Using neural networks to estimate wind turbine power generation

This paper uses data collected at Central and South West Services Fort Davis wind farm (USA) to develop a neural network based prediction of power produced by each turbine. The power generated by electric wind turbines changes rapidly because of the continuous fluctuation of wind speed and direction. It is important for the power industry to have the capability to perform this prediction for diagnostic purposes-lower-than-expected wind power may be an early indicator of a need for maintenance. In this paper, characteristics of wind power generation are first evaluated in order to establish the relative importance for the neural network. A four input neural network is developed and its performance is shown to be superior to the single parameter traditional model approach     

A general method to estimate wind farm power using artificial neural networks

An artificial neural network (ANN) is trained and validated using a large dataset of observations of wind speed, direction, and power generated at an offshore wind farm (Lillgrund in Sweden). In its traditional form, the ANN is used to generate a new two‐dimensional power curve, which predicts with high accuracy (bias ～?0.5% and absolute error ～2%) the power of the entire Lillgrund wind farm based on wind speed and direction. By contrast, manufacturers only provide one‐dimensional power curves (i.e., power as a function of wind speed) for a single turbine. The second innovative application of the ANN is the use of a geometric model (GM) to calculate two simple geometric properties to replace wind direction in the ANN. The resulting GM‐ANN has the powerful feature of being applicable to any wind farm, not just Lillgrund. A validation at an onshore wind farm (Nørrekær in Denmark) demonstrates the high accuracy (bias ～?0.7% and absolute error ～6%) and transfer‐learning ability of the GM‐ANN.     

Use of artificial neural networks for transport energy demand modeling

The paper illustrates an artificial neural network (ANN) approach based on supervised neural networks for the transport energy demand forecasting using socio-economic and transport related indicators. The ANN transport energy demand model is developed. The actual forecast is obtained using a feed forward neural network, trained with back propagation algorithm. In order to investigate the influence of socio-economic indicators on the transport energy demand, the ANN is analyzed based on gross national product (GNP), population and the total annual average veh-km along with historical energy data available from 1970 to 2001. Comparing model predictions with energy data in testing period performs the model validation. The projections are made with two scenarios. It is obtained that the ANN reflects the fluctuation in historical data for both dependent and independent variables. The results obtained bear out the suitability of the adopted methodology for the transport energy-forecasting problem.     

Fuzzy neural network output maximization control for sensorless wind energy conversion system

This paper presents the design of an online training fuzzy neural network (FNN) controller with a high-performance speed observer for the induction generator (IG). The proposed output maximization control is achieved without mechanical sensors such as the wind speed or position sensor, and the new control system will deliver maximum electric power with light weight, high efficiency, and high reliability. The estimation of the rotor speed is designed on the basis of the sliding mode control theory.     

A study into the accuracy of using meteorological wind data to estimate turbine generation output

Meteorological (met) station data is used as the basis for a number of influential studies into the impacts of the variability of renewable resources. Real turbine output data is not often easy to acquire, whereas meteorological wind data, supplied at a standardised height of 10 m, is widely available. This data can be extrapolated to a standard turbine height using the wind profile power law and used to simulate the hypothetical power output of a turbine. Utilising a number of met sites in such a manner can develop a model of future wind generation output. However, the accuracy of this extrapolation is strongly dependent on the choice of the wind shear exponent α. This paper investigates the accuracy of the simulated generation output compared to reality using a wind farm in North Rhins, Scotland and a nearby met station in West Freugh. The results show that while a single annual average value for α may be selected to accurately represent the long term energy generation from a simulated wind farm, there are significant differences between simulation and reality on an hourly power generation basis, with implications for understanding the impact of variability of renewables on short timescales, particularly system balancing and the way that conventional generation may be asked to respond to a high level of variable renewable generation on the grid in the future.     

The application of artificial neural networks to mapping of wind speed profile for energy application in Nigeria

Modelling and prediction of wind speed are essential prerequisites in the sitting and sizing of wind power applications. The profile of wind speed in Nigeria is modelled using artificial neural network (ANN). The ANN model consists of 3-layered, feed-forward, back-propagation network with different configurations, designed using the Neural Toolbox for MATLAB. The monthly mean daily wind speed data monitored at 10 m above ground level for a period of 20 years (1983–2003) for 28 ground stations operated by the Nigeria Meteorological Services (NIMET) were used as training (18 stations) and testing (10 stations) dataset. The geographical parameters (latitude, longitude and altitude) and the month of the year were used as input data, while the monthly mean wind speed was used as the output of the network. The optimum network architecture with minimum Mean Absolute Percentage Error (MAPE) of 8.9% and correlation coefficient (r) between the predicted and the measured wind speed values of 0.9380 was obtained. The predicted monthly wind speed ranged from 0.9–13.1 m/s with an annual mean of 4.7 m/s. The model predicted wind speed values are given in the form of monthly maps, which can be easily used for assessment of wind energy potential for different locations within Nigeria.     

