Wind Speed Estimation Research Articles

WITHDRAWN: Nonlinear Kalman filters based rotor effective wind speed estimation: An experimental comparison

This article has been withdrawn at the request of the editor. The Publisher apologizes for any inconvenience this may cause. The full Elsevier Policy on Article Withdrawal can be found at http://www.elsevier.com/locate/withdrawalpolicy .

Adaptive Interval Type-2 Fuzzy Controller for Variable-speed Wind Turbine

In this paper, an adaptive interval type-2 fuzzy controller is proposed for variable-speed and variable-pitch wind turbines. Because of attractive features of the well-known wind turbine baseline controller, the proposed controller acts as an augmented controller and works in parallel to the baseline controller. As typical variable-speed wind turbines have different controllers for different operation regions, for each operation region, a dedicated interval type-2 fuzzy controller is designed. Because of the uncertainty in wind speed measurement, modern control techniques try to estimate this value. However, in contrast to these modern control techniques, the proposed controller is independent of the wind speed estimation. Thus, there is a better saving in cost and computational burden. To evaluate the effectiveness of the proposed controller, simulations are conducted with wind profiles which span all operation regions. Results show that, compared with the baseline controller, the proposed controller enhances power generations and reduces mechanical loads concurrently.

A SAR-Based Parametric Model for Tropical Cyclone Tangential Wind Speed Estimation

The tangential wind speed increases from the center to the eyewall of tropical cyclones (TC) along the radial direction and begins to decay when it extends outward. The tangential wind profile model is one of the most effective and widely used methods to reconstruct the TC radial wind speed. This paper proposes a parametric tangential wind profile (TWP) model based on high-spatial-resolution SAR imagery. The new model functions are piecewise with maximum tangential wind speed as a threshold, and all of them are designed as nonlinear. Notably, the derivative at the segmentation threshold is zero to ensure a smooth transition of the estimated wind speed profile. With the SAR-derived azimuth-averaged wind speed, we can determine the model parameters and get the tangential wind speed. The TWP model outperforms the commonly used SMRV model, as it better resolves the tangential wind profile shape as depicted by both SAR-derived winds and hurricane hunter Stepped-Frequency Microwave Radiometer (SFMR) derived winds. A comprehensive analysis of the TWP model parameters is carried out by fitting tangential winds for 620 hurricane hunter flights. Interestingly, the tangential wind profiles for major hurricanes show a similar shape. The proposed TWP model can be used for improved TC characterization and forecasting purposes.

Learning Physics Property Parameters of Fabrics and Garments With a Physics Similarity Neural Network

Predicting the physics properties of deformable objects such as garments and fabrics is a challenge in robotic research. Directly measuring their physics properties in a real environment is difficult [1]. Therefore, learning and predicting the physics property parameters of garments and fabrics can be conducted in simulated environments. However, garments have collars, sleeves, pockets and buttons that change the way garments deform and simulating these is time-consuming. Therefore, in this paper, we propose to predict the physics parameters of real fabrics and garments by learning the physics similarities between simulated fabrics via a Physics Similarity Network (PhySNet). For this, we estimate wind speeds generated by an electric fan and area weights to predict the bending stiffness parameters of real fabrics and garments. We found that PhySNet coupled with a Bayesian optimiser can predict physics property parameters and improve state-of-art by 34.0% for fabrics and 68.1% for garments.

Wind Speed Extraction From First-Order Sea Echoes Using a Small-Aperture Multifrequency High-Frequency Radar

Wind speed inversion is a challenging work in the field of ocean surface remote sensing with high-frequency (HF) radars. Recently, the wind speed inversion method based on the first-order HF radar sea echoes attracts much attention. However, most methods, which use radar data at a fixed operating frequency, provide a limited range for wind speed measurement. To overcome the drawback, a wind speed inversion method based on radar data collected with a multifrequency HF radar is proposed. This new method first fits the relationship between the wind speed and the power of the broad-beam first-order HF radar sea echoes and then combines the fit models with the multifrequency HF radar data to estimate wind speed. The proposed method makes use of the information contained in the first-order sea echoes of various operating frequencies so as to improve the performance of the wind speed measurements. Finally, a comparison between radar-estimated and anemometer-measured wind speeds is made to validate the proposed method. Compared with the anemometer data, the wind speeds estimated by the proposed method have a root-mean-square error (RMSE) of 2.27 m/s.

