Different from power prediction for a single wind farm, the regional wind power prediction is to predict the total power of multiple wind farms located in the specific region. The regional wind power prediction involves more data which implicate abundant information on spatiotemporal correlations and nonlinearity. So that addressing the massive data and extracting representative features become the crucial issues to construct an effective regional wind power prediction model. This paper proposes a spatiotemporal quantile regression (SQR) algorithm to perform short-term nonparametric probabilistic prediction of regional wind power, incorporating the advantages of the hybrid neural network (HNN) and quantile regression (QR). In the approach, the high dimensional input data are reorganized into a feature graph that is ready for feature extraction by the HNN. Therefore, the advantages of HNN can be utilized to extract the representative features and construct nonlinear regression models. Meanwhile, by following the QR rules, the model obtains quantiles and perform probabilistic prediction. By properly addressing the explanatory variable selection issue, the approach provides a specific solution for regional wind power probabilistic prediction with the massive input data. Test results on a region with 10 wind farms demonstrate the effectiveness of the proposed approach.

Recent research has demonstrated that the rotor angle stability of a power system can be assessed by identifying the sign of the system's maximal Lyapunov exponent (MLE). A positive (negative) MLE implies unstable (stable) rotor angle dynamics. However, because the MLE may fluctuate between positive and negative values for a long time after a severe disturbance, determining the system stability is difficult when observing a positive or negative MLE without knowing its further fluctuation trend. In this paper, a new approach for online rotor angle stability assessment is proposed to address this problem. The MLE is estimated by a recursive least squares based method from real-time rotor angle measurements, and two critical parameters, the Theiler window, and the MLE estimation initial time step are carefully chosen to make sure that the calculated MLE curves present distinct features for different stability conditions. By using the proposed stability assessment criteria, the developed approach can provide a timely and reliable assessment of rotor angle stability. Extensive tests on the New England 39-bus system and the Northeast Power Coordinating Council 140-bus system verify the effectiveness of the proposed approach.

Different from individual wind power forecast and regional wind power forecast (RWPF), one of the most significant research articles for alleviating negative influence on power systems aims to estimate the generation of multiple wind farms in the specific region, which is a valuable complement of the wind power forecast. This article proposes a nonparametric probabilistic method for RWPF, a quantile regression neural network (QRNN), enhancing the abilities of nonlinear mapping and massive data dealing. On this basis, the deep quantile regression is proposed to improve the performance of the QRNN. In this approach, the local-connected method is applied to the input layer of the neural network for tackling the challenge of the massive data. A ramp function is designed to avoid multiple quantile curves crossing problem. To improve the model's generalization capability, a smoothing method is applied to the loss function for achieving differentiability everywhere. By properly constructing the model, the approach provides a specific solution for RWPF with the massive input information. The test results on a region with ten wind farms demonstrate the effectiveness of the proposed approach.

Wind generation forecast approaches are typically based on fixed model structures, which might fit well with existing data but fall short in forecasting out-of-sample data. The reason lies in that predetermined models cannot always capture the actual dynamic features of wind farms, resulting in unavoidable model errors. In this paper, a novel ultra-short-term wind generation forecast approach based on multivariate empirical dynamic modeling (MEDM) is proposed. The time series of wind generation and the explanatory variables are applied for attractor reconstruction according to Takens’ theorem. By this means, the dynamics of the original wind generation system can be accurately captured by the trajectories of the state variables in the reconstructed space. Then, using a simplex projection approach, wind generation can be directly forecasted with the trajectories. A unique feature of the proposed approach is that its effectiveness depends solely on the potential context hidden in the sequences of the state variables rather than the correctness of the predetermined models. Consequently, this approach can faithfully depict the dynamics of wind farms and improve the forecast accuracy. Case studies illustrate that the proposed approach can obtain advanced forecast results comparing with the benchmark methods.

Object detection is a fundamental computer vision task that simultaneously predicts the category and localization of the targets of interest. Recently one-stage (also termed “dense”) detectors have gained much attention over two-stage ones due to their simple pipeline and friendly application to end devices. Dense object detectors basically formulate object detection as dense classification and localization (i.e., bounding box regression). The classification is usually optimized by Focal Loss and the box location is commonly learned under Dirac delta distribution. A recent trend for dense detectors is to introduce an individual prediction branch to estimate the quality of localization, which facilitates the classification to improve detection performance. This paper delves into the representations of the above three fundamental elements: quality estimation, classification and localization. Three problems are discovered in existing practices, including (1) the inconsistent usage of the quality estimation and classification between training and inference, (2) the inflexible Dirac delta distribution for localization, and (3) the deficient and implicit guidance for accurate quality estimation. To address these problems, we design new representations for these elements. Specifically, we merge the quality estimation into the class prediction vector to form a joint representation, use a vector to represent arbitrary distribution of box locations, and extract discriminant feature descriptors from the distribution vector for more reliable quality estimation. The improved representations eliminate the inconsistency risk and accurately depict the flexible distribution in real data, but contain continuous labels, which is beyond the scope of Focal Loss. We then propose Generalized Focal Loss (GFocal) that generalizes Focal Loss from its discrete form to the continuous version for successful optimization. Extensive experiments demonstrate the effectiveness of our method, without sacrificing the efficiency both in training and inference. Based on GFocal, we construct a considerably fast and lightweight detector termed NanoDet under mobile settings, which is 1.8 AP higher, 2x faster and 6x smaller than scaled YoloV4-Tiny.

This article presents a model predictive control (MPC) based multitime scale co-optimized dispatch for integrated electricity and natural gas system (IEGS) considering the bidirectional interactions and renewable uncertainties. In the proposed model, optimal dispatch is extended into three substages optimization problems of day-ahead, intraday, and real-time to coordinate the economy and accuracy of operations. The different optimization strategy is designed in each dispatching stage according to the operating characteristics of multienergy coupled units and high-penetration intermittent wind power. In the day-ahead stage, the unit commitment is tackled to arrange the start-stop plan of slow-response units, maintained in intraday and real-time stages to ensure full-term optimal economic operation. In the intraday stage, MPC-based rolling optimization is implemented to obtain the unit output scheme with the lowest operating cost target. Furthermore, the output scheme is rolling adjusted in the real-time stage based on the ultrashort-term forecast value, and a closed-loop feedback mechanism is introduced to achieve accurate power balance. In each stage, stochastic decision-making is executed based on the stochastic scenario method, efficiently capturing the uncertainty in IEGS. Besides, the proposed nonlinear model is transformed into a mixed-integer second-order cone programming problem using McCormick relaxation, which can be efficiently solved by commercial solvers. Simulation results on IEEE39-Gas20 and IEEE118-GAS40 test systems demonstrate the superiority of the proposed method in operational economy and wind power utilization, and also verify the effectiveness of the method to accurately track random fluctuations without obvious computational burden.