Development of small domestic wind turbine with scoop and prediction of its annual power output

Based on an unperturbed airflow assumption and using a set of validated modelling methods, a series of activities were carried out to optimise an aerodynamic design of a small wind turbine for a built up area, where wind is significantly weaker and more turbulent than those open sites preferable for wind farms. These activities includes design of the blades using a FORTRAN code; design of the nose cones and nacelles, which then constituted the rotor along with the blades; optimisation of the rotor designs in the virtual wind tunnel developed in the first part of the study; and finally, estimation of the annual power output of this wind turbine calculated using hourly wind data of a real Scottish Weather Station. The predicted annual output of the finalised rotor was then compared with other commercial turbines and result was rather competitive.     

Gas turbine sensor validation through classification with artificial neural networks

Modern power plants are all strongly dependent on reliable and accurate sensor readings for monitoring and control, thus making sensors an important part of any plant. Failing sensors can force a plant or component into non-optimal operation, cause complete shut-down of operation or in the worst case result in damage to components. Given their importance, sensors need regular calibration and maintenance, a time-consuming and therefore costly process. In this paper a method is presented for evaluating sensor accuracy which aims to minimize the need for calibration and at the same time avoid shut-downs due to sensor faults etc. The proposed method is based on training artificial neural networks as classifiers to recognize sensor drifts. The method is evaluated on two types of gas turbines, i.e., one single-shaft and one twin-shaft machine. The results show the method is capable of early detection of sensor drifts for both types of machines as well as accurate production of soft measurements. The findings suggest that the use of artificial neural networks for sensor validation could contribute to more cost-effective maintenance as well as to increased availability and reliability of power plants.     

On comparing three artificial neural networks for wind speed forecasting

Wind speed forecasting is critical for wind energy conversion systems since it greatly influences the issues such as the scheduling of a power system, and the dynamic control of the wind turbine. In this paper, we present a comprehensive comparison study on the application of different artificial neural networks in 1-h-ahead wind speed forecasting. Three types of typical neural networks, namely, adaptive linear element, back propagation, and radial basis function, are investigated. The wind data used are the hourly mean wind speed collected at two observation sites in North Dakota. The performance is evaluated based on three metrics, namely, mean absolute error, root mean square error, and mean absolute percentage error. The results show that even for the same wind dataset, no single neural network model outperforms others universally in terms of all evaluation metrics. Moreover, the selection of the type of neural networks for best performance is also dependent upon the data sources. Among the optimal models obtained, the relative difference in terms of one particular evaluation metric can be as much as 20%. This indicates the need of generating a single robust and reliable forecast by applying a post-processing method.     

Application of artificial neural networks for modeling of biohydrogen production

In this study, an artificial neural network (ANN) model was developed to estimate the hydrogen production profile with time in batch studies. A back propagation artificial neural network ANN configuration of 5–6–4–1 layers was developed. The ANN inputs were the initial pH, initial substrate and biomass concentrations, temperature, and time. The model training was done using 313 data points from 26 published experiments. The correlation coefficient between the experimental and estimated hydrogen production was 0.989 for training, validating, and testing the model. Results showed that the trained ANN successfully predicted the hydrogen production profile with time for new data with a correlation coefficient of 0.976.     

Greek long-term energy consumption prediction using artificial neural networks

In this paper artificial neural networks (ANN) are addressed in order the Greek long-term energy consumption to be predicted. The multilayer perceptron model (MLP) has been used for this purpose by testing several possible architectures in order to be selected the one with the best generalizing ability. Actual recorded input and output data that influence long-term energy consumption were used in the training, validation and testing process. The developed ANN model is used for the prediction of 2005–2008, 2010, 2012 and 2015 Greek energy consumption. The produced ANN results for years 2005–2008 were compared with the results produced by a linear regression method, a support vector machine method and with real energy consumption records showing a great accuracy. The proposed approach can be useful in the effective implementation of energy policies, since accurate predictions of energy consumption affect the capital investment, the environmental quality, the revenue analysis, the market research management, while conserve at the same time the supply security. Furthermore it constitutes an accurate tool for the Greek long-term energy consumption prediction problem, which up today has not been faced effectively.     

