06 Fakultät Luft- und Raumfahrttechnik und Geodäsie
Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/7
Browse
11 results
Search Results
Item Open Access Nonlinear model predictive control of floating wind turbines(2013) Schlipf, David; Sandner, Frank; Raach, Steffen; Matha, Denis; Cheng, Po WenIn this work a nonlinear model predictive control method for a floating wind turbine is presented. A reduced nonlinear model including disturbance preview of wind and waves is derived and implemented to compute optimal input trajectories for collective pitch and the generator torque. A cost functional is introduced which fulfills all desired constraints and controller goals for above rated wind conditions. The controller is tested for extreme and fatigue load cases and a significant reduction of the power and rotor speed deviations is obtained. Furthermore, ultimate tower loads and damage equivalent loads on shaft and blades are decreased. Although more detailed testing is necessary, this preliminary results show the advantages of nonlinear model predictive control for floating wind turbines.Item Open Access Nonlinear model predictive control of floating wind turbines with individual pitch control(2014) Raach, Steffen; Schlipf, David; Sandner, Frank; Matha, Denis; Cheng, Po WenIn this work a nonlinear model predictive controller with individual pitch control for a floating offshore wind turbine is presented. An aerodynamic model of the collective pitch control approach is extended by describing pitching and yawing moments based on rotor disk theory. This extension is implemented in a reduced nonlinear model of the floating wind turbine including disturbance preview of wind speed, linear vertical and horizontal wind shear, and wave height to compute optimal input trajectories for the individual pitch control inputs and the generator torque. An extended cost functional for individual pitch control is proposed based on the collective pitch control approach. The controller is evaluated in aero-servo-hydro-elastic simulations of a 5MW reference wind turbine disturbed by a three-dimensional stochastic turbulent wind field. Results show a significant blade fatigue load reduction compared to a baseline controller through minimizing yawing and pitching moments on the rotor hub while maintaining the advantages of the model predictive control approach with collective pitch control.Item Open Access Lidar-based wake tracking for closed-loop wind farm control(2017) Raach, Steffen; Schlipf, David; Cheng, Po WenThis work presents two advancements towards closed-loop wake redirection of a wind turbine. First, a model-based wake-tracking approach is presented, which uses a nacelle-based lidar system facing downwind to obtain information about the wake. The method uses a reduced-order wake model to track the wake. The wake tracking is demonstrated with lidar measurement data from an offshore campaign and with simulated lidar data from a simulation with the Simulator fOr Wind Farm Applications (SOWFA). Second, a controller for closed-loop wake steering is presented. It uses the wake-tracking information to set the yaw actuator of the wind turbine to redirect the wake to a desired position. Altogether, the two approaches enable a closed-loop wake redirection.Item Open Access Lidar-assisted wake redirection control(2019) Raach, Steffen; Cheng, Po Wen (Prof. Dr.)Wind energy has developed to a competitive energy source and is intended to play an important role in the world-wide sustainable energy supply. The size of wind turbines has tremendously increased and turbines have been clustered to wind farms in order to share infrastructure and moreover reach energy production capacities of conventional power plants. However, the operation of wind turbines in a wind farm hasn't changed compared to single turbine operation. Possible interactions between wind turbines through the wind are not considered in the operation strategy of the turbines. The wake of a wind turbine can negatively influence the operation of a downwind turbine because the wind speed is reduced and the turbulence intensity is increased in the wake. Currently, each wind turbine is maximizing its power output independently of the other wind turbines. In a total consideration this may result in a suboptimal power output due to the interaction of wind turbines. In the case a wake impinges a second wind turbine, that turbine produces less power and the structural loads may increase. Both effects negatively impact the operation of the wind turbine and therefore it makes sense to avoid wake interactions. This task demands new wind park control concepts which take wind turbine interactions into account. In current research activities the wind farm is treated as the total system. Different investigations in wind farm control have identified two promising operation concepts: axial induction control and wake redirection control. Whereas the concept of induction control tries to minimize the impact of wakes on other wind turbines, the basic idea of wake redirection control is to redirect the wake of a wind turbine by yawing the wind turbine and