Design Tool for Wind Turbine Control Algorithms

The modular Control Design Tool is written in Matlab® language and uses ... Based on input data and intended control targets, design modules (Matlab® ...Missing:
1MB taille 31 téléchargements 418 vues
Design Tool for Wind Turbine Control Algorithms Advanced wind turbine control algorithms have become more important over the last years in order to deal with the high requirements to wind turbines. An open source design tool enables possibilities to develop industrial control algorithms and to utilize the benefits of more advanced control solutions. The modular design tool offers design and evaluation of control algorithms, linking to aero-elastic codes and implementation into the turbine controller.

WIND

RECORD

CONTROLLER

WIND TURBINE

Control Design Tool The objectives of the Control Design Tool project, as launched by ECN Wind Energy in 2000, is stimulating the use of advanced control algorithms in wind turbines and making the required theory, the design and the evaluation of the algorithms more accessible. The modular Control Design Tool is written in Matlab® language and uses the graphical toolbox Simulink® for time domain simulation. This enables possibilities to modify or add design modules easily (’open source approach’). Finally, RealTime Workshop® is able to compile the control code to DLL or dedicated hardware.

Design procedure As shown in figure 1, the design procedure consists of three sequential design activities.

Figure 1: Design procedure for control algorithms Based on input data and intended control targets, design modules (Matlab® m-code) process the input data and derive additional data to well-structured and unambiguous data definitions. During this activity, logging messages and verification plots are generated for observation, assessment and reporting purposes. Based on (linear) control analysis within the design modules, the capability of actuators and sensors are proved, even as control loop stability and robustness. Time domain simulations will evaluate the behaviour of the control structure. Both the (discrete) control structure and the (continuous) wind turbine model are implemented as simulation modules in Simulink®, even as a stochastic wind signal. For more detailed analysis (load set calculation, certification), the control algorithm should be linked to an aero-elastic code. This conversion step is fully automated by using RealTimeWorkshop® Finally, the parameterised control algorithm should be converted towards an, often dedicated, code for implementation in the real-time turbine controller. The design tool will facilitate this conversion to a high degree¯ Design and evaluation models Currently, the design tool comprises generic modules for the design of classic feedback control algorithms, for a variable speed turbine by means of collective pitch actuation and generator torque control. Additional modules are available for improved damping of the drive train and power optimisation (figure 2). Tower damping features are under consideration.

J

Figure 2: Top level scheme of wind turbine control The design tool modules have been applied for the development of several industrial control algorithms. The design results have been validated successfully, both in practice by implementation in a modern industrial turbine and by simulations with the aero-elastic code PHATAS. Results Simulation results of a typical multi MW wind turbine are shown in figure 3, around rated wind speed (max. power) and at high wind speed operation (nom. power).

~ {r~n]

~ {rnm]

Figure 3: Simulation results Future developments Extensions for the design of individual pitch control algorithms are being developed. These algorithms allow for much better load reductions. Points of departure are the integrated linear wind turbine models, which are obtained from (reduced) multi body models of coupled dynamics by the analysis tool ’Frequency Domain Analysis of Offshore Wind Turbines (TURBU)’. These developments will be added in shod-term to the design tool as design modules and evaluation models.

ECN Wind Energy P.O. Box 1 NL-1755 ZG Petten

Tel: +31 224 564913 Fax: +31 224 568214

Contact: E.L. van der Hooft vanderhooft @ecn.nl