The unsteady motion of the atmosphere incurs nonlinear and spatiotemporally coupled uncertainties in the wind power prediction (WPP) of multiple wind farms. This brings both opportunities and challenges to wind power probabilistic forecasting (WPPF) of a wind farm cluster or region, particularly when wind power is highly penetrated within the power system. This paper proposes an Improved Deep Mixture Density Network (IDMDN) for short-term WPPF of multiple wind farms and the entire region. In this respect, a deep multi-to-multi (m2m) mapping Neural Network model, which adopts the beta kernel as the mixture component to avoid the density leakage problem, is established to produce probabilistic forecasts in an end-to-end manner. A novel modified activation function and several general training procedures are then introduced to overcome the unstable behavior and NaN (Not a Number) loss issues of the beta kernel function. Verification of IDMDN is based on an open-source dataset collected from seven wind farms, and comparison results show that the proposed model improves the WPPF performance at both wind farm and regional levels. Furthermore, a laconic and accurate probabilistic expression of predicted power at each time step is produced by the proposed model.
In a wind farm some wind turbines may be affected by multiple upwind wakes. The commonly used approach in engineering to simulate the interaction effect of different wakes is to combine the single analytical wake model and the interaction model. The higher turbulence level and shear stress profile generated by upwind turbines in the superposed area leads to faster wake recovery. The existing interaction models are all analytical models based on some simple assumptions of superposition, which cannot characterize this phenomenon. Therefore, in this study, a mixing coefficient is introduced into the classical energy balance interaction model with the aim of reflecting the effect of turbulence intensity on velocity recovery in multiple wakes. An empirical expression is also given to calculate this parameter. The performance of the new model is evaluated using data from the Lillgrund and the Horns Rev I offshore wind farms, and the simulations agree reasonably with the observations. The comparison of different interaction model simulation results with measured data show that the calculation accuracy of this new interaction model is high, and the mean absolute percentage error of wind farm efficiency is reduced by 5.3% and 1.58%, respectively, compared to the most commonly used sum of squares interaction model.
To meet the increasing wind power forecasting (WPF) demands of newly built wind farms without historical data, physical WPF methods are widely used. The computational fluid dynamics (CFD) pre-calculated flow fields (CPFF)-based WPF is a promising physical approach, which can balance well the competing demands of computational efficiency and accuracy. To enhance its adaptability for wind farms in complex terrain, a WPF method combining wind turbine clustering with CPFF is first proposed where the wind turbines in the wind farm are clustered and a forecasting is undertaken for each cluster. K-means, hierarchical agglomerative and spectral analysis methods are used to establish the wind turbine clustering models. The Silhouette Coefficient, Calinski-Harabaz index and within-between index are proposed as criteria to evaluate the effectiveness of the established clustering models. Based on different clustering methods and schemes, various clustering databases are built for clustering pre-calculated CFD (CPCC)-based short-term WPF. For the wind farm case studied, clustering evaluation criteria show that hierarchical agglomerative clustering has reasonable results, spectral clustering is better and K-means gives the best performance. The WPF results produced by different clustering databases also prove the effectiveness of the three evaluation criteria in turn. The newly developed CPCC model has a much higher WPF accuracy than the CPFF model without using clustering techniques, both on temporal and spatial scales. The research provides supports for both the development and improvement of short-term physical WPF systems.
9 February 2020 in International Transactions on Electrical Energy Systems
As wind power is becoming a major energy source of power systems, the risk of power shortages due to its intermittent low power output is growing. Accurate forecasting of low wind power is crucial for mitigating these impacts. However, conventional methods struggle with few-sample issues due to the infrequent occurrence of low wind power, limiting accuracy improvements. To address this, a self-supervised pre-learning method is proposed to forecast low wind power occurrence period and output, leveraging the similarities and differences among low output samples to enhance forecasting accuracy. Low wind power output is decomposed into low wind power events (LWPE), which represent the occurrence timeframe, and low wind power processes (LWPP), which represent the power sequences. For LWPE forecasting, a siamese residual shrinkage network based on contrastive learning is introduced. This network pre-learns LWPE features from sample pairs to mitigate the impact of imbalanced sample distribution. For LWPP forecasting, a pattern recognition-based embedded forecasting framework is proposed, embedding typical LWPP fluctuations into the prediction network to improve fit under limited sample conditions. A case study on 3 wind farm clusters shows that this method improves LWPP forecasting accuracy from 84.99%-86.6% to 89.97%, outperforming traditional methods without pre-learning.
The quantitative evaluation of cluster wind power output volatility and source-load timing matching is vital to the planning and operation of the future power system dominated by new energy. However, the existing volatility evaluation methods of cluster wind power output do not fully consider timing volatility, or are not suitable for small sample data scenarios. Meanwhile, the existing source-load timing matching evaluation indicator ignores the impact of wind power permeability on the timing matching degree between wind power output and load. Therefore, the authors propose quantitative evaluation methods of cluster wind power output volatility and source-load timing matching in regional power grid. Firstly, the volatility-based smoothing coefficient is defined to quantitatively evaluate the smoothing effect of wind-farm cluster power output. Then, the source-load timing matching coefficient considering wind power permeability is proposed to quantitatively evaluate the timing matching degree of regional wind power output and load, and the corresponding function model of volatility-based smoothing coefficient and source-load timing matching coefficient is established. Finally, the validity and applicability of the proposed methods are verified by MATLAB software based on the actual power output of 10 wind farms and actual grid load in a certain grid dispatching cross-section of northeast China. The results demonstrated that the proposed volatility-based smoothing coefficient can accurately represent the smoothing effect of wind farm cluster power output while maintaining the volatility continuity of wind power output time series and without affect from the data sample size. The source-load timing matching coefficient can accurately characterize the difference in the timing matching degree between wind power output and grid load under different wind power permeability and the influence degree on grid load.