Spatially transferable machine learning wind power prediction models: [formula omitted]logit random forests

Due to the features of being fluctuant, intermittent, and stochastic of wind power, interconnection of large capacity wind farms with the power grid will bring about impact on the safety and stability of power systems. Based on the real-time wind power data, wind power prediction model using Elman neural network is proposed. At the same time in order to overcome the disadvantages of the Elman neural network for easily fall into local minimum and slow convergence speed, this paper put forward using the GA algorithm to optimize the weight and threshold of Elman neural network. Through the analysis of the measured data of one wind farm, shows that the forecasting method can improve the accuracy of the wind power prediction, so it has great practical value.

The fluctuations of wind power due to the uncontrollability and uncertainty of wind speed cause great do harm to the grid power. To smooth the power to grid from wind power system, the energy storage system is paralleled in the grid-connected side of wind farm, and a real time rolling horizon optimization scheme based on wind power prediction is taken. The simulation is on the base on the real output power and predicted power of Muling wind farm, where the predicted wind power is obtained base on the Back Propagation Neutral Network. The results of the simulation show that the rolling horizon optimization smooth scheme based on predicted wind power is superior to the average-power smooth scheme.

Data mining and wind power prediction: A literature review

Wind power prediction accuracy is closely related to the quality of wind power data. However, most of traditional prediction methods do not consider the cybersecurity of raw data. So, it would be important to study the prediction accuracy under false data injection attack (one type of cybersecurity attacks) and its influence on the economy and security of power system. In this paper, very short-term forecasting approach of wind power generation is studied for choosing the optimal prediction model. Different scenarios of false data injection attack are then simulated upon wind power data. Wind power prediction accuracy after being affected by false data injection is observed. The robustness of wind power prediction model is evaluated. Finally, wind power prediction information is applied to day-ahead unit commitment to evaluate the impact of cybersecurity of wind power forecasting on power system economy and security.

Wind power forecasting is a critical technology for promoting the effective integration of wind energy. To enhance the accuracy of wind power predictions, this paper introduces a novel wind power prediction model that considers the evolving relationships of multi-scale variables and temporal dependencies. In this paper, a multi-scale frequency decomposition module is designed to split the raw data into high-frequency and low-frequency parts. Subsequently, features are extracted from the high-frequency information using a multi-scale temporal graph neural network combined with an adaptive graph learning module and from the low-frequency data using an improved bidirectional temporal network. Finally, the features are integrated through a cross-attention mechanism. To validate the effectiveness of the proposed model, extensive comprehensive experiments were conducted using a wind power dataset provided by the State Grid. The experimental results indicate that the MSE of the model proposed in this paper has decreased by an average of 7.1% compared to the state-of-the-art model and by 48.9% compared to the conventional model. Moreover, the improvement in model performance becomes more pronounced as the prediction horizon increases.

With the gradual warming of the global climate, wind energy, as one of the representatives of clean energy, has been rapidly developed in countries worldwide. However, wind power generation has strong intermittency and variability, constantly impacting the stability of the power grid. In order to improve the accuracy of wind power prediction, this paper proposes a BP neural network model optimized by genetic algorithm (GA) for wind power prediction. This model uses GA to optimize the initial weights and thresholds of the BP neural network to improve the computational accuracy and convergence speed of the model, effectively avoiding the problem of a single BP neural network model easily falling into local minima and the slow training speed of GA algorithm. By preprocessing and normalizing meteorological data such as weather, temperature, and wind direction, a wind power prediction model is constructed, and the prediction results show that the model effectively improves the computational accuracy and convergence speed compared to the traditional BP neural network model. This result proves the feasibility of using GA to optimize BP neural network for wind power prediction in this paper.

The accuracy of short-term wind power forecast is important for the power system operation. Based on the real-time wind power data, a wind power prediction model using wavelet neural network (WNN) is proposed. In order to overcome such disadvantages of WNN as easily falling into local minimum, this paper put forward using Particle Swarm Optimization (PSO) algorithm to optimize the weight and threshold of WNN. It’s advisable to use Support Vector Machine (SVM) to comparatively do prediction and put two outcomes as input vector for Generalized Regression Neural Network (GRNN) to do nonlinear combination forecasting. Simulation shows that combination prediction model can improve the accuracy of the short-term wind power prediction.

Wind power generation is currently one of the most promising power generation technologies. It is particularly important to improve the prediction accuracy of wind power output, which can effectively reduce the impact on the grid when wind power is connected to the grid. Based on the fractal model, this paper integrates it with the wind power prediction model, and combines the custom K nearest neighbor algorithm to evaluate the prediction effect using multi-dimensional indicators. Finally, taking the data of a wind farm in northwest china as an example, compared it with the prediction model of random forest, support vector machine and gradient boosting decision tree prediction model to verify the effectiveness of the prediction algorithm in this paper.

