Wind Speed Power Research Articles

A PSO-LSTM Model of Offshore Wind Power Forecast considering the Variation of Wind Speed in Second-Level Time Scale

To enable power generation companies to make full use of effective wind energy resources and grid companies to correctly schedule wind power, this paper proposes a model of offshore wind power forecast considering the variation of wind speed in second-level time scale. First, data preprocessing is utilized to process the abnormal data and complete the normalization of offshore wind speed and wind power. Then, a wind speed prediction model is established in the second time scale through the differential smoothing power sequence. Finally, a rolling PSO-LSTM memory network is authorized to realize the prediction of second-level time scale wind speed and power. An offshore wind power case is utilized to illustrate and characterize the performance of the wind power forecast model.

The physics of galactic winds driven by cosmic rays – II. Isothermal streaming solutions

ABSTRACT We use analytical calculations and time-dependent spherically symmetric simulations to study the properties of isothermal galactic winds driven by cosmic rays (CRs) streaming at the Alfvén velocity. The simulations produce time-dependent flows permeated by strong shocks; we identify a new linear instability of sound waves that sources these shocks. The shocks substantially modify the wind dynamics, invalidating previous steady state models: the CR pressure pc has a staircase-like structure with dpc/dr ≃ 0 in most of the volume, and the time-averaged CR energetics are in many cases better approximated by pc ∝ ρ1/2, rather than the canonical pc ∝ ρ2/3. Accounting for this change in CR energetics, we analytically derive new expressions for the mass-loss rate, momentum flux, wind speed, and wind kinetic power in galactic winds driven by CR streaming. We show that streaming CRs are ineffective at directly driving cold gas out of galaxies, though CR-driven winds in hotter ISM phases may entrain cool gas. For the same physical conditions, diffusive CR transport (Paper I) yields mass-loss rates that are a few-100 times larger than streaming transport, and asymptotic wind powers that are a factor of ≃4 larger. We discuss the implications of our results for galactic wind theory and observations; strong shocks driven by CR-streaming-induced instabilities produce gas with a wide range of densities and temperatures, consistent with the multiphase nature of observed winds. We also quantify the applicability of the isothermal gas approximation for modelling streaming CRs and highlight the need for calculations with more realistic thermodynamics.

Wind power short-term power forecasting based on Improved Grey Wolf algorithm and Optimized Generalized Regression Neural Network

Aiming at the problem of wind turbine output prediction, a wind power prediction method based on Improved Gray Wolf algorithm and optimized generalized regression neural network is proposed in this paper. Firstly, according to the daily similarity of wind speed and wind power, cluster analysis is used to classify the data. Considering that the degree of each factor affecting wind power output changes, based on the selection of similar days, an improved gray wolf algorithm is introduced to optimize the weight of each influencing factor. The two models of the first mock exam are selected to input the radial single mode function RBF and the back propagation (BP) network to predict the output of the wind turbine separately. The prediction results of the two models are input to the generalized regression neural network optimized by the Wolf Wolf algorithm and the nonlinear combination forecasting is carried out. The basis models are used to predict the output of the wind turbine. The example analysis of an area shows that the model can be closer to the real value in the peak and valley of the prediction curve and has higher prediction accuracy than the combined prediction model of single BP, RBF and non optimized general regression neural network (GRNN).

Active power control strategy of wind farm considering fatigue load of wind turbines

For the purpose of maximize the utilization of wind energy and reduce the generation cost of wind farms, an active power control strategy of wind farm considering the fatigue load intensity of wind turbines was proposed. Turbulence intensity, average wind speed and active power command were used to evaluate the comprehensive fatigue load strength of the wind turbine. Then an improved particle swarm optimization algorithm was applied to optimize the active power distribution of each wind turbine, which maximize wind energy capture and keep the fatigue load intensity of each wind turbine within a certain allowable range. The effectiveness is proved by the analysis of an example. The results show that the control strategy is helpful to reduce the fatigue load intensity and reduce the operation and maintenance of the WF and the cost of power generation, which can improve the economy of the power system.

