Wind Forecasts Research Articles

Integration of path planning and following control for the stratospheric airship with forecasted wind field data

The absence of real-time airspeed sensors, which was more often ignored in previous studies, and low dynamic characteristics render stratospheric airship control challenging. This study creatively overcomes the aforementioned problems in an integrated path planning and following control scheme using forecasted wind field data. Herein, an efficient and practicable path planning algorithm is designed. Further, a smooth vector field guidance law is proposed for solving the problem of complex path following. Subsequently, an event-triggered neural network-based adaptive tracking controller is designed, considering the wind forecast error influence. Finally, these three parts are organically integrated to achieve autonomous flight. The stability of the closed-loop system and the exclusion of Zeno behavior are rigorously proved. The simulation results reveal that the convergence rate is 63.8% improved, essentially exhibiting better optimization, the tracking errors are eliminated within 80 s, and 99.4% control input updating times are saved.

Hierarchical model predictive control for islanded and grid-connected microgrids with wind generation and hydrogen energy storage systems

This paper presents a novel energy management strategy (EMS) to control a wind-hydrogen microgrid which includes a wind turbine paired with a hydrogen-based energy storage system (HESS), i.e., hydrogen production, storage and re-electrification facilities, and a local load. This complies with the mini-grid use case as per the IEA-HIA Task 24 Final Report, where three different use cases and configurations of wind farms paired with HESS are proposed in order to promote the integration of wind energy into the grid. Hydrogen production surpluses by wind generation are stored and used to provide a demand-side management solution for energy supply to the local and contractual loads, both in the grid-islanded and connected modes, with corresponding different control objectives. The EMS is based on a hierarchical model predictive control (MPC) in which long-term and short-term operations are addressed. The long-term operations are managed by a high-level MPC, in which power production by wind generation and load demand forecasts are considered in combination with day-ahead market participation. Accordingly, the hydrogen production and re-electrification are scheduled so as to jointly track the load demand, maximize the revenue through electricity market participation and minimize the HESS operating costs. Instead, the management of the short-term operations is entrusted to a low-level MPC, which compensates for any deviations of the actual conditions from the forecasts and refines the power production so as to address the real-time market participation and the short time-scale equipment dynamics and constraints. Both levels also take into account operation requirements and devices’ operating ranges through appropriate constraints. The mathematical modeling relies on the mixed-logic dynamic (MLD) framework so that the various logic states and corresponding continuous dynamics of the HESS are considered. This results in a mixed-integer linear program which is solved numerically. The effectiveness of the controller is analyzed by simulations which are carried out using wind forecasts and spot prices of a wind farm in center-south of Italy.

Modeling chronic oil pollution from ships

Stochastic simulations of virtual oil spills from ships were performed for the Adriatic Sea over 2017–2020, applying the European Marine Observation and Data Network vessel densities as a proxy for starting locations of operational spillage. The MEDSLIK-II oil spill model was run using high-resolution currents provided by the Copernicus Marine Service and the European Centre for Medium-Range Weather Forecasts winds. Chronic exposure to operational oil spills was reported in terms of hazard indices for five vessel groups: pleasure and passenger ships, cargo and service vessels, the fishing fleet, tankers, and other ships. The northernmost Adriatic expectedly showed the highest hazard values, including the areas of Trieste and Venice, where cargo and service ships were the dominant polluters. The Croatian coastal waters were more chronically polluted than the Italian coastal waters; the predominant contribution was from coastwise pleasure and passenger ships.

Machine Learning–Adjusted WRF Forecasts to Support Wind Energy Needs in Black Start Operations

Abstract The push for increased capacity of renewable sources of electricity has led to the growth of wind-power generation, with a need for accurate forecasts of winds at hub height. Forecasts for these levels were uncommon until recently, and that, combined with the nocturnal collapse of the well-mixed boundary layer and daytime growth of the boundary layer through the levels important for energy generation, has contributed to errors in numerical modeling of wind generation resources. The present study explores several machine learning algorithms to both forecast and correct standard WRF Model forecasts of winds and temperature at hub height within wind turbine plants over several different time periods that are critical for the anticipation of potential blackouts and aiding in black start operations on the power grid. It was found that mean square error for day-2 wind forecasts from the WRF Model can be improved by over 90% with the use of a multioutput neural network, and that 60-min forecasts of WRF error, which can then be used to adjust forecasts, can be made with an LSTM with great accuracy. Nowcasting of temperature and wind speed over a 10-min period using an LSTM produced very low error and especially skillful forecasts of maximum and minimum values over the turbine plant area.

