Mads Mølgaard Pedersen

Dr. Mads Mølgaard Pedersen

orcid

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 August 2023 in Wind Energy Science
Verified
Journal Article
Open access
PDF file
JQJulian Quick
Julian Quick
PRPierre-Elouan Rethore
Pierre-Elouan Rethore
MPMads Mølgaard Pedersen
Mads Mølgaard Pedersen
RRRafael Valotta Rodrigues
Rafael Valotta Rodrigues
MFMikkel Friis-Møller
Mikkel Friis-Møller
It is important to optimize wind turbine positions to mitigate potential wake losses. To perform this optimization, atmospheric conditions, such as the inflow speed and direction, are assigned probability distributions according to measured data, which are propagated through engineering wake models to estimate the annual energy production (AEP). This study presents stochastic gradient descent (SGD) for wind farm optimization, which is an approach that estimates the gradient of the AEP using Monte Carlo simulation, allowing for the consideration of an arbitrarily large number of atmospheric conditions. SGD is demonstrated using wind farms with square and circular boundaries, considering cases with 100, 144, 225, and 325 turbines, and the results are compared to a deterministic optimization approach. It is shown that SGD finds a larger optimal AEP in substantially less time than the deterministic counterpart as the number of wind turbines is increased.

12 November 2020 in Wind Energy Science
Verified
Journal Article
Open access
PDF file
MPMads Mølgaard Pedersen
Mads Mølgaard Pedersen
GLGunner C. Larsen
Gunner C. Larsen
The objective of this paper is to investigate the joint optimization of wind farm layout and wind farm control in terms of power production. A successful fulfilment of this goal requires the following: (1) an accurate and fast flow model, (2) selection of the minimum set of design parameters that rules or governs the problem, and (3) selection of an optimization algorithm with good scaling properties. For control of the individual wind farm turbines with the aim of wind farm production optimization, the two most obvious strategies are wake steering based on active wind turbine yaw control and wind turbine derating. The present investigation is limited to wind turbine derating. A high-speed linearized computational fluid dynamics (CFD) Reynolds-averaged Navier–Stokes (RANS) solver models the flow field and the crucial wind turbine wake interactions inside the wind farm. The actuator disc method is used to model the wind turbines, and utilizing an aerodynamic model, the design space of the optimization problem is reduced to only three variables per turbine – two geometric and one carefully selected variable specifying the individual wind turbine derating setting for each mean wind speed and direction. The full design space is spanned by these (2N+NdNsN) parameters, where N is the number of wind farm turbines, Nd is the number of direction bins, and Ns is the number of mean wind speed bins. This design space is decomposed into two subsets, which in turn define a nested set of optimization problems to achieve a significantly faster optimization procedure compared to a direct optimization based on the full design space. Following a simplistic sanity check of the platform functionality regarding wind farm layout and control optimization, the capability of the developed optimization platform is demonstrated on a Swedish offshore wind farm. For this particular wind farm, the analysis demonstrates that the expected annual energy production can be increased by 4 % by integrating the wind farm control into the design of the wind farm layout, which is 1.2 % higher than what is achieved by optimizing the layout only.