In the 21st century, energy demand has become a global problem, China's wind power energy jumps forward from start to large-scale development and its use also shows the characteristic of diversified development and thus requires the research of its related technology and personnel to ensure that the new energy wind power industry can further develop healthily.Based on this, this paper study from research on the development prediction and evaluation model of new energy wind power.The continued development of new energy wind power industry, on one hand, has a higher requirement of technology and needs some technically sophisticated technical personnel, on the other hand, the continuous improvement of technology level is bound to promote the continuous development of the biotechnology industry.So according to the current new wind energy development needs, study, evaluate and forecast new energy technologies talent demand and skill progress levels, which has extremely important practical significance for the development of new energy wind power.

To reduce the influence of the random fluctuation on wind power prediction, a new ultra-short-term wind power prediction model, based on wavelet decomposition (WD), variational mode decomposition (VMD), and least-squares support vector machine (LSSVM), is proposed in this paper. The method is based on the double decomposition and LSSVM, where the wind power sequence is decomposed by WD into low- and high-frequency components, which are further decomposed by VMD to obtain many modal components with tendency and periodicity. Multiple LSSVM prediction models are then established with historical wind power data and weather data as the inputs to obtain the predicted values of the multiple modal components. The final predicted values of wind power are achieved by data fusion of outputs of these LSSVM models. The experimental results show that the MAPE (mean absolute percentage error) of the combined prediction model is 4.66%, which is the best compared with nine benchmark models. This demonstrates the high performance of the proposed WD-VMD-LSSVM model for short-term prediction of wind power.

The demand on energy supply increases continuously as economies develop and world population grows. Most of the energy in the world comes from the fossil fuels and the increase in energy demand accelerates the rate of depleting the fossil fuel resources. As fossil fuel is finite and also due to the environmental considerations, renewable energy (RE) resources have been actively developed as the alternative sources for electricity generation. Renewable sources used for large-scale electricity production can include wind, photovoltaic (PV), waves and so on. Among all the RE resources, wind power generation is witnessing rapid developments in recent years. Unfortunately, wind tends to be unsteady. The fluctuating and uncertain output powers from wind farms can be problematic, especially if there is a high penetration level of the renewable source in a grid system. The fluctuations may lead to excessive voltage variations and unacceptable degree of frequency deviations. Moreover, the unsteady and uncertain input wind power can cause firm power dispatch commitment from wind farms a difficult task. Expensive generation reserves have to be provided for grid system to ensure adequate level of system security and reliability is guaranteed. In fact, power generation from wind farms has often been excluded in the dispatch planning of the grids. Unless other viable techniques can be found, the above issues can constitute major impediment to the successful large-scale integration of wind power generation into grid systems. In this connection, the use of energy storage system (ESS) is one possible solution to mitigate the negative impacts of the unsteady wind power on electricity supply systems and to improve on the dispatchability of the wind power. In this investigation, in order to smoothen the fluctuations in the wind and to realize the dispatch planning of the wind power in a way similar to that of conventional generators, it is proposed that a three-level ESS is incorporated in a large-scale wind power generation scheme. In essence, at cluster-level of wind turbine generators (WTG), the solution involves having the high- and mid-frequency components of the aggregated wind power routed to supercapacitors (SC) and battery banks respectively of a battery-supercapacitor hybrid energy storage system (BSHESS) by a high-pass filter (HPF) and a band-pass filter (BPF). It then results in the smoothening of the wind power harnessed by the cluster. In this thesis, the BSHESS control scheme to reduce the wind power perturbations has been shown. A statistical method is also developed to determine the capacities of the minimum-cost BSHESS to meet the power smoothing objective at pre-specified probability level. Application of the proposed design approach is demonstrated using data obtained from an existing wind farm. As for dispatch planning, attention is to use a low-pass filter (LPF) to extract the low-frequency intrinsic mode and residue functions of wind power. The relatively slow-changing characteristics of the low-frequency wind power components, which contain most of the energy in the wind power, allow accurate forecasts of the power components to be obtained using artificial neural network (ANN). Dispatch planning of the wind power is then realized through buffering the low-frequency power fluctuations by a proposed power flows control strategy applied to a pumped hydroelectric storage (PHS) system. Focusing on low-frequency wind power for dispatch planning is a distinct advantage over the dispatch planning methods proposed by other researchers. Furthermore, a statistical method to determine the power and energy capacities of the PHS is included. The efficacy of the developed approach to dispatch planning is again illustrated based on data obtained from an existing wind farm. In all the above works, suitable cut-off frequencies for the HPF, BPF and LPF are derived based on the developed concept of minimum overlap energy and the outcome of empirical mode decomposition (EMD) analysis. EMD method is used to gain insights into the frequency-time characteristics of wind power. In this manner, the wind power can then be grouped into high-, mid- and low-frequency bands.