REDUCING CLIMATE CHANGE BY INSTALLING A NEW PHOTOVOLTAIC POWER PLANT IN BULGARIA

Three new roof-mounted 10 kWp grid-connected photovoltaic (PV) power plants have been constructed in the Technology Park at the Technical University of Gabrovo, Bulgaria, as part of a project "Competence Center – Intelligent Mechatronic, Eco, and Energy Saving Systems and Technologies". Three different PV modules types have been used: monocrystalline silicon (mono-Si), cadmium telluride (CdTe), and copper indium gallium selenide (CIGS). New three power plants, together with the existing amorphous silicon and poly-crystalline silicon photovoltaic power plants at the TU-Gabrovo enhanced simultaneous testing of five different photovoltaic materials. A small 500 Wp mono-Si photovoltaic thermal hybrid solar collectors (PVT) PV system has also been constructed. The power plants feature a monitoring system for the meteorological and electrical operating parameters, which measures, displays, and stores data on solar radiation, temperature, wind speed, currents, voltages, and electrical power of each power plant. PV plants components' technical characteristics are given in the paper. The schemes describing the basic wiring diagram, disposition of the three PV subsystems on the roof of the building at the technology center have also been presented. The initial comparative software data for monitoring of meteorological and electrical operating characteristics of the three different types of PV subsystems are shown as well. According to the specific ecological equivalent of energy resources and energy for the region of Bulgaria, the data are presented on the saved CO2 emissions from the avoided production and transmission of electricity owing to the operation of photovoltaic power plants.

A Combined Forecasting System Based on Modified Multi-Objective Optimization for Short-Term Wind Speed and Wind Power Forecasting

Wind speed and wind power are two important indexes for wind farms. Accurate wind speed and power forecasting can help to improve wind farm management and increase the contribution of wind power to the grid. However, nonlinear and non-stationary wind speed and wind power can influence the forecasting performance of different models. To improve forecasting accuracy and overcome the influence of the original time series on the model, a forecasting system that can effectively forecast wind speed and wind power based on a data pre-processing strategy, a modified multi-objective optimization algorithm, a multiple single forecasting model, and a combined model is developed in this study. A data pre-processing strategy was implemented to determine the wind speed and wind power time series trends and to reduce interference from noise. Multiple artificial neural network forecasting models were used to forecast wind speed and wind power and construct a combined model. To obtain accurate and stable forecasting results, the multi-objective optimization algorithm was employed to optimize the weight of the combined model. As a case study, the developed forecasting system was used to forecast the wind speed and wind power over 10 min from four different sites. The point forecasting and interval forecasting results revealed that the developed forecasting system exceeds all other models with respect to forecasting precision and stability. Thus, the developed system is extremely useful for enhancing forecasting precision and is a reasonable and valid tool for use in intelligent grid programming.

A review of wind speed and wind power forecasting with deep neural networks

The use of wind power, a pollution-free and renewable form of energy, to generate electricity has attracted increasing attention. However, intermittent electricity generation resulting from the random nature of wind speed poses challenges to the safety and stability of electric power grids when wind power is integrated into grids on large scales. Therefore, accurate forecasting of wind speed and wind power (WS/WP) has gradually taken on a key role in reducing wind power fluctuations in system dispatch planning. With the development of artificial intelligence technologies, especially deep learning, increasing numbers of deep learning-based models are being considered for WS/WP forecasting due to their superior ability to deal with complex nonlinear problems. This paper comprehensively reviews the various deep learning technologies being used in WS/WP forecasting, including the stages of data processing, feature extraction, and relationship learning. The forecasting performance of some popular models is tested and compared using two real-world wind datasets. In this review, three challenges to accurate WS/WP forecasting under complex conditions are identified, namely, data uncertainties, incomplete features, and intricate nonlinear relationships. Moreover, future research directions are summarized as a guide to improve the accuracy of WS/WP forecasts.

An ultra-short-term wind speed forecasting model based on time scale recognition and dynamic adaptive modeling

The forecast of wind speed is prerequisite for wind power prediction, which is one of the most effective means of promoting wind power absorption. However, when modeling for wind speed sequences with different fluctuations, most existing researches ignore the influence of time scale of wind speed fluctuation period, let alone the low compatibility between training and testing samples that severely limit the training performance of forecasting model. To improve the accuracy of wind speed and wind power forecasting, an ultra-short-term wind speed forecasting model based on time scale recognition and dynamic adaptive modeling is proposed in this paper. First, a series of wind processes are divided from the historical wind speed sequence according to the natural variation characteristics of wind speed. Second, we divide all the wind processes into two patterns based on their time scale, and an SVC model with input features extracted from meteorological data is built to identify the time scale of the current wind process. Third, for a specifically identified wind process, the complex network algorithm is applied in data screening to select high compatible training samples to train the forecast model dynamically for current input. Simulation indicates that the proposed approach presents higher accuracy than benchmark models using the same forecasting algorithms but without considering the time scale and data screening.