Using Tianwen-1 Differential One-way Range Signals to Probe Corotating Interaction Regions: Case Studies

The study reports the first detection of corotating interaction regions (CIRs) in the solar wind by the differential one-way range (DOR) signals transmitted by Tianwen-1. Differential phase delay is calculated for the DOR signals. Variations of the total electron content (TEC) along the signal path between Tianwen-1 and the ground stations are then derived from changes in the differential phase delay. Large TEC variations are found for 2021 February 11, 18, and 19 even after the influence from the Earth’s ionosphere has been removed. In situ solar wind measurements and the corresponding simulation results further show that the TEC variations are caused by the CIRs that passed through the signal path on these days. The study demonstrates that CIR structures can be probed using deep-space spacecraft range signals. Such measurements can be used to constrain global solar wind forecast models.

Forecasting of Solar and Wind Resources for Power Generation

Solar and wind are now the fastest-growing power generation resources, being ecologically benign and economical. Solar and wind forecasts are significantly noteworthy for their accurate evaluation of renewable power generation and, eventually, their ability to provide profit to the power generation industry, power grid system and local customers. The present study has proposed a Prophet-model-based method to predict solar and wind resources in the Doomadgee area of Far North Queensland (FNQ), Australia. A SARIMA modelling approach is also implemented and compared with Prophet. The Prophet model produces comparatively less errors than SARIMA such as a root mean squared error (RMSE) of 0.284 and a mean absolute error (MAE) of 0.394 for solar, as well as a MAE of 0.427 and a RMSE of 0.527 for wind. So, it can be concluded that the Prophet model is efficient in terms of its better prediction and better fitting in comparison to SARIMA. In addition, the present study depicts how the selected region can meet energy demands using their local renewable resources, something that can potentially replace the present dirty and costly diesel power generation of the region.

Effects of assimilation of YOPP‐SH additional radiosonde observations on analyses and forecasts over Antarctica in austral summer

AbstractRadiosonde observations over Antarctica and the surrounding oceans were enhanced during the Year of Polar Prediction in the Southern Hemisphere (YOPP‐SH) summer Special Observing Period (SOP). Observing System Experiments (OSEs) were conducted in a continuous cycling framework using the Weather Research and Forecasting (WRF) Model and its data assimilation system. Routinely available observations were assimilated in the CTL (control) experiment, and special radiosonde observations from the YOPP‐SH SOP were additionally assimilated in the YOPP experiment. The results were compared to investigate the effects of additional radiosonde observations on analyses and forecasts over and around Antarctica. Verifications against ERA5 re‐analysis, radiosonde observations, and Automatic Weather Station (AWS) observations show overall positive effects of additional radiosonde observations. These positive effects are most noticeable in temperature at lower levels at earlier forecast lead times; afterward, wind forecast improvements at upper levels are the most noticeable. Although routine and special radiosonde observations are concentrated over the eastern and coastal regions of Antarctica (compared to the western and inland regions), the effects of the extra data spread in longitudinal and latitudinal directions; therefore, the effects on the forecasts are not limited to only the areas near the radiosonde observations. A case study reveals how cyclone forecasts are improved through the assimilation of the additional YOPP‐SH SOP radiosonde observations. This study provides insights into future observation strategies in Antarctica, such as horizontal/vertical observation locations, observation variables, and so forth to maximize effects of new observations on forecasts over Antarctica.

A Comprehensive Review on Machine Learning Techniques for Forecasting Wind Flow Pattern

The wind is a crucial factor in various domains such as weather forecasting, the wind power industry, agriculture, structural health monitoring, and so on. The variability and unpredictable nature of the wind is a challenge faced by most wind-energy-based sectors. Several atmospheric and geographical factors influence wind characteristics. Many wind forecasting methods and tools have been introduced since early times. Wind forecasting can be carried out short-, medium-, and long-term. The uncertainty factors of the wind challenge the accuracy of techniques. This article brings the general background of physical, statistical, and intelligent approaches and their methods used to predict wind characteristics and their challenges—this work’s objective is to improve effective data-driven models for forecasting wind-power production. The investigation and listing of the effectiveness of improved machine learning models to estimate univariate wind-energy time-based data is crucially the prominent focus of this work. The performance of various ML predicting models was examined using ensemble learning (ES) models, such as boosted trees and bagged trees, Support Vector Regression (SVR) with distinctive kernels etc. Numerous neural networks have recently been constructed for forecasting wind speed and power due to artificial intelligence (AI) advancement. Based on the model summary, further directions for research and application developments can be planned.