Estimating the velocity of pyroclastic density currents using an operational dual-PRF radar

Pyroclastic density currents are one of the deadliest hazards produced by a volcano. Understanding their dynamics and generation mechanisms is critical for developing better hazard mitigation strategies. This study presents a method for retrieving velocity profiles across a natural moving PDC, applied here to a PDC generated by collapsing column during the eruption of Sinabung Volcano, Indonesia, on 19 February 2018 at the onset time of 08:53. We used an operational dual Pulse Repetition Frequency (PRF) weather radar, located ~7.8 km to the SE of the volcano, to estimate the velocity profile components of the volcanic plume: updraft, fallout, and horizontal advection. Doppler radar data was post-processed by applying two different filters: median and Laplacian, to correct errors associated with dealiased Doppler velocities. The Laplacian filter method was more effective in correcting the dealiasing errors by producing a more continuous velocity field without over smoothing its values. Following the dealiasing process, the velocity profile components were retrieved according to radar parameters such as Doppler velocities, copolar correlation, and reflectivity intensity factor. Initially, the pyroclastic clast was released at a lower exit velocity of ~120 m/s (84 s after the onset). A maximum of ~190 m/s exit velocity was then observed at 08:57:52 (292 s after onset). Lower exit velocity in the initial phase and less than 10 m/s estimated wind speed are the main factors causing the partial collapse of the plume at ~2.5 km height above the vent. The part of the collapsing column was associated with a more than 50 dBZ reflectivity intensity factor of fallout velocity exceeding −50 m/s at 126 s after the onset. Dilute PDCs were observed until 09:09:34 (994 s after onset), moving downslope at SE sector at a maximum velocity of −84 m/s (i.e., in the direction of the radar) . The extracted velocity components are essential parameters in the numerical model of PDCs and tephra dispersal, enforcing the benefit of weather radar to complement the remote monitoring system of volcanic hazards.

A data-driven approach to support voltage profiles & loss reduction in wind generator integrated active distribution network considering solid-state transformers with twofold reactive power compensation

ABSTRACT The paper proposes an optimal planning strategy for an active distribution network (ADN) with wind energy-based distributed generation (WDG) by considering the twofold reactive power compensation (TRPC) feature of a three-stage type solid-state transformer (SST). A model for wind speed and load demand estimation for each hour of the day is formulated by using the partition around medoids (PAM) technique over an annual database of load demand and wind speed. A multi-objective mixed-integer non-linear programming (MINLP) approach for optimally locating and sizing the SST installations in the radial distribution network (RDN) is presented to address the dual objectives of voltage profile improvement (VPI) and energy loss reduction (ELR). Along with the SST parameters, the locations and number of wind turbine (WT) units to be integrated into the system are treated as optimization variables. Furthermore, the distribution power flow accounts for the distribution transformer’s (DT) operating losses as well as the approximate losses of the SST. The presented approach is tested on a 33-bus RDN and the results are reported for multiple case studies. The programming was developed in the MATLAB R2020a environment and the paretosearch algorithm (PSA) is used to address the MINLP problem. Other standard multi-objective optimization algorithms, including multi-objective multi-verse optimization (MOMVO), non-dominated sorting genetic algorithm (NSGA-II), and multi-objective salp swarm algorithm (MSSA) are used to compare the performance of the proposed method. The outcomes of the peak hour evaluation, with a 20% over-rating, show the active and reactive power demand on the MV side of the substation to have decreased by 56.3% and 30.1% respectively. The active line loss has been lowered by 76.95% and the reactive line loss has decreased by 76.14%. The absolute minimum voltage has increased by 7.314%. Furthermore, according to the annual technical evaluation, there was a 29.42% reduction in active energy served by the DTs and SSTs.

