A large-scale wind power grid connection reduces the level of inertia in a power system and increases the out-of-limit risk of transient frequency deviations. However, the existing frequency prediction models are not suitable for a fast and highly accurate online prediction of wind power system characteristics. Therefore, further optimizing the extreme value prediction method that predicts frequency deviations is necessary for evaluating the frequency stability of such systems. This paper introduces an online prediction method for predicting the frequency deviation extremum based on physical data fusion; this method is based on a wide-area measurement system and considers the influence of the wind power grid connection on the frequency response process. First, a transient frequency analysis model of physical data fusion was developed based on data bonding of the open-loop decoupling model of active power and frequency using wide-area measurement information. Second, the predicted extreme value of frequency deviation is updated in real-time using the model. A “prediction value error index” is proposed to quantify the prediction accuracy and guide the adaptive dynamic output of the online model. Finally, an example is provided to verify the prediction speed and accuracy of the extremum frequency deviation predicted by the proposed online prediction method.
Rotor angle instability impedes the safe and stable operation of AC and high-voltage direct current (HVDC) hybrid power grids. Rapid identification of rotor angle instability is the premise of emergency control, such as unit switching and HVDC power increase at the sending end. To distinguish between the rotor angle and voltage-dominant stable state more effectively, the sBTTC index is modified to weaken the phase factor of the original index. The dominant stable state is determined based on the correlation coefficient between the sBTTC index and the voltage factor, correlation coefficient of the phase factor, and ratio of the two coefficients. Setting the pedal voltage of the branch to be greater than that of the branch is one of the necessary conditions for judging rotor angle stability. This paper discusses four typical forms of rotor angle instability, namely rotor angle one swing instability, rotor angle two swing instability, rotor angle reverse swing instability, and rotor angle low-frequency amplitude oscillation instability. The simulation results show that the discrimination method can accurately distinguish between the different forms.
Predicting the level of in advance in new power systems is essential to eliminate the risk of a weak system inertia, and black-box machine learning models, which have insufficient interpretability, are widely used for system-inertia predictions. Therefore, this paper introduces a short-term prediction method, based on interpretable extreme gradient boosting (XGBoost), for power system inertia. Based on the analysis of the system inertia response characteristics, the method selects the power system operation and meteorological data as input features. The interpretation mechanism of XGBoost was constructed based on Shapley additive explanation values. By calculating the Shapley value to quantify the importance of each feature, the model prediction results can be deconstructed into multiple dimensions. Simulations were performed using a realistic photovoltaic system, and the results showed that the proposed method can effectively predict the short-term inertia of a power system as well as elucidate the influence of the features on the predicted results.
The impedance-based analysis method is an effective tool for studying the wideband oscillation problems of “double-high” power systems. However, traditional impedance modeling methods have deficiencies such as the necessity to obtain a detailed structure and control parameters of the system in advance, complex modeling process, and inconvenience for online applications. To address these challenges, an improved BP neural network based impedance model identification method for a doubly-fed induction generator (DFIG) based wind turbine is proposed herein. An evaluation coefficient was defined to evaluate the accuracy of the neural network identification model, and the structure of the BP neural network was improved based on this. Under the improved network structure, accurate multi-condition impedance model identification can be achieved with less data. By comparing the identification impedance errors of both the proposed and the traditional BP neural network methods, the superiority of the proposed method was demonstrated. To verify the feasibility of the proposed method for online stability analysis, the stability of the DFIG-based wind turbine grid integration system under different grid short-circuit ratios and wind turbine output active power conditions was analyzed using the identified impedance. The accuracy of the proposed method was subsequently verified using time-domain simulations.
With the background of dual-carbon targets, countless new energy power prediction models with regional characteristics and performance characteristics have emerged. Distinguishing and utilizing the numerous existing prediction models are challenges for prediction personnel in practical applications. Therefore, this paper proposes a new energy power prediction model inference method based on heterogeneous graph learning, which divides the new energy power prediction into a basic model layer and model inference layer. In the basic model layer, prediction models with different characteristics are trained using datasets with different regional characteristics. In the model inference layer, a heterogeneous graph representation method of “input information node-input edge-model node-potential connection edge-prediction result” is used to fuse the heterogeneous information, and the optimal model inference is achieved through a heterogeneous graph attention network. Thus, accurate new energy prediction results are obtained. For a case study involving a wind farm in southwest China, the wind power prediction error of the proposed method was <9%, and the prediction performance for direct migration to new sites without data was superior to that of other methods.