提高风出力预测精度的储能系统模糊控制策略

风气象信息的精细化程度不够高会造成风电场风出力预测精度的降低,给电网调度增加了难度。在风电场中配置锌溴电池储能系统是提高风电场日前预报精度的有效措施,对储能装置控制是其关键问题。文章采用模糊控制方法,搭建了储能系统模糊控制规则库,根据电池储能系统荷电状态SOC的变化来控制储能充放电功率;并将所提出的控制策略在新疆达坂城风电场-储能联合发电系统中进行了仿真验证,与传统控制策略进行了对比分析。研究结果表明,采用模糊控制策略的储能系统能够更有效地提高风出力短期预测精度,85%的预测值达到了国家电网要求。     

Validation of a hybrid modeling approach to floating wind turbine basin testing

Hybrid modeling combining physical tests and numerical simulations in real time opens new opportunities in floating wind turbine research. Wave basin testing is an important validation step for floating support structure design, but current methods are limited by scaling problems in the aerodynamic loadings. Applying wind turbine loads with an actuation system controlled by a simulation that responds to the basin test offers a way to avoid scaling problems and reduce cost barriers for floating wind turbine design validation in realistic coupled conditions. In this work, a cable‐based hybrid coupling approach is developed and implemented for 1:50‐scale wave basin tests with the DeepCwind semisubmersible floating wind turbine. Tests are run with thrust loads provided by a numerical wind turbine model. Matching tests are run with physical wind loads using an above‐basin wind maker. When the numerical submodel is set to match the aerodynamic performance of the physical scaled wind turbine, the results show good agreement with purely physical wind‐wave tests, validating the hybrid model approach. Further hybrid model tests with simulated true‐to‐scale dynamic thrust loads and wind turbulence show noticeable differences and demonstrate the value of a hybrid model approach for improving the true‐to‐scale realism of floating wind turbine basin tests.     

Fault detection of wind energy conversion systems using recurrent neural networks

Structure of wind energy conversion systems (WECSs) must be robust against faults. In order to accurately study WECSs during occurrence of faults and to explore the impact of faults on each component of the WECSs, a detailed model is required in which both mechanical and electrical parts of the WECSs are properly involved. In addition, a fault detection system (FDS) is required to diagnose the occurred faults at the appropriate time in order to ensure a safe system operation, avoid heavy economic losses, prevent damage to adjacent relevant systems and facilitate timely repair of failed components. This can be performed by subsequent actions through fast and accurate detection of faults. In this paper, by utilising a comprehensive dynamic model of the WECS, an FDS is presented using dynamic recurrent neural networks. In industrial processes, dynamic neural networks are known as a good mathematical tool for fault detection. The proposed FDS detects faults of the generator's angular velocity sensor, pitch angle sensors and pitch actuators. The presented FDS has high capability of fault detection in short time and it has much low false alarms rate. Simulation results verify validity and usefulness of the proposed fault detection scheme.     

Global sensitivity analysis of wind turbine power output

The dynamics of wind turbine behavior are complex and a critical area of study for the wind industry. Identification of factors that cause changes in turbine performance can sometimes prove to be challenging, whereas other times, it can be intuitive. The quantification of the effect that these factors have is valuable for making improvements to both power performance and turbine health. In commercial farms, large quantities of meteorological and performance data are commonly collected to monitor daily operations. These data can also be used to analyze the relationship between each parameter in order to better understand the interactions that occur and the information contained within these signals. In this global sensitivity analysis, a neural network is used to model select wind turbine supervisory control and data acquisition system parameters for an array of turbines from a commercial wind farm that exhibit signs of wake interaction. An extended Fourier amplitude sensitivity test is then performed for 2 years of 10‐min averaged data. The study examines the primary and combined sensitivities of power output to each selected parameter for two turbines in the array. The primary sensitivities correspond to single parameter interactions, whereas combined sensitivities account for interactions between multiple parameters simultaneously. Highly influential parameters such as wind speed and rotor rotation frequency produce expected results; the extended Fourier amplitude sensitivity test method proved effective at quantifying the sensitivity of a wide range of more subtle inputs. These include blade pitch, yaw position, main bearing and ambient temperatures as well as wind speed and yaw position standard deviation. The technique holds promise for application in full‐scale wake studies where it might be used to determine the benefits of emerging power optimization strategies such as active wake management. The field of structural health monitoring can also benefit from this method. Copyright © 2013 John Wiley & Sons, Ltd.     

