An Improved 2DCNN With Focal Loss Function for Blade Icing Detection of Wind Turbines Under Imbalanced SCADA Data

2021 ◽  
Author(s):  
Dandan Peng ◽  
Chenyu Liu ◽  
Wim Desmet ◽  
Konstantinos Gryllias
Keyword(s):  
Power Generation ◽  
Deep Learning ◽  
Wind Turbine ◽  
Loss Function ◽  
Wind Turbines ◽  
Cold Climate ◽  
Detection Methods ◽  
Physical Parameters ◽  
Learning Abilities ◽  
Imbalance Problem

Abstract The deployment of wind power plants in cold climate becomes ever more attractive due to the increased air density resulting from low temperatures, the high wind speeds, and the low population density. However, the cold climate conditions bring some additional challenges as itt can easily cause wind turbine blades to freeze. The frizzing ice on blades not only increases the energy required for the rotation of blades, resulting in a reduction in the power generation, but also increases the amplitude of the blades’ vibrations, which may cause the blade to break, affecting the power generation performance of the wind turbine and poses a threat to its safe operation. Current published blade icing detection methods focus on studying the blade icing mechanism, building the model and then judging if it is iced or not. These models vary with different wind turbines and working conditions, so expertise knowledge is required. However, deep learning techniques may solve the abovementioned problem based on their excellent feature learning abilities but until now, there are only few studies on wind turbine blade icing detection based on the deep learning technology. Therefore, this paper proposes a novel blade icing detection model, named two-dimensional convolutional neural network with focal loss function (FL-2DCNN). The network takes the raw data collected by the Supervisory Control and Data Acquisition (SCADA) system as input, automatically learns the correlation between the different physical parameters in the dataset, and captures the abnormal information, in order to accurately output the detection results. However, the amount of normal data collected by SCADA systems is usually much larger than the one of blade icing fault data, leading to a serious data imbalance problem. This problem makes it difficult for the network to obtain enough features related to the blade icing fault. Therefore the focal loss function is introduced to the FL-2DCNN to solve the aforementioned data imbalanced problem. The focal loss function can effectively balance the importance of normal samples and icing fault samples, so that the network can obtain more icing-related feature information from the icing fault samples, and thus the detection ability of the network can be improved. The experimental results of the proposed FL-2DCNN based on real SCADA data of wind turbines show that the proposed FL-2DCNN can effectively solve the sample imbalance problem and has significant potential in the blade icing detection task compared with other deep learning methods.

scholarly journals Unsupervised transfer learning for condition monitoring of a fleet of wind turbines

2021 ◽  
Vol 13 (1) ◽  
Author(s):  
Dandan Peng ◽  
Chenyu Liu ◽  
Wim Desmet ◽  
Konstantinos Gryllias
Keyword(s):  
Deep Learning ◽  
Health Status ◽  
Wind Turbine ◽  
Transfer Learning ◽  
Condition Monitoring ◽  
Wind Turbines ◽  
Working Environment ◽  
Generative Adversarial Networks ◽  
Physical Parameters ◽  
Scada Systems

The condition monitoring and health status prediction of a fleet of wind turbines are essential for the safety of wind turbines. At present, the Supervisory Control And Data Acquisition (SCADA) system has been widely used in wind turbines, which can monitor and collect various physical information and sensor information of wind turbines in real-time. Due to the fact that the amount of data obtained by SCADA systems is extremely large, developing an intelligent decision-making system based on deep learning is a very valuable research. Therefore, this paper is committed to exploring a health status prediction algorithm of wind turbines based on deep learning and SCADA systems. However, yet in actual industrial applications, it is very time-consuming and expensive to obtain a large amount of labeled data. In addition, as failures rarely occur, there is a serious sample imbalance problem in the datasets. More importantly, due to the difference in working environment and physical parameters, there are significant differences in the feature distribution of different wind turbines data, which lead to a significant drop in the performance of the deep learning model on unknown wind turbines. Therefore, we propose an unsupervised transfer learning algorithm based on Generative Adversarial Networks for wind turbine health status prediction (WT-GAN). WT-GAN can not only remove the domain shift between wind turbines, but also it is an unsupervised learning method. This practically means that only the unlabeled data for the target domain is required, which solves the problem of labeling data. In order to evaluate the effectiveness of WT-GAN on the condition monitoring of a fleet of wind turbines, we apply this method to one dataset about blade icing detection of wind turbines. The experimental results prove that the proposed method can predict the health of the wind turbine well. In addition, it can significantly reduce the domain shift among different wind turbines, thereby achieving excellent performance on unknown wind turbines.