therefore forcing a misalignment to the wind direction. Using this technique, interactions between wind turbines can be minimized. The current concept is based on an open-loop methodology in which a reduced-order wake model is used to pre-calculate the optimal yaw angles. This results in two disadvantages: The uncertainty which a simplified model introduces and the missing possibility to react to disturbances. This work introduces the concept of lidar-based closed-loop wake redirection control which can adapt to uncertainties and react on disturbances. Therefore it extends the concept of wake redirection control with a new closed-loop methodology. This work contributes various aspects to enable a lidar-based closed-loop wake redirection control. It first presents the general concept. Then it is separated in two subtasks: the measurement and the control tasks. This separation helps to focus on the specific questions of each task. First the measurement task is addressed and solutions are provided to process lidar measurement data to a useful signal for the wake redirection controller. Different methodologies are presented to track the wake position using lidar measurement data and the concept of model-based wake tracking is described in detail. Afterwards the control task is considered. Three different controller synthesis concepts are applied to wake redirection and controllers are synthesized. The different controllers are analyzed and the performances are assessed. Then the controllers are verified in different simulation tools. Mainly simulations are performed with a medium-fidelity computational fluid dynamics simulation tool. In addition the concept is implemented in a Large-Eddy simulation tool to demonstrate the adaption to disturbances and model uncertainties. Altogether the work introduces lidar-based closed-loop wake redirection control. It demonstrates the feasibility of the concept as well as the adaptivity of the controller to model uncertainties and disturbances. The different aspects of the concept are considered and methodologies for wake position estimation are provided and controllers are designed. Finally, recommendations are given to realize the concept in reality, and open questions are highlighted which require deeper investigations.Item Open Access Optimization of a feed-forward controller using a CW-lidar system on the CART3(2015) Haizmann, Florian; Schlipf, David; Raach, Steffen; Scholbrock, Andrew; Wright, Alan; Slinger, Chris; Medley, John; Harris, Michael; Bossanyi, Ervin; Cheng, Po WenThis work presents results from a new field-testing campaign conducted on the three-bladed Controls Advanced Research Turbine (CART3) at the National Renewable Energy Laboratory in 2014. Tests were conducted using a commercially available, nacelle-mounted continuous-wave lidar system from ZephIR Lidar for the implementation of a lidar-based collective pitch feed-forward controller. During the campaign, the data processing of the lidar system was optimized for higher availability. Furthermore, the optimal scan distance was investigated for the CART3 by means of a spectra-based analytical model and found to match the lidar's capabilities well. Throughout the campaign the predicted correlation between the lidar measurements and the turbine's reaction was confirmed from the measured data. Additionally, the baseline feedback controller's gains were tuned based on a simulation study that included the lidar system to achieve further load reductions. This led to some promising first results, which are presented at the end of this paper.Item Open Access FlexDyn - a new OpenFAST structural dynamics module for a general, user defined wind turbine topology(2021) Lemmer, Frank; Pérez Brovia, Santiago; Skandali, Danai; Raach, SteffenIn the present work, FlexDyn, a new structural dynamics module for the OpenFAST framework is developed. FlexDyn can generate structural equations of motion through a formalism, given user-defined rigid and elastic bodies and associated Degrees of Freedom (DOFs). The Newton-Euler formalism uses beam models with shape functions for a reduced-order representation in the same way as ElastoDyn of OpenFAST. The equations of motion are formulated in minimal coordinates, equally to ElastoDyn. FlexDyn is fully integrated into the OpenFAST framework with a coupling to AeroDyn and the new SubDyn module for FE representations of floating substructures (Jonkman, et al., 2020), among others. The formalism was previously implemented and verified in the low-order aero-hydro-servo-elastic code SLOW (Lemmer, et al., 2020). The objective of the presentation is to show the methodology of the formalized generation of equations of motion and first results of the new FlexDyn module for OpenFAST. The use case is an improved aero-elastic model, which includes the torsional DOF of the blades. The torsional DOF is not included in the ElastoDyn module but can potentially contribute to the motion and load response of the blades. The fidelity level of this use case of FlexDyn is higher than that of ElastoDyn but still below that of BeamDyn, which is a full FE representation of