The active distribution network with a large proportion of distributed renewable power sources has frequent changes in operation mode, necessitating a fast and accurate operation mode identification method. Therefore, this paper proposes a method for identifying the operation mode of the distribution network online using real-time measurement data from smart meters. First, the node voltage and power data obtained from smart meters are preprocessed to calculate the voltage difference between two nodes, and the voltage-difference data and power data of each historical time segment are downscaled using t-distributed stochastic neighbor embedding. Then, a few historical sampling moments are selected, and the correspondence between the data of these moments and the distribution network operation mode is established manually. Finally, the real-time measurement data are downscaled to determine how the active distribution network is operating at the current moment according to known correspondences. Case studies verify the feasibility and effectiveness of the proposed method, which can maintain an identification accuracy of >90% when the accuracy of the measurement device is not below the 0.2 level. The proposed method can quickly and accurately identify the operation mode of an active distribution network online by using voltage and power data collected by smart meters.
Unmanned aerial vehicle (UAV) inspection plays an increasingly important role in the daily inspection of 220kV transmission lines. However, UAV inspection has a series of problems, such as inaccurate positioning and the difficulty of identifying and tracking inspection targets. Therefore, an adaptive method for UAV transmission-line patrol positioning and target tracking based on the fusion of information from inertial measurement units, global navigation satellite systems (GNSS), visual odometry, and other multichannel sensors is proposed. To address the difficulty of autonomous localization caused by GNSS signal interruption and the lack of feature points for visual sensors in complex scenes, a robust adaptive localization algorithm based on federated Kalman filtering is proposed, which ensures the accuracy of autonomous navigation and localization for UAVs. Additionally, to achieve automatic target recognition and tracking of transmission lines, an optimal threshold selection algorithm for changing scenarios is proposed for predictive and continuous parameter selection, to increase the accuracy of transmission-line tracking. A typical case in Tangshan was used as a pilot to verify the effectiveness of the proposed method and optimize the workflow of UAV inspection for transmission lines, laying a foundation for multidimensional intelligent transmission-line inspection.
Aiming at the regional operation of power systems and the uneven regional distribution of multitype flexible resources, an optimal dispatching strategy for flexible resources considering flexible mutual aid among regional grids is proposed. First, a comprehensive flexible supply capacity model considering multitype flexible resources was established, and a power constraint model of the tie line was realized considering the level of flexible mutual aid. Second, an optimal dispatching model of a multiregional power grid based on a flexible mutual aid mechanism is accomplished, in which the flexibility balance mechanism is combined with the coordinated operation of multiple flexible resources, and fuzzy opportunity constraints are introduced to describe the impacts of net load uncertainties on the flexible supply-demand balance and power balance. Finally, three interconnected IEEE-39-node systems are used as examples to verify the proposed dispatching model. The results show that the proposed flexible mutual aid dispatching model can promote a flexible supply-demand balance of the power grid, effectively reduce the system operation cost and carbon emissions, and thus has positive significance for the economic, safe, and low-carbon operation of the power system.
In gas-fired power plant carbon capture (GPPCC), carbon dioxide (CO2) is captured and converted to methane (CH4) via power-to-gas (P2G). This paper proposes connecting GPPCC, carbon storage device (CS), and P2G with a virtual power plant (i.e., C2P-VPP). First, a C2P-VPP structure and a mathematical model were constructed. Subsequently, the dispatch cost, carbon emissions, and output fluctuation are minimized as the optimal objectives, and the robust stochastic optimal theory is employed to characterize the uncertainty of wind power plants (WPPs) and photovoltaic power plants, and establish a C2P-VPP multi-objective stochastic optimal dispatching model. Finally, the CIGRE medium-voltage distribution system was selected for the analysis. The results show that the proposed optimal decision model can exert the electrocarbon-electric cycle optimal effect of C2P and provide an effective tool to aid decisionmakers in balancing the C2P-VPP dispatching scheme.