scholarly journals Numerical Investigations of the Effects of the Rotating Shaft and Optimization of Urban Vertical Axis Wind Turbines

Energies ◽  
2018 ◽  
Vol 11 (7) ◽  
pp. 1870 ◽  
Author(s):  
Lidong Zhang ◽  
Kaiqi Zhu ◽  
Junwei Zhong ◽  
Ling Zhang ◽  
Tieliu Jiang ◽  
...  
Keyword(s):  
Wind Turbine ◽  
Wind Turbines ◽  
Suction Side ◽  
Vertical Axis ◽  
Power Coefficient ◽  
Small Scale ◽  
Physical Parameters ◽  
Rotating Shaft ◽  
Vertical Axis Wind Turbines ◽  
Negative Effect

The central shaft is an important and indispensable part of a small scale urban vertical axis wind turbines (VAWTs). Normally, it is often operated at the same angular velocity as the wind turbine. The shedding vortices released by the rotating shaft have a negative effect on the blades passing the wake of the wind shaft. The objective of this study is to explore the influence of the wake of rotating shaft on the performance of the VAWT under different operational and physical parameters. The results show that when the ratio of the shaft diameter to the wind turbine diameter (α) is 9%, the power loss of the wind turbine in one revolution increases from 0% to 25% relative to that of no-shaft wind turbine (this is a numerical experiment for which the shaft of the VAWT is removed in order to study the interactions between the shaft and blade). When the downstream blades pass through the wake of the shaft, the pressure gradient of the suction side and pressure side is changed, and an adverse effect is also exerted on the lift generation in the blades. In addition, α = 5% is a critical value for the rotating shaft wind turbine (the lift-drag ratio trend of the shaft changes differently). In order to figure out the impacts of four factors; namely, tip speed ratios (TSRs), α, turbulence intensity (TI), and the relative surface roughness value (ks/ds) on the performance of a VAWT system, the Taguchi method is employed in this study. The influence strength order of these factors is featured by TSRs > ks/ds > α > TI. Furthermore, within the range we have analyzed in this study, the optimal power coefficient (Cp) occurred under the condition of TSR = 4, α = 5%, ks/ds = 1 × 10−2, and TI = 8%.

scholarly journals A numerical investigation of the effects of ice accretion on the aerodynamic and structural behavior of offshore wind turbine blade

2020 ◽  
pp. 0309524X2098322
Author(s):  
Oumnia Lagdani ◽  
Mostapha Tarfaoui ◽  
Mourad Nachtane ◽  
Mourad Trihi ◽  
Houda Laaouidi
Keyword(s):  
Wind Turbine ◽  
Turbine Blade ◽  
Wind Turbines ◽  
Failure Modes ◽  
Leading Edge ◽  
Wind Turbine Blade ◽  
Cold Climate ◽  
Offshore Wind ◽  
Structural Behavior ◽  
Offshore Wind Turbine

In recent years, several wind turbines have been installed in cold climate sites and are menaced by the icing phenomenon. This article focuses on two parts: the study of the aerodynamic and structural performances of wind turbines subject to atmospheric icing. Firstly, the aerodynamic analysis of NACA 4412 airfoil was obtained using QBlade software for a clean and iced profile. Finite element method (FEM) was employed using ABAQUS software to simulate the structural behavior of a wind turbine blade with 100 mm ice thickness. A comparative study of two composite materials and two blade positions were considered in this section. Hashin criterion was chosen to identify the failure modes and determine the most sensitive areas of the structure. It has been found that the aerodynamic and structural performance of the turbine were degraded when ice accumulated on the leading edge of the blade and changed the shape of its profile.

scholarly journals A Survey of Deep Learning Approaches on Wind Turbine Condition Monitoring