Model-based wind turbine control design with power tracking capability: A wind-tunnel validation

Owing to the aerodynamic conversion process, wind turbines exhibit a nonlinear behavior and encounter turbulent operating conditions that demand well-defined closed-loop dynamics to withstand accumulated loading over the lifetime. Wind excitation is the main disturbance and driving force for the system and determines the necessary operating strategy, but it usually represents an unmeasurable quantity. In this study, we used a linear-matrix-inequalities-based control and observer design to operate a variable-speed, variable-pitch wind turbine in a wind tunnel experiment at different reproducible inflow conditions while relying on a wind speed estimate obtained from a disturbance observer. The computational complexity of the stability framework incorporating the reconstruction of the unknown wind speed is reduced by exploiting characteristics of the modeling approach based on a convex combination of linear submodels. The assumption used in the proposed stability consideration is evaluated based on measurement data. We introduced an extended operating range compared to the commonly considered operating trajectory of wind turbines in the control design. A controller based on Takagi–Sugeno modeling is used to operate the turbine at challenging power tracking requirements demonstrating the capability to support fast stabilization of the electrical grid while discussing the loading and operational constraints observed during the experiments.

Constrained power reference control for wind turbines

AbstractThe cost of wind energy can be reduced by controlling the power reference of a turbine to increase energy capture, while maintaining load and generator speed constraints. We apply standard torque and pitch controllers to the direct inputs of the turbine and use their set points to change the power output and reduce generator speed and blade load transients. A power reference controller increases the power output when conditions are safe and decreases it when problematic transient events are expected. Transient generator speeds and blade loads are estimated using a gust measure derived from a wind speed estimate. A hybrid controller decreases the power rating from a maximum allowable power. Compared to a baseline controller, with a constant power reference, the proposed controller results in generator speeds and blade loads that do not exceed the original limits, increases tower fore‐aft damage equivalent loads by 1%, and increases the annual energy production by 5%.

Site or regional design wind speeds?

Estimates of high return period design wind speeds are required to achieve an acceptable probability of failure. These can be found by fitting an appropriate distribution to observed data but is complicated by spatial variation of the wind climate, therefore it is unclear which data are relevant at a particular location. We aim to frame this problem in terms of model selection and the bias variance trade-off. Existing models for incorporating site versus regionally averaged statistics are expressed in terms of key parameters, which allow the errors associated with each model to be derived. It is found that using a characteristic wind speed in codification combines the site statistics (low bias) with regionally averaged statistics (low variance). The return period of the characteristic wind speed is shown to act as a parameter controlling the relative weighting of these models. This insight is used to develop an optimal (minimum mean square error) estimator of the design wind speed, which is shown to vary based on both the available quantity of data at a particular site and how well this data corresponds to the regional average. Some practical advantages of this optimal model are then demonstrated at several South African stations.

Investigation of Machine Learning Using Satellite-Based Advanced Dvorak Technique Analysis Parameters to Estimate Tropical Cyclone Intensity

AbstractSeveral simple and computationally inexpensive machine learning models are explored that can use advanced Dvorak technique (ADT)-retrieved features of tropical cyclones (TCs) from satellite imagery to provide improved maximum sustained surface wind speed (MSW) estimates. ADT (version 9.0) TC analysis parameters and operational TC forecast center best track datasets from 2005 to 2016 are used to train and validate the various models over all TC basins globally and select the best among them. Two independent test sets of TC cases from 2017 to 2018 are used to evaluate the intensity estimates produced by the final selected model called the “artificial intelligence (AI)” enhanced advanced Dvorak technique (AiDT). The 2017–18 MSW results demonstrate a global RMSE of 7.7 and 8.2 kt (1 kt ≈ 0.51 m s−1), respectively. Basin-specific MSW RMSEs of 8.4, 6.8, 7.3, 8.0, and 7.5 kt were obtained with the 2017 dataset in the North Atlantic, east/central Pacific, northwest Pacific, South Pacific/south Indian, and north Indian Ocean basins, respectively, with MSW RMSE values of 8.9, 6.7, 7.1, 10.4, and 7.7 obtained with the 2018 dataset. These represent a 30% and 23% improvement over the corresponding ADT RMSE for the 2017–18 datasets, respectively, with the AiDT error reduction significant to 99% in both sets. The AiDT model represents a notable improvement over the ADT performance and also compares favorably to more computationally expensive and complex machine learning models that interrogate satellite images directly while still preserving the operational familiarity of the ADT.

