The ever-increasing number of wind farms has brought challenges and opportunities in wind power forecasting techniques to take advantage of interdependencies between hundreds of spatially distributed wind farms, e.g., over a region. In this paper, a Sparsity-Controlled Vector Autoregressive (SC-VAR) model is introduced to obtain sparse model structures in a spatio-temporal wind power forecasting framework by reformulating the original VAR model into a constrained Mixed Integer Non-Linear Programming (MINLP) problem. It allows controlling the sparsity of the coefficient matrices in direct manner. However this original SC-VAR is difficult to implement due to its complicated constraints and the lack of guidelines for setting its parameters. To reduce the complexity of this MINLP and to make it possible to incorporate prior expert knowledge to benefit model building and forecasting, the original SC-VAR is modified and a Correlation-Constrained SC-VAR (CCSC-VAR) is proposed based on spatial correlation information about wind farms. Our approach is evaluated based on a case study of very-short-term forecasting for 25 wind farms in Denmark. Comparison is performed with a set of traditional local methods and spatio-temporal methods. Results show the proposed CCSC-VAR has better overall performance than both the original SC-VAR and other benchmark methods, taking into account all evaluation indicators, including sparsity-control ability, sparsity, accuracy and efficiency.

This paper presents a fast and reliable method to evaluate risk of single or multiple commutation failures for multi-infeed HVDC systems following faults occurring at the receiving-end ac system. First, a fast calculation method of multi-infeed interaction factor, describing the interactivity and mutual impact between two inverter AC buses on is proposed. It is then refined to being the AC-DC system voltage interaction factor (ADVIF) to represent the relationship in bus voltages between inverter and AC buses at the receiving-end systems. Then, a critical AC-DC system voltage interaction factor (CADVIF) based on the minimum extinction angle criteria is proposed. By calculating and comparing ADVIFs and CADVIFs, which can be obtained quickly, it is possible to identify AC buses that, if a fault occurs, would cause commutation failures at one or more HVDC systems. The validity and accuracy of the proposed approach are demonstrated by comparing with simulations results using a two-infeed HVDC test system and an actual large power grid.

Large-scale wind power cluster with distributed wind farms has generated the active power dispatch and control problems in the power system. In this paper, a novel hierarchical model predictive control (HMPC) strategy based on dynamic active power dispatch is proposed to improve wind power schedule and increase wind power accommodation. The strategy consists of four layers with refined time scales, including intra-day dispatch, real-time dispatch, cluster optimization and wind farm modulation layer. A dynamic grouping strategy is specifically developed to allocate the schedule for wind farms in cluster optimization layer. In order to maximize wind power output, downward spinning reserve and transmission pathway utilization are developed in wind farm modulation layer. Meanwhile, a stratification analysis approach for ultra-short-term wind power forecasting error is presented as feedback correction to increase forecasting accuracy. The proposed strategy is evaluated by a case study in the IEEE network with wind power cluster integration. Results show that wind power accommodation has been enhanced by use of the proposed HMPC strategy, compared with the conventional dispatch and allocation methods.

Accurate and generalized collaborative prediction of multi-cluster renewable energy power generation is both an inevitable trend and urgent demand as the growth of multi-region interconnected power grids with wind and photovoltaic (PV) power. In this paper, a novel heterogeneous spatiotemporal graph convolution network (HSTGCN) is proposed for ultra-short-term multi-modal prediction oriented to wind-PV power, which sufficiently considers spatiotemporal correlations in tens of wind farms or PV stations of each neighboring region and effectively coordinates the heterogeneities of different power generation types in different regions. This approach first designs a dynamical heterogeneous graph structure including modes, nodes, and edges to give a unified framework evolving over time for different interdependencies in the multi-cluster wind and PV sites, and then develops a hierarchical spatiotemporal learning mechanism to enhance representation power for multi-cluster prior information from temporal and spatial dimensions, integrating 2-D CNN with different sizes of filters and GCN embedded a specially designed lightweight graph convolution attention module (GCAM). Experiments including 57 operating wind farms and PV stations from 4 regions distributed over a broad spatial scale demonstrate the generalization and interpretation of HSTGCN compared with other commonly considered benchmarks.

Deep learning (DL) is a useful tool for power system stability assessment (PSSA) and dominant instability mode (DIM) identification. However, when faced with operational variability, the performance of DL models degrades. This paper proposes a bidirectional active transfer learning (Bi-ATL) framework for more adaptive PSSA and DIM identification, where the DL model is easier to adapt to unlearned operating conditions with fewer newly labeled instances. At the instance level, forward active learning and backward active learning are integrated to progressively build a mixed instance set by actively including the most label-worthy instances of new operating conditions and actively eliminating the most useless original operating condition instances. Then at the model parameter level, the mixed instance set is utilized to fine-tune the original DL model to new operating conditions. The Bi-ATL framework synthesizes three-way information of the instances and model of the original operating condition, and a few labeled instances of new operating conditions for more efficient adaptation. Intensive case studies conducted on a benchmark power system (CEPRI 36-bus system) and a real-world large-scale power system (Northeast China Power System-2131 bus) validate the efficacy and efficiency of the Bi-ATL framework as well as the role of the three-way information.

With the expansion of power system simulation scale, intelligent stability analysis based on simulation data becomes more and more important. However, the changeable operating conditions and fault conditions will seriously affect the performance of the intelligent analysis model. This paper proposes an advanced unsupervised transfer learning (UTL) method combining dynamic graph attention network (DGAT) and data augmentation to identify the instability mode towards multiple known and unknown scenarios. For the pre-training stage, the DGAT model with data augmentation is used for instability mode identification based on the basic dataset. For the transfer learning stage, it starts when faced with new operating conditions or fault conditions different from the basic dataset. The UTL scheme takes the distribution differences among different categories of source domain and target domain as the loss to fine-tune the pre-trained model. Moreover, considering the difficulty of sample-labeling, the UTL scheme can make the pre-trained model adapt to them without additional labeled samples. Case studies are conducted on 8-machine 36-bus system and Northeast China Power Grid. The results verify the superiority of the DGAT model with data augmentation and confirm the UTL scheme can help the model adapt to multiple unknown scenarios.