2021 ◽  
Vol 2132 (1) ◽  
pp. 012015
Author(s):  
Sijia Li
Keyword(s):  
Deep Learning ◽  
Wind Turbine ◽  
Condition Monitoring ◽  
Wind Turbines ◽  
Control Charts ◽  
Original Data ◽  
Learning Approaches ◽  
Monitoring Methods ◽  
Effectiveness And Efficiency ◽  
Operational Data

Abstract Current physics-based wind turbine monitoring methods often need extra sensors installed on wind turbines, thus increasing the operation and maintenance (O&M) cost. Besides, physical methods are only effective under some constraints. The real effectiveness needs to be further checked in real conditions. Recent advances in data acquisition systems allow collection of large volumes of operational data of wind turbines. Learning knowledge from the data allows us to do monitoring in another direction. In this paper, a survey of deep learning algorithms applied to wind turbine condition monitoring is given. Compared with original data, more meaning features were extracted through feature extraction of deep learning. Monitoring these new signals, outliers were detected by applying suitable control charts. Several industrial cases confirmed the effectiveness and efficiency of these frameworks.

Power Generation From Wind Turbines in a Solar Chimney

2011 ◽  
Author(s):  
Tudor Foote ◽  
Ramesh Agarwal
Keyword(s):  
Power Generation ◽  
Wind Turbine ◽  
Wind Power ◽  
Wind Turbines ◽  
Stokes Equations ◽  
Computational Study ◽  
Boussinesq Approximation ◽  
Navier Stokes ◽  
Solar Chimney ◽  
Wind Speeds

In past several years, several studies have shown that the shrouded wind turbines can generate greater power compared to bare turbines. A solar chimney not only generates an upward draft of the wind inside the solar tower but also creates a shroud around the wind turbine. There is large number of empty silos on farms, especially in mid-western U.S. They can be used as a solar chimney with minor modifications at very modest cost. The objective of this study is to determine the potential of these silos/chimneys in generating wind-power by installing a wind turbine inside the silo. An analytical/computational study is performed to evaluate this potential by employing the well known commercial Computational Fluid Dynamics (CFD) software FLUENT. An actuator disc model is used to model the turbine. Calculations are performed for three cases using the dimensions of a typical silo and assuming Class 3 wind velocity: (a) bare turbine (without enclosing silo), (b) turbine enclosed by a cylindrical silo, and (c) the turbine enclosed by the cylindrical silo with a diffuser at the top of the silo. The incompressible Navier-Stokes equations with Boussinesq approximation and a two equation realizable k–ε model are employed in the calculations. Cp and generated power are calculated for the three cases. It was found that the silo increases the Cp beyond the Betz’s limit significantly and as a result the generated power; this effect is consistent with that found in the recent literature that the shrouded wind-turbines can generate greater power than the bare turbines. The inclusion of a diffuser on top of the silo further increases the generated power and Cp. The results reported here are for typical silo dimensions and wind speeds; the results for silos with different dimensions and wind speeds can be easily generated. This study shows the potential of using abandoned silos in mid-west for wind power generation.

Variable Speed Direct Drive Wind Controller for a Rim Driven Wind Turbine

2012 ◽  
Author(s):  
Chase Hubbard ◽  
Rob Hovsapian ◽  
Srinivas Kosaraju
Keyword(s):  
Power Generation ◽  
Control System ◽  
Wind Turbine ◽  
Wind Turbines ◽  
Aerodynamic Force ◽  
Rotor Blades ◽  
Power Coefficient ◽  
Variable Speed ◽  
Direct Drive ◽  
Significant Difference

Multi-blade shaft driven wind turbines depend greatly on the angle of attack as an important factor that the control system monitors such that a maximum amount of aerodynamic force is seen by the rotor blades. This is one significant difference when controlling a Rim Driven Wind Turbine (RDWT). The controller for a RDWT is required to simply point the tower such that it is facing the wind for maximum power generation. This is achieved by incorporating a Variable Speed Direct Drive (VSDD) wind operation control system to control the power production and safe operation of the RDWT. Another consideration for the control system is its integration with the generator. Since the power generation is rim driven, thus operating at a higher variable speed. With information related to the wind turbine’s diameter and the wind speed at any given time it can be calculated how much power can be potentially generated. This can then be in turn relayed to the generator from the wind turbine controller. This information can be relayed using controller-controller communication (such as an analog voltage signal or protocol based communication such as MODBUS RTU or TCP/IP) representing the power coefficient from Betz’ Law. A feasibly controllable system implements a signal from the overall wind turbine controller that in turn supplies the generator with how much power is available in the system to maximize power generation for a broad range of traditionally unrealizable wind conditions (3 m/s to 30 m/s). Rim Driven Wind Turbines represent an evolution in fundamental design of how the wind can be harnessed for power. This paper will discuss the VSDD’s unique design and aspects of maintaining controllability thorough out the overall system operation.