Wind lidar signal denoising method based on singular value decomposition and variational mode decomposition.

A denoising method based on singular value decomposition (SVD) and variational mode decomposition (VMD) is proposed for wind lidar. Utilizing the covariance matrix based lidar signal simulation model, the performance of VMD, SVD, and VMD-SVD is evaluated. The results show that the VMD-SVD method is of better performance, and the output signal-to-noise ratio (SNR) is about 12 dB at the input SNR of -9dB. The actual lidar signals processing is performed with this combined denoising method, and the detection range and wind speed at pulse accumulation numbers of 50,100, and 300 are compared. We set the wind speed resulting from noisy signal with pulse accumulation number of 300 as the reference wind speed, and the mean value and standard deviation of wind differences are analyzed. The results show that the denoising method can not only increase the detection range while ensuring the accuracy of wind speed estimation but also achieve the same detection distance with fewer pulse accumulations, thereby improving the temporal resolution. For the pulse accumulation number of 50, the detection range is extended to 24 km from 18.45 km, and the standard deviation of speed difference is 0.88 m/s; for the same detection range, the temporal resolution is increased by about 6 times.

Novel rotor effective wind speed estimation method for light detection and ranging application

The ‘Cyclops’ effect of the light detection and ranging (LiDAR) system means that the LiDAR system can only measure wind speed at one point in space at a certain moment. However, for the purposes of wind turbine control, considering the wind information at the rotor plane with time stamps can be more precise. The wind speed attenuates due to the induction zone; however, Taylor's frozen hypothesis is commonly used in the application of LiDAR for the rotor effective wind speed estimation. The wind turbine is controlled directly by LiDAR measured data with constant time delay, which weakens LiDAR's effectiveness. In this paper, a novel LiDAR data pre-processing method is demonstrated. The Medici induction zone model is further refined with initial wind speed and different measured distances for accurate estimation. Furthermore, it makes online variable time delay of LiDAR data computation possible with the solution of the separated differential equation. Finally, the novel method is fully verified through field test results. The results show that the initial wind speed affected the attenuation factor within 1.4 times rotor radius and the error between the experimental value and the theoretical value is within 0.1 m/s.

Off-season lee wave cloud over the Arsia Mons in Mars: A study based on Mars Colour Camera (MCC)

The present analysis reported the lee wave interpretation on Arsia Mons Elongated Cloud (AMEC) that used to appear over the lee side of the Arsia Mons. The present work also reported the wavelength, the cloud's height, and wind speed of the observed lee wave structure. The estimated wind speed is 86 ± 4.9 m/s, with a wavelength of 60 ± 0.3 km at an altitude of 55 ± 7 km. This is the largest lee wave structure that appears during the dust storm season (Ls = 230 to Ls = 300) and strongly contributes to the planet's Albedo level. Calculated TOA reflectance varies from 0.05 to 0.15 for the blue channel and 0.02 to 0.09 for the red channel. Estimated AOD ranges from 1.8 (red channel) to 1.5 (blue channel), indicating the contribution of dust and water ice crystal. The estimated angstrom exponent (α) value signifies coarse mode particle presence in the cloud with an effective radius of 3.2 μm. The presence of water ice crystal contributes to the albedo level's increment to 0.8 and signifies the formation temperature for AMEC to be around 190 K.

Wind Energy Potential Assessment Based-on WRF Four-Dimensional Data Assimilation System and Cross-Calibrated Multi-Platform Dataset

Indonesia has a target of achieving 23% of renewable energy share in the total energy mix in 2025. However, Indonesia does not have accurate and comprehensive data on renewable energy potentials, especially wind energy. This article aims to assess the theoretical potential of wind speed and to visualize the wind speed by province for the entire Indonesia. Our assessment relied on the Weather Research and Forecasting (WRF) model using Four-Dimensional Data Assimilation technique, also known as Nudging Newtonian relaxation. The robustness of our analysis is confirmed by using high-resolution data from the National Centers for Environmental Prediction–Final (NCEP - FNL) and Cross-Calibrated Multi-Platform (CCMP) Reanalysis satellite data. This study shows the WRF method is a feasible option to estimate wind speed data.