Ultra-short-term Wind Power Prediction Model Based on VMD Decomposition and LSTM

With the rising demand for environmental protection, wind energy has made considerable progress as a clean energy source. Due to the uncertainty of wind power generation, accurate forecasting of wind power is conducive to achieving stable grid-connected control of wind turbines. This paper proposes an ultra-short-term wind power forecasting model composed of variational modal decomposition (VMD) and long short-term memory (LSTM) networks. Obtain the recorded historical wind speed and historical power from the wind field data acquisition and monitoring control system (SCADA), and use VMD to decompose and reconstruct the historical wind speed of the wind field to form 5 training data sets. Using historical power, historical wind speed, and predicted wind speed as input, wind power forecasts for the next 4 hours (15 minutes apart) are made. In order to evaluate the prediction ability of the proposed model, two algorithms, LSTM and XGBoost, are used for prediction. Taking root mean square error (RMSE) and mean absolute error (MAE) as evaluation indicators, the results show that the prediction accuracy of wind power is significantly improved after VMD processing, and the performance of LSTM is better than XGBoost in the two evaluation indicators.

Short-Term Wind Power Forecasting Using Mixed Input Feature-Based Cascade-connected Artificial Neural Networks

Accurate short-term wind power forecasting is crucial for the efficient operation of power systems with high wind power penetration. Many forecasting approaches have been developed in the past to forecast short-term wind power. Artificial neural network-based approaches (ANNs) have become one of the most effective and popular short-term wind speed and wind power forecasting approaches in recent years. However, most researchers have used only historical data from a specific station to train the ANNs without considering meteorological variables from many neighboring stations on the forecasting performance. Using additional meteorological variables from neighboring stations can contribute valuable surrounding information to the forecasting model of the target station and improve ANNs performance. In this paper, a mixed input features-based cascade-connected artificial neural network (MIF-CANN) is used to train input features from many neighboring stations without encountering overfitting issues caused by many input features. Multiple ANNs train different combinations of input features in the first layer of the MIF-CANN model to produce preliminary results, then cascading into the second phase of the MIF-CANN model as inputs. The performance of the proposed MIF-CANN model is compared with the ANNs-based spatial correlation models. Simulation results show that the proposed MIF-CANN has better performance than the ANNs-based spatial correlation models.

Employing power spectral density method for investigating atmospheric stability impacts on power generation of a wind farm

ABSTRACT Theoretically, the output power of a wind farm is mainly under influence of wind speed as an atmospheric parameter. However, there are other factors that impact wind power generation. Atmospheric stability is one these features. The main contribution of this study is to investigate the impact of atmospheric stability on efficiency of Manjil wind farm in Iran for summer of 2004. For accomplishing the task, wind speed distributions and power generation amounts of the farm for three different months of the summer were investigated. The investigation showed that in the first and second month wind speed values for 82.3 and 87% of the times were more than 8.8 m/s, respectively. However, power generation in the first month was 4.96% more than the second one. Considering the impact of atmospheric stability on power production, the stability was inspected by employing wind shear exponent as a common metric and Power Spectral Density method as a rather new approach. The results showed that in the first month the atmosphere was instable for 91% of the times. This percent for the second month was 82%. The results support the idea that atmospheric instability has positive impact on efficiency of the wind farm.

Assessment of wind energy resources in the urat area using optimized weibull distribution

Accurate assessment and analysis of wind energy resources are the key to wind farm planning. In this study, we evaluate the wind energy resource potential in the Urat area of Inner Mongolia, China, based on data measured from four 70–100 m/s wind tower. We first model the wind speed by the Weibull distribution where the parameters are estimated by various numerical and intelligent optimization methods. In particular, we introduce the Cross Entropy Method (CEM) and find that it outperforms other methods. Based on the Weibull model, we then evaluate six wind energy metrics for the Urat area. It is shown that the annual average wind speed and wind power density are up to 9.33 m/s and 941 W/m2, respectively, with over 7000 h of annual available time. The wind is relative stable for all seasons except for a mild reduction in speed and a change of direction in summer. The low power production cost and rapid payback period make the Urat area a promising location for deploying wind farms.