scholarly journals Simulative Investigation of the Radar Cross Section of Wind Turbines

2019 ◽  
Vol 9 (19) ◽  
pp. 4024 ◽  
Author(s):  
Sebastian Hegler ◽  
Dirk Plettemeier
Keyword(s):  
Power Generation ◽  
Wind Turbine ◽  
Cross Section ◽  
Wind Turbines ◽  
Wind Power Generation ◽  
Rotor Blades ◽  
Design Parameters ◽  
Complex Nature ◽  
Radar Systems ◽  
Turbine Design

Wind-power generation is one of the fundamental sources of renewable energy. However, due to the increasing size of wind turbines, they cause unwanted interference with radar systems for civic protection, especially for on-shore locations. This paper presents parameter studies performed on different wind-turbine models, with a focus on differences of the aerodynamical shape of the rotor blades. Numerical simulation is employed to estimate the influence of different wind-turbine design parameters, with the aim of deriving strategies to minimize wind-turbine influence on radar systems for civic protection. Due to the complex nature of the aerodynamic shape of the blade, a general model cannot be derived from the studies. However, further steps to eventually achieve this goal are outlined.

scholarly journals Design procedures and experimental verification of an electro-thermal deicing system for wind turbines

2021 ◽  
Vol 6 (5) ◽  
pp. 1291-1309
Author(s):  
David Getz ◽  
Jose Palacios
Keyword(s):  
Wind Turbine ◽  
Power Density ◽  
Turbine Blade ◽  
Wind Turbines ◽  
Rotor Blade ◽  
Wind Turbine Blade ◽  
Cold Climate ◽  
Ice Formation ◽  
Ice Thickness ◽  
Ice Accretion

Abstract. There has been a substantial growth in the wind energy power capacity worldwide, and icing difficulties have been encountered in cold climate locations. Rotor blade icing has been recognized as an issue, and solutions to mitigate accretion effects have been identified. Wind turbines are adapting helicopter rotor and propeller ice protection approaches to reduce aerodynamic performance degradation related to ice formation. Electro-thermal heating is one of the main technologies used to protect rotors from ice accretion, and it is one of the main technologies being considered to protect wind turbines. In this research, the design process required to develop an ice protection system for wind turbines is discussed. The design approach relies on modeling and experimental testing. Electro-thermal heater system testing was conducted at the Adverse Environment Rotor Test Stand at Penn State, where wind turbine representative airfoils protected with electro-thermal deicing were tested at representative centrifugal loads and flow speeds. The wind turbine sections tested were half-scale models of the 80 % span region of a generic 1.5 MW wind turbine blade. The icing cloud impact velocity was matched to that of a 1.5 MW wind turbine at full power production. Ice accretion modeling was performed to provide an initial estimate of the power density required to de-bond accreted ice at a set of icing conditions. Varying icing conditions were considered at −8 ∘C with liquid water contents of the cloud varying from 0.2 to 0.9 g/m3 and water droplets from 20 µm median volumetric diameter to 35 µm. Then, ice accretion thickness gradients along the span of the rotor blade for the icing conditions were collected experimentally. Given a pre-determined maximum power allocated for the deicing system, heating the entire blade was not possible. Heating zones were introduced along the span and the chord of the blade to provide the required power density needed to remove the accreted ice. The heating sequence for the zones started at the tip of the blade, to allow de-bonded ice to shed off along the span of the rotor blade. The continuity of the accreted ice along the blade span means that when using a portioned heating zone, ice could de-bond over that specific zone, but the ice formation could remain attached cohesively as it is connected to the ice on the adjacent inboard zone. To prevent such cohesive retention of de-bonded ice sections, the research determined the minimum ice thickness required to shed the accreted ice mass with the given amount of power availability. The experimentally determined minimum ice thickness for the varying types of ice accreted creates sufficient tensile forces due to centrifugal loads to break the cohesive ice forces between two adjacent heating zones. The experimental data were critical in the design of a time sequence controller that allows consecutive deicing of heating zones along the span of the wind turbine blade. Based on the experimental and modeling efforts, deicing a representative 1.5 MW wind turbine with a 100 kW power allocation required four sections along the blade span, with each heater section covering 17.8 % span and delivering a 2.48 W/in.2 (0.385 W/cm2) power density.