The impact of Aeolus winds on near-surface wind forecasts over tropical ocean and high-latitude regions

Abstract. To detect global wind profiles and improve numerical weather prediction (NWP), the European Space Agency (ESA) launched the Aeolus satellite carrying a spaceborne Doppler wind lidar in 2018. After the successful launch, the European Centre for Medium-Range Weather Forecasts (ECMWF) performed the observing system experiments (OSEs) to evaluate the contribution of Aeolus data to NWP. This study aims to assess the impact of Aeolus wind assimilation in the ECMWF model on near-surface (10 m height) wind forecasts over tropical ocean regions by taking buoy measurements for reference and over high-latitude regions by taking weather station data for reference for the year 2020. The assessments were conducted mainly through inter-comparison analysis. The results show that Aeolus data assimilation has a limited impact on sea surface wind forecasts for tropical regions when compared with buoy measurements. For the high-latitude regions in the Northern Hemisphere, Aeolus is able to improve near-surface wind forecasts. This positive impact is more evident as the forecast time step is extended, during the first half year of 2020 and during the winter months. In addition, the v component tends to benefit more from the Aeolus observations than the u component. For the Southern Hemisphere, a few error reductions are observed but exist randomly. Overall, this in situ data-based assessment expands our understanding of the role of Aeolus data assimilation with the global NWP model in predicting near-surface wind for tropical oceans and high-latitude regions.

Adaptive stepsize estimation based accelerated gradient descent algorithm for fully complex-valued neural networks

Nesterov accelerated gradient (NAG) method is an efficient first-order algorithm for optimization problems. To ensure the convergence, it usually takes a relatively conservative constant as the stepsize. However, the choice of stepsize has a great impact on the optimization process. In this paper, two adaptive stepsize estimation methods are proposed for complex-valued NAG algorithm for efficient training of fully complex-valued neural networks. The basic idea of the first one is to adaptively determine suitable stepsize by estimating the local smoothness constant of the loss function with the norm of approximate complex Hessian matrix. Its validity is theoretically analyzed by means of the decomposition of complex matrix. Furthermore, by introducing a new parameter design method for multi-step quasi-Newton condition, an improved stepsize estimation is presented. Experimental results on pattern recognition, channel equalization, wind forecasting and synthetic aperture radar (SAR) target classification demonstrate the effectiveness of the proposed methods.

Multi-model ensemble forecasting of 10-m wind speed over eastern China based on machine learning optimization

Wind substantially impacts human activity and electricity generation. Thus, accurately forecasting the short-term wind speed is of profound societal and economic significance. Based on 100 weather stations in eastern China, the authors first evaluate the performance of the 10-m wind forecast products from five operational forecast models. Among them, the Japan Meteorological Agency (JMA) model performs best in reducing the forecasting errors. Then, the authors establish a 10-m wind speed multimodel ensemble forecast based on the five numerical models’ outputs and machine learning methods, combining dynamic and statistical methods. Feature engineering and machine learning algorithm optimization are conducted for each site separately. The forecast performance of this method is compared to the JMA model and multimodel ensemble forecast by ridge regression at lead times of 24–96 h. The results demonstrate that the multimodel ensemble method based on machine learning optimization can reduce the forecast error of JMA by more than 39%, and the improvement in forecast skill is most evident in November. In addition, it performs better than the ensemble forecast by ridge regression.

Quantification of natural gas and other hydrocarbons from production sites in northern West Virginia using tracer flux ratio methodology.

Tracer flux ratio (TFR) methodology performed downwind of 15 active oil and natural gas production sites in Ohio County, West Virginia sought to quantify air pollutant emissions over two weeks in April 2018. In coordination with a production company, sites were randomly selected depending on wind forecasts and nearby road access. Methane (CH4), ethane (C2H6), and tracer gas compounds (acetylene and nitrous oxide) were measured via tunable infrared direct absorption spectroscopy. Ion signals attributed to benzene (C6H6) and other volatile gases (e.g., C7 - C9 aromatics) were measured via proton-transfer reaction time-of-flight mass spectrometry. Short-term whole facility emission rates for 12 sites are reported. Results from TFR were systematically higher than the sum of concurrent on-site full flow sampler measurements, though not all sources were assessed on-site in most cases. In downwind plumes, the mode of the C2H6:CH4 molar ratio distribution for all sites was 0.2, which agreed with spot sample analysis from the site operator. Distribution of C6H6:CH4 ratios was skew but values between 1 and 5 pptv ppbv-1 were common. Additionally, the aromatic profile has been attributed to condensate storage tank emissions. Average ratios of C7 - C9 to C6H6 were similar to other literature values reported for natural gas wells.