Sensitivity of southern California wind energy to turbine characteristics

Using output from a high‐resolution meteorological simulation, we evaluate the sensitivity of southern California wind energy generation to variations in key characteristics of current wind turbines. These characteristics include hub height, rotor diameter and rated power, and depend on turbine make and model. They shape the turbine's power curve and thus have large implications for the energy generation capacity of wind farms. For each characteristic, we find complex and substantial geographical variations in the sensitivity of energy generation. However, the sensitivity associated with each characteristic can be predicted by a single corresponding climate statistic, greatly simplifying understanding of the relationship between climate and turbine optimization for energy production. In the case of the sensitivity to rotor diameter, the change in energy output per unit change in rotor diameter at any location is directly proportional to the weighted average wind speed between the cut‐in speed and the rated speed. The sensitivity to rated power variations is likewise captured by the percent of the wind speed distribution between the turbines rated and cut‐out speeds. Finally, the sensitivity to hub height is proportional to lower atmospheric wind shear. Using a wind turbine component cost model, we also evaluate energy output increase per dollar investment in each turbine characteristic. We find that rotor diameter increases typically provide a much larger wind energy boost per dollar invested, although there are some zones where investment in the other two characteristics is competitive. Our study underscores the need for joint analysis of regional climate, turbine engineering and economic modeling to optimize wind energy production. Copyright © 2012 John Wiley & Sons, Ltd.     

Transport energy demand modeling of South Korea using artificial neural network

Artificial neural network models were developed to forecast South Korea's transport energy demand. Various independent variables, such as GDP, population, oil price, number of vehicle registrations, and passenger transport amount, were considered and several good models (Model 1 with GDP, population, and passenger transport amount; Model 2 with GDP, number of vehicle registrations, and passenger transport amount; and Model 3 with oil price, number of vehicle registrations, and passenger transport amount) were selected by comparing with multiple linear regression models. Although certain regression models obtained better R-squared values than neural network models, this does not guarantee the fact that the former is better than the latter because root mean squared errors of the former were much inferior to those of the latter. Also, certain regression model had structural weakness based on P-value. Instead, neural network models produced more robust results. Forecasted results using the neural network models show that South Korea will consume around 37 MTOE of transport energy in 2025.     

Dynamic wind turbine output power reduction under varying wind speed conditions due to inertia

The inertia of wind turbines causes a reduction in their output power due to their inability to operate at the turbine maximum co‐efficient of performance point under dynamic wind conditions. In this paper, this dynamic power reduction is studied analytically and using simulations, assuming that a steady‐state optimal torque control strategy is used. The concepts of the natural and actual turbine time‐constant are introduced, and typical values for these parameters are examined. It is shown that for the typical turbine co‐efficient of performance curve used, the average turbine speed can be assumed to be determined by the average wind speed. With this assumption, analytical expressions for the power reduction with infinite and then finite turbine inertia are determined for sine‐wave wind speed variations. The results are then generalized for arbitrary wind speed profiles. A numerical wind turbine system simulation model is used to validate the analytical results for step and sine‐wave wind speed variations. Finally, it is used with real wind speed data to compare with the analytical predictions. Copyright © 2012 John Wiley & Sons, Ltd.     

Assessment of optimum tip speed ratio in wind turbines using artificial neural networks

Wind turbine blade design depends on several factors, such as turbine profile used, blade number, power factor, and tip speed ratio. The key to designing a wind turbine is to assess the optimal tip speed ratio (TSR). This will directly affect the power generated and, in turn, the effectiveness of the investment made. TSR is suggested to be taken between 7 and 8 and in practice generally taken as 7 for a 3-blade network-connected wind turbine. However, the optimal TSR is dependent upon the profile type used and the blade number and could fall out of the boundaries suggested. Therefore, it has to be assessed accordingly. In this study, the optimal TSR and the power factor of a wind turbine are predicted using artificial neural networks (ANN) based on the parameters involved for NACA 4415 and LS-1 profile types with 3 and 4 blades. The ANN structure built is found to be more successful than the conventional approach in estimating the TSR and power factor.     

Nonlinear model identification of wind turbine with a neural network

A nonlinear model of wind turbine based on a neural network (NN) is described for the estimation of wind turbine output power. The proposed nonlinear model uses the wind speed average, the standard deviation and the past output power as input data. An anemometer with a sampling rate of one second provides the wind speed data. The NN identification process uses a 10-min average speed with its standard deviation. The typical local data collected in September 2000 is used for the training, while those of October 2000 are used to validate the model. The optimal NN configuration is found to be 8-5-1 (8 inputs, 5 neurons on the hidden layer, one neuron on the output layer). The estimated mean square errors for the wind turbine output power are less than 1%. A comparison between the NN model and the stochastic model mostly used in the wind power prediction is done. This work is a basic tool to estimate wind turbine energy production from the average wind speed.