7 February 2024 in Wind Energy Science
Verified
Journal Article
Open access
PDF file
RRRafael Valotta Rodrigues
Rafael Valotta Rodrigues
MPMads Mølgaard Pedersen
Mads Mølgaard Pedersen
JSJens Peter Schøler
Jens Peter Schøler
JQJulian Quick
Julian Quick
PRPierre-Elouan Réthoré
Pierre-Elouan Réthoré
As the use of wind energy expands worldwide, the wind energy industry is considering building larger clusters of turbines. Existing computational methods to design and optimize the layout of wind farms are well suited for medium-sized plants; however, these approaches need to be improved to ensure efficient scaling to large wind farms. This work investigates strategies for covering this gap, focusing on gradient-based (GB) approaches. We investigated the main bottlenecks of the problem, including the computational time per iteration, multi-start for GB optimization, and the number of iterations to achieve convergence. The open-source tools PyWake and TOPFARM were used to carry out the numerical experiments. The results show algorithmic differentiation (AD) as an effective strategy for reducing the time per iteration. The speedup reached by AD scales linearly with the number of wind turbines, reaching 75 times for a wind farm with 500 wind turbines. However, memory requirements may make AD unfeasible on personal computers or for larger farms. Moreover, flow case parallelization was found to reduce the time per iteration, but the speedup remains roughly constant with the number of wind turbines. Therefore, top-level parallelization of each multi-start was found to be a more efficient approach for GB optimization. The handling of spacing constraints was found to dominate the iteration time for large wind farms. In this study, we ran the optimizations without spacing constraints and observed that all wind turbines were separated by at least 1.4 D. The number of iterations until convergence was found to scale linearly with the number of wind turbines by a factor of 2.3, but further investigation is necessary for generalizations. Furthermore, we have found that initializing the layouts using a heuristic approach called Smart-Start (SMAST) significantly reduced the number of multi-starts during GB optimization. Running only one optimization for a wind farm with 279 turbines initialized with SMAST resulted in a higher final annual energy production (AEP) than 5000 optimizations initialized with random layouts. Finally, estimates for the total time reduction were made assuming that the trends found in this work for the time per iteration, number of iterations, and number of multi-starts hold for larger wind farms. One optimization of a wind farm with 500 wind turbines combining SMAST, AD, and flow case parallelization and without spacing constraints takes 15.6 h, whereas 5000 optimizations with random initial layouts, finite differences, spacing constraints, and top-level parallelization are expected to take around 300 years.

14 March 2024 in Wind Energy Science
Verified
Journal Article
Open access
PDF file
JRJavier Criado Risco
Javier Criado Risco
RRRafael Valotta Rodrigues
Rafael Valotta Rodrigues
MFMikkel Friis-Møller
Mikkel Friis-Møller
JQJulian Quick
Julian Quick
MPMads Mølgaard Pedersen
Mads Mølgaard Pedersen
PRPierre-Elouan Réthoré
Pierre-Elouan Réthoré
Wind farm layout optimization is usually subjected to boundary constraints of irregular shapes. The analytical expressions of these shapes are rarely available, and, consequently, it can be challenging to include them in the mathematical formulation of the problem. This paper presents a new methodology to integrate multiple disconnected and irregular domain boundaries in wind farm layout optimization problems. The method relies on the analytical gradients of the distances between wind turbine locations and boundaries, which are represented by polygons. This parameterized representation of boundary locations allows for a continuous optimization formulation. A limitation of the method, if combined with gradient-based solvers, is that wind turbines are placed within the nearest polygons when the optimization is started in order to satisfy the boundary constraints; thus the allocation of wind turbines per polygon is highly dependent on the initial guess. To overcome this and improve the quality of the solutions, two independent strategies are proposed. A case study is presented to demonstrate the applicability of the method and the proposed strategies. In this study, a wind farm layout is optimized in order to maximize the annual energy production (AEP) in a non-uniform wind resource site. The problem is constrained by the minimum distance between wind turbines and five irregular polygon boundaries, defined as inclusion zones. Initial guesses are used to instantiate the optimization problem, which is solved following three independent approaches: (1) a baseline approach that uses a gradient-based solver; (2) approach 1 combined with the relaxation of the boundaries, which allows for a better design space exploration; and (3) the application of a heuristic algorithm, “smart-start”, prior to the gradient-based optimization, improving the allocation of wind turbines within the inclusion polygons based on the potential wind resource and the available area. The results show that the relaxation of boundaries combined with a gradient-based solver achieves on average +10.2 % of AEP over the baseline, whilst the smart-start algorithm, combined with a gradient-based solver, finds on average +20.5 % of AEP with respect to the baseline and +9.4 % of AEP with respect to the relaxation strategy.

