Joint Project TurboWind

  Copyright: © W2E Wind to Energy, 2020

Adaptive robust control of wind turbines based on estimation of turbulence intensity

 

Motivation

Wind energy is the key technology for converting power generation in Germany to renewable energies. With increasing market share and reduced subsidies for wind power, their electricity generation costs will have a significant influence on consumers' electricity prices in the future. One lever for reducing the LCOE of wind turbines is to reduce the amount of material used and to increase the service life, in particular by reducing the dynamic mechanical load on the rotor and tower through control engineering interventions. Conventional control strategies will therefore no longer be sufficient due to further regulatory requirements to be expected for wind turbines in the future, such as grid-supporting operation.

One solution is a modern, model predictive control (MPC), in which several criteria are simultaneously weighted within a cost function and thus optimal manipulated variables are specified depending on the operating point. In this control methodology, a dynamic, mathematical model of the plant is used to predict the behavior and thus the future behavior of the control variables within the control system. This provides the possibility to act proactively on future conditions. Likewise, direct consideration of operating constraints within the MPC is possible. Furthermore, the wind situation can be explicitly considered in the control criteria, which facilitates the development of difficult sites where, for example, a highly fluctuating or generally weak wind prevails.

Compared to conventional control strategies, which have been used all over the world for several decades, this corresponds to a technical revolution. However, due to the completely new development, acceptance in the industry can only be achieved through experimental demonstrations in order to outweigh the high technological and thus economic risks.

 

Project Goals and Methods

The overall objective of the project is to enable the prototypical use of modern model-based control methods on real wind turbines.

In this project, extensive practical tests will be carried out to investigate the effects of the individual target criteria on the control system and how heavily these can be weighted in each case so that robustly stable behavior results. For this purpose, a prototype wind turbine from W2E with a capacity of 3 MW will be used and a corresponding concept for safe operation management during the control tests will be developed. Based on the preliminary results, an adaptive, robust control method is implemented which covers the complete operating range of current wind turbines. Subsequently, the control approach is analyzed by means of iterative tests, first simulatively and finally experimentally on a real wind turbine.

This procedure allows a direct comparison between classical and optimized MPC-based control. Furthermore, the adaptive control concept and the turbulence estimation can be optimized and improved based on the experimental measurement data, so that the control concept covers all relevant operating ranges.

 

Innovations and Perspectives

The following scientific and technological sub-objectives are targeted in this project:

• Development of improved structural dynamic models to achieve higher prediction accuracy even under suboptimal environmental conditions.
• Extension of an existing MPC to an adaptive, robust multi-criteria control method for the complete operating range of a wind turbine (full load, part load and transition range)
• Optimization and extension of an estimation of the current operating situation, in particular the time and spatially resolved wind speed and turbulence intensity
• Design and implementation of a hardware-in-the-loop test bench based on a simulation computer and industrial controllers (PLC) to realize a safe test environment and to ensure a consistent controller design procedure
• Conceptual design of a safety and operational management concept for the safe execution of experimental control tests on a wind turbine and for subsequent productive use
• Based on this, integration of the control algorithms on an existing prototype turbine with 3 MW and integration of a corresponding sensor concept for data processing
• Execution of experimental measurement campaigns for optimization and evaluation of the improvement potential with regard to the necessary model depth, parameterization, weighting factors and required sensor technology.

Following the project, an adaptive control concept based on turbulence estimation is available, which can be used for retrofitting existing plants as well as for new developments. Furthermore, the findings of a detailed evaluation of experimental control tests and corresponding comparisons can be used to improve the accuracy of future simulation studies.

 

Project partner
W2E Wind to Energy GmbH