scholarly journals Short-term wind speed forecasting system using deep learning for wind turbine applications

2020 ◽  
Vol 10 (6) ◽  
pp. 5779
Author(s):  
Gokhan Erdemir ◽  
Aydin Tarik Zengin ◽  
Tahir Cetin Akinci
Keyword(s):  
Deep Learning ◽  
Wind Speed ◽  
Wind Turbine ◽  
Wind Turbines ◽  
Learning Algorithm ◽  
Integrated System ◽  
Short Term ◽  
Wind Speed Forecasting ◽  
Short Term Forecasting

It is very important to accurately detect wind direction and speed for wind energy that is one of the essential sustainable energy sources. Studies on the wind speed forecasting are generally carried out for long-term predictions. One of the main reasons for the long-term forecasts is the correct planning of the area where the wind turbine will be built due to the high investment costs and long-term returns. Besides that, short-term forecasting is another important point for the efficient use of wind turbines. In addition to estimating only average values, making instant and dynamic short-term forecasts are necessary to control wind turbines. In this study, short-term forecasting of the changes in wind speed between 1-20 minutes using deep learning was performed. Wind speed data was obtained instantaneously from the feedback of the emulated wind turbine's generator. These dynamically changing data was used as an input of the deep learning algorithm. Each new data from the generator was used as both test and training input in the proposed approach. In this way, the model accuracy and enhancement were provided simultaneously. The proposed approach was turned into a modular independent integrated system to work in various wind turbine applications. It was observed that the system can predict wind speed dynamically with around 3% error in the applications in the test setup applications.

scholarly journals A Modified Reynolds-Averaged Navier–Stokes-Based Wind Turbine Wake Model Considering Correction Modules

Energies ◽  
2020 ◽  
Vol 13 (17) ◽  
pp. 4430
Author(s):  
Yuan Li ◽  
Zengjin Xu ◽  
Zuoxia Xing ◽  
Bowen Zhou ◽  
Haoqian Cui ◽  
...  
Keyword(s):  
Power Generation ◽  
Power Systems ◽  
Wind Turbine ◽  
Turbulence Model ◽  
Wind Turbines ◽  
Navier Stokes ◽  
Aerodynamic Model ◽  
Wake Model ◽  
Turbine Wake ◽  
Wind Turbine Wake

Increasing wind power generation has been introduced into power systems to meet the renewable energy targets in power generation. The output efficiency and output power stability are of great importance for wind turbines to be integrated into power systems. The wake effect influences the power generation efficiency and stability of wind turbines. However, few studies consider comprehensive corrections in an aerodynamic model and a turbulence model, which challenges the calculation accuracy of the velocity field and turbulence field in the wind turbine wake model, thus affecting wind power integration into power systems. To tackle this challenge, this paper proposes a modified Reynolds-averaged Navier–Stokes (MRANS)-based wind turbine wake model to simulate the wake effects. Our main aim is to add correction modules in a 3D aerodynamic model and a shear-stress transport (SST) k-ω turbulence model, which are converted into a volume source term and a Reynolds stress term for the MRANS-based wake model, respectively. A correction module including blade tip loss, hub loss, and attack angle deviation is considered in the 3D aerodynamic model, which is established by blade element momentum aerodynamic theory and an improved Cauchy fuzzy distribution. Meanwhile, another correction module, including a hold source term, regulating parameters and reducing the dissipation term, is added into the SST k-ω turbulence model. Furthermore, a structured hexahedron mesh with variable size is developed to significantly improve computational efficiency and make results smoother. Simulation results of the velocity field and turbulent field with the proposed approach are consistent with the data of real wind turbines, which verifies the effectiveness of the proposed approach. The variation law of the expansion effect and the double-hump effect are also given.

Sign in / Sign up

Export Citation Format

Share Document