1 May 2017 in Journal of Physics: Conference Series
Verified
Conference Paper
Open access
PDF file
TLT.J. Larsen
T.J. Larsen
GLG.C. Larsen
G.C. Larsen
MPMads Mølgaard Pedersen
Mads Mølgaard Pedersen
KEK. Enevoldsen
K. Enevoldsen
HMHenrik Madsen
Henrik Madsen
This paper presents a comparison between measured and simulated tower loads for the Danish offshore wind farm Nysted 2. Previously, only limited full scale experimental data containing tower load measurements have been published, and in many cases the measurements include only a limited range of wind speeds. In general, tower loads in wake conditions are very challenging to predict correctly in simulations. The Nysted project offers an improved insight to this field as six wind turbines located in the Nysted II wind farm have been instrumented to measure tower top and tower bottom moments. All recorded structural data have been organized in a database, which in addition contains relevant wind turbine SCADA data as well as relevant meteorological data-e.g. wind speed and wind directionfrom an offshore mast located in the immediate vicinity of the wind farm. The database contains data from a period extending over a time span of more than 3 years. Based on the recorded data basic mechanisms driving the increased loading experienced by wind turbines operating in offshore wind farm conditions have been identified, characterized and modeled. The modeling is based on the Dynamic Wake Meandering (DWM) approach in combination with the state-of-the-art aeroelastic model HAWC2, and has previously as well as in this study shown good agreement with the measurements. The conclusions from the study have several parts. In general the tower bending and yaw loads show a good agreement between measurements and simulations. However, there are situations that are still difficult to match. One is tower loads of single-wake operation near rated ambient wind speed for single wake situations for spacing's around 7-8D. A specific target of the study was to investigate whether the largest tower fatigue loads are associated with a certain downstream distance. This has been identified in both simulations and measurements, though a rather flat optimum is seen in the measurements.

20 November 2017 in Wind Energy Science
Verified
Journal Article
Open access
PDF file
MPMads Mølgaard Pedersen
Mads Mølgaard Pedersen
TLTorben J. Larsen
Torben J. Larsen
HMHenrik Madsen
Henrik Madsen
GLGunner Chr. Larsen
Gunner Chr. Larsen
In this paper an alternative method to evaluate power performance and loads on wind turbines using a blade-mounted flow sensor is investigated. The hypothesis is that the wind speed measured at the blades has a high correlation with the power and loads such that a power or load assessment can be performed from a few hours or days of measurements.In the present study a blade-mounted five-hole pitot tube is used as the flow sensor as an alternative to the conventional approach, where the reference wind speed is either measured at a nearby met mast or on the nacelle using lidar technology or cup anemometers. From the flow sensor measurements, an accurate estimate of the wind speed at the rotor plane can be obtained. This wind speed is disturbed by the presence of the wind turbine, and it is therefore different from the free-flow wind speed. However, the recorded wind speed has a high correlation with the actual power production as well as the flap-wise loads as it is measured close to the blade where the aerodynamic forces are acting.Conventional power curves are based on at least 180 h of 10 min mean values, but using the blade-mounted flow sensor both the observation average time and the overall assessment time can potentially be shortened. The basis for this hypothesis is that the sensor is able to provide more observations with higher accuracy, as the sensor follows the rotation of the rotor and because of the high correlation between the flow at the blades and the power production. This is the research question addressed in this paper.The method is first tested using aeroelastic simulations where the dependence of the radial position and effect of multiple blade-mounted flow sensors are also investigated. Next the method is evaluated on the basis of full-scale measurements on a pitch-regulated, variable-speed 3.6 MW wind turbine.It is concluded that the wind speed derived from the blade-mounted flow sensor is highly correlated with the power and flap-wise bending moment and that the method has advantages over the traditional approach where the met-mast wind speed is used as reference, e.g. the capability of measuring the shear, veer and turbulence. The aeroelastic simulations show that the assessment time can be reduced, but this reduction cannot be confirmed from the current measurement database due to sensor problems and practical circumstances. Measuring the wind speed at the rotor plane comes with a price as the wind speed is affected by the induction which may be sensitive to the changes you want to evaluate, e.g. different vortex generator configurations. Furthermore it is concluded that a robust instrument and measurement system is required to obtain accurate and reliable wind speed recordings from pitot-tube measurements.