Improved CYGNSS Wind Speed Retrieval Using Significant Wave Height Correction

This article presents the methodology for an improved estimation of the sea surface wind speed measured by the cyclone global navigation satellite system (CYGNSS) constellation of satellites using significant wave height (SWH) information as external reference data. The methodology consists of a correcting 2D look-up table (LUT) with inputs: (1) the CYGNSS wind speed given by the geophysical model function (GMF); and (2) the collocated reference SWH given by the WW3 model, which is forced by winds from the European Centre for Medium-Range Weather Forecasts (ECMWF) organization. In particular, the analyzed CYGNSS wind speeds are the fully developed seas (FDS) obtained with the GMF 3.0, and the forcing winds are the ECMWF forecast winds. Results show an increase in sensitivity to large winds speeds and an overall reduction in the root mean square difference (RMSD) with respect to the ECMWF winds from 2.05 m/s to 1.74 m/s. The possible influence of the ECWMF winds on the corrected winds (due to their use in the WW3 model) is analyzed by considering the correlation between: (1) the difference between the ECMWF winds and those from another reference; and (2) the difference between the corrected CYGNSS winds and those from the same reference. Results using ASCAT, WindSat, Jason, and AltiKa as references show no significant influence.

Improved Calibration of Wind Estimates from Advanced Scatterometer MetOp-B in Korean Seas Using Deep Neural Network

Satellite-based observations of sea wind are useful for forecasting marine weather and performing marine disaster management. Meteorological Operational Satellite-B (MetOp-B) is one of the satellites that provide wind products through a scatterometer named the Advanced Scatterometer (ASCAT). Since the linear regression method has been conventionally employed for calibrating remotely-sensed wind data, deviations and biases remain un-resolved to some degree. For coastal applications, these remotely-sensed wind data need to be calibrated again using local in-situ measurements in order to provide more accurate and realistic information. Thus, this study proposed a new method to calibrate ASCAT-based wind speed estimates using artificial neural networks. Herein, a deep neural network (DNN) model was applied. Wind databases collected during a period from 2012 to 2019 by the MetOp-B ASCAT and ten buoy stations in Korean seas were considered for deep learning-based calibration. ASCAT-based wind data and in-situ measurements were collocated in space and time. They were then separated into training and validation sets. A DNN model was designed and trained using multiple input variables such as observation location, sensing date and time, wind speed, and wind direction of the training set. The DNN-based best fit calibration model was evaluated using the validation set. The mean of biases between ASCAT-based and in-situ wind speeds was found to be decreased from 0.41 to 0.05 m/s on average for all buoy locations. The root mean squared error (RMSE) of wind speed was reduced from 1.38 m/s to 0.93 m/s. Moreover, the DNN-based calibration considerably improved the quality of wind speeds of less than 4 m/s, and of high wind speeds of 11–25 m/s. These results suggest that ASCAT-based observations can accurately represent real wind fields if a DNN-based calibration approach is applied.