Risk-averse and flexi-intelligent scheduling of microgrids based on hybrid Boltzmann machines and cascade neural network forecasting

The future of energy flexibility in microgrids (MGs) is steering towards a highly granular control of the end-user customers. This calls for more highly accurate uncertainty forecasting and optimal management of risk and flexibility options. This paper presents a novel data-driven model to optimize the operation of MGs based on a risk-averse flexi-intelligent energy management system (RFEMS), considering the rising challenge of global climate change. It considers the presence of renewables, a diesel generator, and flexibility resources (FRs) containing a demand response program (DRP), distributed electric vehicles (EVs), and electric springs (ESs). In the first phase, the proposed model, by means of a novel hybrid deep-learning (DL) model, forecasts uncertain parameters associated with wind and solar generations, load demand, and day-ahead energy market price. The architecture of the proposed hybrid forecasting model is composed of several stacked restricted Boltzmann machines and a cascade neural network. In the second phase, the MG operator (MGO), based on the obtained uncertainty forecasting results, in the context of a hybrid risk-controlling model, optimizes the MG operation using the provided demand-side flexibility. The proposed optimization problem is linearized stochastic programming with robust concepts, subject to AC optimal power flow constraints, MG flexibility restrictions, and operating limits of local resources. Finally, the efficiency of the proposed RFEMS by using real German datasets on a 33-bus test MG is analyzed. Numerical results demonstrate the superior performance of the proposed forecasting model over several hybrid DL models. In particular, the root mean square error (RMSE) index for wind, solar, load, and price forecasting is improved by 53.35%, 73.24%, 80.24%, and 58.1%, respectively. Further analysis of the proposed RFEMS reveals that operating indices in two 33-bus and 69-bus test networks are significantly improved. It paves pathways to risk-averse, flexible, and economic operation of smart active distribution networks.

An integrated ultra short term power forecasting method for regional wind–pv–hydro

Renewable-based multi energy power system is the main trend for power system in the future. However, the randomness and fluctuation of wind and photovoltaic power, as well as the seasonality of hydropower, have an increasingly prominent impact on the stability of power system. Accurate power forecasting technology is the key to solve the above problems. At the same time, the output characteristics of heterogeneous energy sources are very different, and the existing forecasting methods are difficult to fully exploit their spatio-temporal correlation characteristics, which limits the improvement of prediction accuracy. In this paper, an integrated ultra short term power forecasting method for regional wind–pv–hydro is proposed, Firstly, it quantifies the differences and similarities among wind, pv and hydro in different scenarios based on empirical mode decomposition, and achieves the extraction of homogeneous features, on this basis, the integrated power forecasting models for wind, pv and hydro based on long short-term memory neural networks is constructed , and achieves regional-level integrated forecasting of wind, pv and hydro. The results show that the forecasting accuracy and modeling efficiency of the proposed integrated forecasting method are significantly improved compared with the traditional independent forecasting method, the forecasting accuracy is increased by 1%–3%, the modeling efficiency is increased by 6 times.

Toward a High-Resolution Wave Forecasting System for the Changjiang River Estuary

Based on a high-resolution unstructured SWAN model and GFS forecast wind, an operational wave forecasting system is conducted for the Changjiang River Estuary (CRE). The performance of the wave forecasting system is evaluated by comparing it with the altimeter observations and in situ wave buoys. The present operational system shows good accuracy in reproducing the seasonal and the synoptic-scale wave characteristics over the CRE. The forecasting capability in three different horizons, including 24 h, 48 h, and 72 h forecasts, is evaluated. Waves over the CRE exhibit distinct seasonal variability. Larger waves occur in both the summer and winter when typhoons and cold weather events affect the CRE. In contrast, waves with longer wave periods take place mainly in the wind transition seasons, i.e., the spring and fall, and the wave directions are more dispersed in these seasons. A seasonal varied forecasting capability is also revealed: better in the winter and spring than in the summer and fall and better during cold weather events than during typhoons. A cross comparison with the model analysis suggests that there is a systematic difference between wave measurements by Jason-3 and Sentinel-3A/3B. The significant wave height from Jason-3 compares best with the model analysis and forecasts and is systematically lower than Sentinel-3A/3B in lower wave conditions (<4 m) in the East China Sea. Substantial discrepancies exist among the three altimeters when the significant wave height exceeds 4 m, and further efforts are needed to discern their merits.