Wind power has been valued by countries for its renewability and cleanness and has become most of the focus of energy development in all countries. However, due to the uncertainty and volatility of wind power generation, making the grid-connected wind power system presents some serious challenges. Improving the accuracy of wind power prediction has become the focus of current research. Therefore, this paper proposes a combined short-term wind power prediction model based on T-LSTNet_markov to improve prediction accuracy. First, perform data cleaning and data preprocessing operations on the original data. Second, forecast using T-LSTNet model in original wind power data. Finally, calculate the error between the forecast value and the actual value. The k-means++ method and Weighted Markov process are used to correct errors and to get the result of the final prediction. The data that are collected from a wind farm in Inner Mongolia Autonomous Region, China, are selected as a case study to demonstrate the effectiveness of the proposed combined models. The empirical results show that the prediction accuracy is further improved after correcting errors.

This paper proposes a short‐term wind power prediction model based on the improved Sparrow Search Algorithm (SSA) and Extreme Learning Machine(ELM) for anomalous wind power information from wind farms. The objective is to enhance the accuracy of short‐term wind power prediction. The model employs the extraction of features utilizing raw wind power history data from wind farms, in conjunction with the application of Variable Importance in Projection indices in Partial Least Squares (PLS‐VIP). As the ELM network model is susceptible to the influence of randomly generated input weights and thresholds at the outset of training, a solution is proposed whereby the input weights and thresholds of the ELM are optimized using SSA. The optimal weights and thresholds identified by SSA are then applied to the ELM model, thus forming the SSA‐ELM model. To address the limitations of traditional SSA, namely its susceptibility to local optimal solutions and poor global search ability, an improved SSA‐ELM algorithm is proposed. The improved SSA‐ELM algorithm introduces chaotic sequences and an exchange learning strategy to the original SSA. The rationale behind incorporating chaotic sequences is to enhance the quality of the initial solution, ensuring a more uniform distribution of sparrow positions and, consequently, a more diverse sparrow population. This, in turn, enables the algorithm to achieve a more effective global search capability through the utilization of the exchange learning strategy. Subsequently, all the data are fed into the SSA‐ELM model for prediction purposes. The simulation results demonstrate that the model exhibits enhanced prediction accuracy and improved practical applicability in wind power prediction. © 2025 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.

With the large-scale wind integration into power grid, the intermittent and stochastic nature of wind power possess a threat to the safety of power grid, and wind power prediction has become one of the most important solutions to the current grid connection problem. However, the prediction error in point prediction of wind power cannot be ignored, so it is necessary to make probability prediction of wind power and improve reliable information to the power grid dispatch department. In this paper, by combing with the wavelet decomposition technology and the long short-term memory (LSTM) network, an ultra short-term probability model is proposed. Firstly, the original time series are smoothed by wavelet decomposition technology, then the LSTM network prediction model of each sub-series sample is developed. The Gaussian distribution function of the prediction error is obtained by using the maximum likelihood estimation method, and finally the probability interval prediction of the future wind power in four hours could be realized. Finally, by using the data collected from a wind farm in northeast china, a numerical example is presented to illustrate the usefulness of the proposed method, which show that combining wavelet decomposition with deep learning method can improve the accuracy of prediction, improve the interval reliability of probability prediction, and enhance the generalization ability of the model.

Large-scale wind power grid connection increases the uncertainty of the power system, which reduces the economy and security of power system operations. Wind power prediction technology provides the wind power sequence for a period of time in the future, which provides key technical support for the reasonable development of the power generation plan and the arrangement of spare capacity. For large-scale wind farm groups, we propose a cluster model of wind power prediction based on multi-task learning, which can directly output the power prediction results of multiple wind farms. Firstly, the spatial and temporal feature matrix is constructed based on the meteorological forecast data provided by eight wind farms, and the dimensionality of each attribute is reduced by the principal component analysis algorithm to form the spatial fusion feature set. Then, a network structure with bidirectional gated cycle units is constructed, and a multi-output network structure is designed based on the Multi-gate Mixture-of-Experts (MMoE) framework to design the wind power group prediction model. Finally, the data provided by eight wind farms in Jilin, China, was used for experimental analysis, and the predicted average normalized root mean square error is 0.1754, meaning the prediction precision meets the scheduling requirement, which verifies the validity of the wind power prediction model.