DETERMINATION OF THE CONDITIONS FOR APPLYING THE TECHNOLOGY OF TARGET GRAVITY BRAKING OF UNCOUPLED CARS

Improving the efficiency of the process of train uncoupling remains one of the mosturgent problems in railway transport. The priority areas for solving this problem today include thedevelopment and implementation of integrated automation systems for hump technologies, the use ofartificial intelligence in decision support systems and procedures for calculating the controlparameters of car retarders. These systems provide high accuracy in designing and implementingcontrol actions, taking into account indeterminacy and multiple random parameters.However, a number of factors that are quite difficult to take into account, predict, or formalize.Such factors include the technical condition of rail car retarders, their random brakingcharacteristics, constant changes in wind speed and direction, the condition of wheel pairs, theresponse rate of hump operators, and others. These factors can have significant impact on the qualityof the marshalling process, even on automated marshalling humps.To increase the efficiency of the train uncoupling technology, researchers have developed anumber of scientific approaches and technical solutions: approaches to control using the equipmentfor controlling the speed of rolling uncoupled cars in automated mode have been formed, new designsof car retarders have been developed and the existing ones were improved, optimizing methods forbraking modes of uncoupled cars, the longitudinal profile of marshalling humps and the design ofhump necks have been proposed.
 The introduction of target gravity braking technology for uncouples cars is one of the ways toimprove the efficiency of the marshalling process. This technology can be implemented provided thata special profile design of the marshalling device is used.To determine the conditions for applying the target gravity braking technology for uncoupledcars, simulation modeling of rolling of model cars in favorable and unfavorable meteorologicalconditions was carried out.The results of simulation modeling showed that the proposed technology can be definitely used,if the number of tracks in the marshalling yard is up to 32 and estimated wind speeds up to 6 m/s,when it provides an uncoupling rate of up to 1,8 m/s. If the marshalling yard has a larger number oftracks and/or in case of powerful winds, the technology can be applied if the uncoupling rate isreduced. In this case, the feasibility of using the technology should be verified with technical andeconomic calculations.

Sensitivity analysis of planetary boundary layer schemes using the WRF model in Northern Colombia during 2016 dry season

This study conducted meteorological simulations in northern Colombia by analyzing different planetary boundary layer (PBL) schemes available in the numerical Weather Research and Forecasting (WRF) model. The study area included three nested domains with horizontal resolutions of 18 km, 6 km, and 2 km, with 38 vertical levels. The evolution and structure of the PBL were analyzed during the driest months (March, April, and May 2016) and in regions with the highest particulate matter concentrations. Sensitivity analysis of the WRF model was performed with two local and two non-local PBL schemes. The results were validated using observations of the surface air temperature, relative humidity, and surface wind speed collected from three meteorological stations in the area. The PBL heights were experimentally determined using radiosonde data provided by a station located in the center of the study area. Variations in PBL heights were estimated using linear regression methods and minimization of statistical errors for the bulk Richardson number, as well as analysis of vertical temperature and wind profiles. The WRF model reliably reproduced the daily values and diurnal cycles of temperature, relative humidity, and wind speed within the PBL and accounted for the influence of topography and sea breezes. Horizontal heat advection dominates the upwelling of air masses when sea breezes are active. The onshore wind direction starts to change from east to northwest, implying a decay in the land breeze regime. All schemes overestimate the mixing height and tend to underestimate surface air temperature values at night. All show wetter conditions and underestimate wind speed. Although the non-local Yonsei University (YSU) scheme shows the best performance, it also shows the largest sources of errors when determining the behavior of the surface layer during stable conditions. Relative humidity and wind speed estimates provided by the local Mellor‐Yamada‐Nakanishi‐Niino (MYNN) scheme were closer to those recorded at the meteorological stations.

On estimating long period wind speed return levels from annual maxima

The uniform risk engineering practices that are increasingly being adopted for structural design require estimates of the extreme wind loads with very low annual probabilities of exceedance, corresponding to return periods of up to 3000-years in some cases. These estimates are necessarily based on observational wind data that typically spans only a few decades. The estimates are therefore affected by both large sampling uncertainty and, potentially, non-negligible biases. Design practices that aim to meet mandated structural reliability criteria take the sampling uncertainty of long period wind speed or wind pressure estimates into account, but reliability could be compromised if estimates are also biased. In many circumstances, estimates are obtained by fitting an extreme value distribution to annual maximum wind speed observed over a few decades. A key assumption implicit in doing so is that wind speed annual maxima are max-stable. Departures from max-stability can exacerbate the uncertainty of long-period return level estimates by inducing systematic estimation bias as well. Observational records, however, are generally too short to assess max-stability. We therefore use wind speed data from a large (50-member) ensemble of CanRCM4 historical simulations over North America to assess whether wind speed annual maxima are max-stable. While results are generally reassuring at the continental scale, disquieting evidence of a lack of max-stability is often found in the central and southern parts of the continent. Results show that when annual maximum wind speeds are not max-stable, long period return level extreme wind speeds tend to be underestimated, which would compromise reliability if used to design infrastructure such as tall buildings and towers.