the blades. For this reason, the computational performance is still in the range of ElastoDyn, taking advantage of the order reduction.Item Open Access Multi-variable feedforward control for floating wind turbines using lidar(2020) Schlipf, David; Lemmer, Frank; Raach, SteffenIn this work a multi-variable feedforward controller for floating wind turbines is presented. The feedforward controller provides a pitch rate and a torque update to a conventional feedback controller based on a wind speed preview. A 10 MW reference wind turbine is used on a semi submersible floating platform to study the potential of the controller. An open-source simulation tool is extended with an realistic lidar simulator and the lidar data processing, feedforward controller, and feedback controller are implemented in modular setup. The lidar measurements are fully motion compensated and combined to provide a preview of the rotor-effective wind speed to the controller. The feedforward controller is designed to minimize structural loads and to decrease the platform pitch motion. In verification and simulation studies the concept is demonstrated and the multi-variable feedforward controller shows a promising improvement in speed regulation and load reduction on the floating wind turbine.Item Open Access Prospects of linear model predictive control on a 10MW floating wind turbine(2015) Lemmer, Frank; Raach, Steffen; Schlipf, David; Cheng, Po WenThe presented research has the objective of supporting the integrated conceptual design of floating offshore wind turbines (FOWT). The dynamics of the multidisciplinary coupled system with the aerodynamics, hydrodynamics, structural dynamics, the catenary mooring lines and the controller shall be represented in simulation models adapted to the current design stage. Here, a linear model-predictive controller (MPC) as an optimal multiple input-multiple output (MIMO) controller is designed for a novel concept of the floating foundation for a 10MW wind turbine. The performance of this controller is easily adjustable by a cost function with multiple objectives. Therefore, the MPC can be seen as a benchmark controller in the concept phase, based on a simplified coupled simulation model with only approximate model information. The linear model is verified against its nonlinear counterpart and the performance of the MPC compared to a SISO PI-controller, which is also designed in this work. The developed models show to be well suited and the linear MPC shows a reduction of the rotor speed overshoot and tower bending from a deterministic gust.Item Open Access Advances on reduced-order modeling of floating offshore wind turbines(2021) Lemmer, Frank; Yu, Wei; Steinacker, Heiner; Skandali, Danai; Raach, SteffenAero-hydro-servo-elastic modeling of Floating Offshore Wind Turbines (FOWTs) is a key component in the design process of various components of the system. Different approaches to order reduction have been investigated with the aim of improving structural design, manufacturing, transport and installation, but also the dynamic behavior, which is largely affected by the blade pitch controller. The present work builds on previous works on the SLOW (Simplified Low-Order Wind Turbine) code, which has already been used for the above purposes, including controller design. While the previous rigid rotor model gives good controllers in most cases, we investigate in the present work the question if aero-elastic effects in the design model can improve advanced controllers. The SLOW model is extended for the flapwise bending and coupled to NREL's AeroDyn, linearized and verified with the OlavOlsen OO-Star Wind Floater Semi 10MW public FOWT model. The results show that the nonlinear and linear reduced-order SLOW models agree well against OpenFAST. The state-feedback Linear Quadratic Regulator (LQR) applied with the same weight functions to both models, the old actuator disk, and the new aero-elastic model shows that the LQR becomes more sensitive to nonlinear excitation and that the state feedback matrix is significantly different, which has an effect on the performance and potentially also on the robustness. Thus modeling uncertainties might even be more critical for the LQR of the higher-fidelity model.Item Open Access Shadow effects in an offshore wind farm - potential of vortex methods for wake modelling(2015) Beyer, Friedemann; Luhmann, Birger; Raach, Steffen; Cheng, Po WenOffshore wind turbines in a wind farm are affected by wakes of upstream turbines and adjacent wind farms depending on the park layout and wind direction. As a result the power output may decrease, while structural loads are increasing. In this research a coupled numerical approach based on multi-body system and free vortex methods is used to simulate shadow effects on the Alpha Ventus wind farm. The AV5 is operating at 12 m/s wind speed at half wake conditions with enabled control system and flexible blades and tower. Results of power output, rotor speed, blade pitch and blade root moment over time and azimuth demonstrate the high impact of the half wake condition on the wind turbine performance and loads.