Evaluation of XGBoost vs. other Machine Learning models for wind parameters identification

Wind energy is one of the most promising renewable energies. But wind is a quite unstable resource due to its continuous variation and random nature. This uncertainty affects the production cost. Therefore, accurate forecasting of wind and energy is very interesting for energy markets. In this work, we test a recent and powerful intelligent technique, extreme gradient boosting (XGBoost), for wind prediction. The forecasting models of some wind features with XGBoost are compared with Support Vector Regression (SVR), Gaussian Process Regression (GPR) and Neural Networks (NN) models. Specifically, the three features predicted are the active power generated by the turbine, the wind speed, and the wind direction. The results conclude that these techniques are useful for wind and energy forecasting, with XGBoost being the most outstanding one, especially for short-term predictions.

Deep hybrid neural net (DHN-Net) for minute-level day-ahead solar and wind power forecast in a decarbonized power system

The need to reduce global carbon emissions has led to a significant increase in clean energy globally. While renewable energy penetration into energy grids and power systems is increasing in many countries, the intermittency and stochastic nature of wind and solar energy resources is still a major challenge. These can affect the safety, stability, and reliability of the energy grid. In existing works of literature, the forecast and prediction of wind energy, solar power, wind power, and solar energy with various models have been considered independently. However, with the rise in solar power and wind power penetration, there exists a gap in literature on the development of models that can simultaneously forecast solar and wind power production. In this paper, two deep hybrid neural networks (DHN-Net) models are developed for the simultaneous forecast of wind and solar power. The novelty of this study is further strengthened as a minute-level timestep is considered for the application of the models developed. The models are trained and tested with data collected from Zone 1 of four different power system operators in the USA. The two DHN-Net models are built on the foundation of artificial, convolutional, and recurrent neural networks (ANN, CNN, and RNN). Results from this study show that the two DHN-Nets can accurately forecast solar and wind power with an R-squared (r2) value of 0.9915, RMSE of 0.01920, and MAE of 0.00736 for data collected from PJM_Zone1. The DHN-Net models recorded a better performance when compared to the benchmark results in literature.

Using Deep Learning for Energy Modeling in Intelligent Structures with Decentralized Power Generation

In this research, we propose new methods to enhance the construction, designs, also hybrid renewable energy plant maintenance by incorporating battery energy storage systems. Our investigates focused by the three key aspects of renewable electricity energy monitoring: size optimization, improved combination forecasts of wind and solar energy production using deep reinforcement learning, and an advanced battery life forecasting model employing Hidden Markov Models and network analysis to ensure equilibrium throughout the battery’s lifespan, with the utilization of the ZSoft Fuzzy method. To address the renewable energy sources of dynamic nature connected with grid, we consider a load requirements significance and select a reliable operational mode accordingly. Additionally, we apply Z-Soft Ridge Regression (ZS-Fuzzy) analysis to investigate if matching the battery capacity with the system would enhance its service life. By resolving architectural challenges, we determine that an optimal mix of renewable energy sources, with the PV plant contributing 37.1% of the total, offers the best solution. Moreover, our findings indicate that a BESS capacity of 12.9% relative to the HREP capacity satisfies the design criteria effectively.

Short Term Wind Forecasting Using Machine Learning Models with Noise Assisted Data Processing Method

In order to address the complex and intermittent nature of wind, this thesis proposes an innovative approach to enhance the accuracy of short-term wind forecasting. By leveraging the strengths of different methods while mitigating their weaknesses, a robust hybrid model is developed. The methodology incorporates Empirical Mode Decomposition (EMD), a data-adaptive denoising technique, to break down the signals into meaningful components called Intrinsic Mode Functions (IMFs), along with a residue. However, EMD is known to suffer from the mode mixing problem, where different scales of signals are erroneously mixed within IMFs, causing signal intermittency. To overcome this challenge, Ensemble Empirical Mode Decomposition is introduced, utilizing an ensemble of white noise to establish a uniform reference frame in the time-frequency space. By doing so, the added noise effectively collates signal portions with similar scales into a single IMF. Subsequently, the IMFs and residue obtained from both EMD and Ensemble Empirical Mode Decomposition are fed into a Convolutional Neural Network (CNN). The performance of this hybrid model is then compared against benchmark models such as Bi LSTM and LSTM. Evaluation is conducted based on two critical factors: performance metrics including MSE, MAE, RMSE, MAPE, and loss, as well as the time required for testing. Through a comprehensive analysis of these factors, the superior performance of the hybrid model is determined, thereby enhancing the prospects of reliable wind forecasting.

Strong surface winds in Storm Eunice. Part 1: storm overview and indications of sting jet activity from observations and model data

Strong surface winds in Storm <i>Eunice</i>. Part 1: storm overview and indications of sting jet activity from observations and model data