the controllers of widely accepted simulators (e.g. NREL FAST) will be performed, to establish a real comparison among different techniques as far as load reduction and performance optimization is concerned.
Advanced Supervisory Control Design
State-of-the-art WT commonly use a single controller with a fixed structure to cover a large spectrum of different operational situations, making it difficult to optimize the controller parameters.
Supervised control algorithms could help in managing such a situation.
- Controller adaptation schemes will be developed using advanced (predicted) information of environmental conditions (e.g. rotor inflow conditions) , in order to perform either the suitable formulation of the control objectives, the correct choice of the control methods more appropriate for the detected conditions and the design of a set of controllers able to manage the specific operative conditions
- Novel approaches will be explored inspired by modern methodologies for the synthesis of Cyber Physical Systems
Fault Management for WT working under damaged/degraded conditions
The purpose of this research line is the design and development of novel state observers and/or residual generators for wind turbines. Tools from nonlinear control theory will be employed, and robustness will be a fundamental feature to be achieved.
In particular, high gain observers and sliding model based approaches will be focussed on. Fault detection techniques non model-based will be also explored, as a consequence of the well known complexity of the WT model. The critical situation of WTs working in possible icing conditions will be explicitly analysed, due to its relevance for the currently operating WTs.