Proactive maintenance of small wind turbines using IoT and machine learning models

ABSTRACT Wind is one of the most important natural resources from which we can generate power through wind turbines, Presently the wind turbines may be small or large but the output power generated is used for various purposes. There may be a possibility of wear and tear of the turbines which may lead to physical damage. There are no proper mechanisms available for monitoring the turbines from remote locations through wireless mode. The implementation of live remote monitoring and intelligent condition monitoring techniques reduces the downtime and would increase the lifetime of the turbines. This paper proposes a novel smart and proactive maintenance system that would aid in diagnosing the major faults with the wind turbines and a prediction analysis tool that would forecast the generation status of the small wind turbines. A Sensor based IoT system (SBIS) will help to monitor the significant parameters of the wind turbines which determine its working conditions like wind speed, vibration, temperature, and output power. A combined approach of Machine Learning techniques for SBIS have been used for proactive maintenance of small wind turbines. The diagnosis part of the system takes input from the sensors and uses a cloud platform for predictive analysis. The machine learning algorithms like Linear Regression (LR), Support Vector Machine (SVM), Optimized Artificial Neural Network (OANN), and XG Boost (XGB) are applied and the results are summarized for prediction. The results of the algorithms are compared for accuracy of the sensed data and it is observed that the OANN algorithm is better performing for proactive maintenance and power prediction of the small wind turbines.

High-resolution mesoscale wind-resource assessment of Fiji using the Weather Research and Forecasting (WRF) model

This study presents a high-resolution mesoscale wind-resource assessment of the small island developing state (SIDS) of Fiji using a 10-year simulation of the Weather Research and Forecasting (WRF) model with convection-permitting resolution. Our analysis evaluates the wind speed and Weibull distributions, diurnal and seasonal wind speed patterns, resource maps of annual and seasonal wind speed, power density, model statistical analysis and interannual wind speed variability. The results reveal that the WRF-model simulated wind resource parameters are in good agreement with observations at 24 existing weather stations. At 55 m above ground, the annual mean wind speed and wind power density varies from 1.5 m/s to 8 m/s and 50 W/m2 to 300 W/m2, respectively, for onshore land areas. Higher wind speeds are observed during austral winter than in austral summer. Forty high wind-resource areas are identified in this study, which were previously unknown. This indicates that there is potential for utility-scale wind power generation at selected locations with wind speed and power density greater than 6.4 m/s and 300 W/m2 (NREL, Wind Power Class 3). An estimated 1000 MW theoretical potential installed capacity is available for utility-scale wind power applications on Viti Levu and Vanua Levu.

Observation and Simulation of Wind Speed and Wind Power Density over Bac Lieu Region

In this study, the WRF (Weather Research and Forecasting) model was used to simulate and investigate diurnal and annual variations of wind speed and wind power density over Southern Vietnam at 2‐km horizontal resolution for two years (2016 and 2017). The model initial and boundary conditions are from the National Centers for Environmental Prediction (NCEP) Final Analyses (FNL). Observation data for two years at 20 m height at Bac Lieu station were used for model bias correction and investigating diurnal and annual variation of wind speeds. The results show that the WRF model overestimates wind speeds. After bias correction, the model reasonably well simulates wind speeds over the research area. Wind speed and wind power density show much higher values at levels of 50–200 m above ground levels than near ground (20 m) level and significantly higher near the coastal regions than inland. Wind speed has significant annual and diurnal cycles. Both annual and diurnal cycles of wind speeds were well simulated by the model. Wind speed is much stronger during daytime than at nighttime. Low-level wind speed reaches the maximum at about 14 LT to 15 LT when the vertical momentum mixing is highly active. Wind speeds over the eastern coastal region of Southern Vietnam are much stronger in winter than in summer due to two main reasons, including (1) stronger large-scale wind speed in winter than in summer and (2) funnel effect creating a local maximum wind speed over the nearshore ocean which then transports high-momentum air inland in winter.

Deep Learning for Wave Energy Converter Modeling Using Long Short-Term Memory

Accurate forecasts of ocean waves energy can not only reduce costs for investment, but it is also essential for the management and operation of electrical power. This paper presents an innovative approach based on long short-term memory (LSTM) to predict the power generation of an economical wave energy converter named “Searaser”. The data for analysis is provided by collecting the experimental data from another study and the exerted data from a numerical simulation of Searaser. The simulation is performed with Flow-3D software, which has high capability in analyzing fluid–solid interactions. The lack of relation between wind speed and output power in previous studies needs to be investigated in this field. Therefore, in this study, wind speed and output power are related with an LSTM method. Moreover, it can be inferred that the LSTM network is able to predict power in terms of height more accurately and faster than the numerical solution in a field of predicting. The network output figures show a great agreement, and the root mean square is 0.49 in the mean value related to the accuracy of the LSTM method. Furthermore, the mathematical relation between the generated power and wave height was introduced by curve fitting of the power function to the result of the LSTM method.

Assessment of wind energy potential over India using high-resolution global reanalysis data

An assessment of wind energy potential based on wind speed data over the Indian subcontinent has been made using high spatio-temporal resolution global reanalysis for the period from 1979 to 2018. Regions of high wind speed exceeding 4.5 m/s are identified over West Rajasthan, West Gujarat, Saurashtra and Kutch, Central Maharashtra, Interior Karnataka, and Rayalaseema. Threshold wind speeds are noted to occur during the daytime, and during the summer months from May through September. Wind speeds and the spatial extent of threshold winds increase rapidly with height below 40 m and then gradually up to 100 m. The wind power density is highest between 50 and 80 m, with the potential highest over Gujarat, Kutch, and Interior Karnataka and moderate over Saurashtra and Rayalaseema. This study also notifies that offshore wind potential is higher than over land, and most of the western parts of India are congenial for low wind farming. The present study clearly delineates wind speed distributions and wind power productivity regions over the entire Indian subcontinent. The results would provide authentic wind speed and wind power potential information that would be useful for the industries, government agencies, and industries concerning wind harness over India.

Impact of Covid-19 pandemic on electricity demand in the UK based on multivariate time series forecasting with Bidirectional Long Short Term Memory

Due to lockdown measures taken by the UK government during the Coronavirus disease 2019 pandemic, the national electricity demand profile presented a notably different performance. The Coronavirus disease 2019 crisis has provided a unique opportunity to investigate how such a landscape-scale lockdown can influence the national electricity system. However, the impacts of social and economic restrictions on daily electricity demands are still poorly understood. This paper investigated how the UK-wide electricity demand was influenced during the Coronavirus disease 2019 crisis based on multivariate time series forecasting with Bidirectional Long Short Term Memory, to comprehend its correlations with containment measures, weather conditions, and renewable energy supplies. A deep-learning-based predictive model was established for daily electricity demand time series forecasting, which was trained by multiple features, including the number of coronavirus tests (smoothed), wind speed, ambient temperature, biomass, solar & wind power supplies, and historical electricity demand. Besides, the effects of Coronavirus disease 2019 pandemic on the Net-Zero target of 2050 were also studied through an interlinked approach.

Multi-step real-time wind speed prediction based on convolution memory network

Multi-step real-time prediction based on the spatial correlation of wind speed is a research hotspot for large-scale wind power grid integration, and this paper proposes a multi-location multi-step wind speed combination prediction method based on the spatial correlation of wind speed. The correlation coefficients were determined by gray relational analysis for each turbine in the wind farm. Based on this, timing-control spatial association optimization is used for optimization and scheduling, obtaining spatial information on the typical turbine and its neighborhood information. This spatial information is reconstructed to improve the efficiency of spatial feature extraction. The reconstructed spatio-temporal information is input into a convolutional neural network with memory cells. Spatial feature extraction and multi-step real-time prediction are carried out, avoiding the problem of missing information affecting prediction accuracy. The method is innovative in terms of both efficiency and accuracy, and the prediction accuracy and generalization ability of the proposed method is verified by predicting wind speed and wind power for different wind farms.

Prediction of Atmospheric Turbulence Characteristics for Surface Boundary Layer using Empirical Spectral Methods

Abstract Atmospheric turbulence is an unsteady phenomenon found in nature and plays significance role in predicting natural events and life prediction of structures. In this work, turbulence in surface boundary layer has been studied through empirical methods. Computer simulation of Von Karman, Kaimal methods were evaluated for different surface roughness and for low (1%), medium (10%) and high (50%) turbulence intensities. Instantaneous values of one minute time series for longitudinal turbulent wind at mean wind speed of 12 m/s using both spectra showed strong correlation in validation trends. Influence of integral length scales on turbulence kinetic energy production at different heights is illustrated. Time series for mean wind speed of 12 m/s with surface roughness value of 0.05 m have shown that variance for longitudinal, lateral and vertical velocity components were different and found to be anisotropic. Wind speed power spectral density from Davenport and Simiu profiles have also been calculated at surface roughness of 0.05 m and compared with k−1 and k−3 slopes for Kolmogorov k−5/3 law in inertial sub-range and k−7 in viscous dissipation range. At high frequencies, logarithmic slope of Kolmogorov −5/3rd law agreed well with Davenport, Harris, Simiu and Solari spectra than at low frequencies.