Electric vehicles (EVs) are promising demand-side regulation resources. This study examines the integration and scheduling of EV load resources considering electric vehicle aggregators (EVAs) in the spot market. First, the EV vehicle-pile charging load differentiation was considered to establish the EV charging model. Second, to maximize the EVA operating income, taking into account the interests of EV users simultaneously, an EVA day-ahead optimization model based on the Stackelberg game is established, and a scenario-based stochastic approach is used to consider the uncertainty of the electricity market price, gaining differentiated pricing results of the EV charging service fee and EVA day-ahead energy purchase plan. Subsequently, based on the energy-sharing mode, all types of EVs are dispatched via rolling optimization scheduling in real time to reduce the unplanned EVA power purchase cost caused by real-time market transactions owing to EV access uncertainty. Finally, considering an EVA in the distribution area as an example, the calculation results prove the effectiveness of the two-stage scheduling strategy proposed in this study, which can cope with the uncertainty of EV access and market price as well as realize a win-win situation between EVA and EV users.
To determine the energy supply potential of distributed energy resources, this study focuses on the dispatching optimization of combined heat and power virtual power plants (VPP). First, combined heat and power (CHP) units and various distributed energy sources were integrated into a combined heat and power virtual power plant. Carbon capture and power-to-gas devices were also introduced to realize the recycling utilization of CO2. Further, carbon and hydrogen storage devices are added to decouple the carbon capture and electro-gas conversion processes. Then, through un-certainty scenario generation and conditional value-at-risk (CVaR) theory, the risks in virtual power plant real-time dispatching are qualified. Finally, with operating cost, carbon emissions, and operating risk as the objectives, a multi-objective stochastic optimization model of a virtual power plant is constructed and solved using a subjective and objective integrated weighting method. The example results demonstrate that the proposed method can promote the consumption of wind and photovoltaic power, besides reduce the carbon emissions of power plants.
Cloud-edge collaboration is a key technological approach for power automation to adapt to the massive access of distributed objects and support the distribution internet of things (IoT) automation business. To fully utilize the resource advantages of cloud-edge collaborative systems and reduce the delay in the distribution of the IoT automation business, this study starts from the perspective of cloud-edge resource collaboration and proposes a cloud-edge architecture for the distribution of IoT, which includes resource elastic configuration and communication security encryption function modules. Then, using the microservice-based power business organization method, spatiotemporal logic and computational load models of the distribution IoT automation business were constructed. On this basis, considering the elastic allocation of container resources and encryption delay of cloud-edge communication, an elastic configuration method for cloud-edge collaborative resources was proposed to reduce business delays through the overall collaborative multi-container resources of the edge and cloud. Finally, the influence of different business spatiotemporal logics was analyzed, and the results of the resource configuration and business delay under different scenario settings were compared to verify the effectiveness of the proposed method.
Energy sharing is an effective method for enhancing the economics of multi-microgrid systems and promoting the local consumption of distributed photovoltaic systems. The efficiency and fairness of multi-microgrid energy sharing rely on a suitable market framework. Therefore, herein, an energy-sharing model, based on the call auction mechanism, is proposed for multiple microgrids. First, an energy-sharing market framework based on the call auction mechanism is established, in which each microgrid submits bids for both power purchase and sale to an energy-sharing service provider. The energy-sharing service provider aims to maximize the social welfare of multiple microgrids by considering the distribution-network costs. The microgrid in this framework only needs to submit the bid price and quantity, which helps protect the privacy of the microgrid. Second, an optimal operation model of the microgrid is established to determine the marginal utility of the microgrid in terms of interactive electric power based on sensitivity analysis and moving average filtering. Because of the large number of microgrids, the energy-sharing market is in a state of perfect competition; thus, a microgrid bidding strategy based on marginal utility is proposed. Finally, an IEEE-33 node distribution system containing 20 microgrids is used to verify the effectiveness and feasibility of the proposed method.
With the background of the new power system, studying the multi-microgrid (MMG) low-carbon scheduling strategy that considers the uncertainty of wind power is conducive to reducing the impact of the uncertainty of wind power on the low-carbon and economic dispatching strategy. Therefore, considering the carbon trading mechanism, an MMG adjustable robust low-carbon economic scheduling model is proposed, which employs meteorological clustering and grouping generation to improve the uncertainty set of wind power. First, the wind power output generation model based on meteorological feature clustering is established by using minimum redundancy and maximum correlation feature selection technology, the CURE algorithm, and a conditional generative adversarial network. Using the fuzzy set of the probability distribution, the uncertainty set of the wind power output under different meteorological types is constructed. Second, to realize the low-carbon economic operation of the microgrid, the stepped carbon trading mechanism is introduced into the microgrid. Meanwhile, considering the impact of wind power uncertainty, a two-stage adjustable robust optimization model based on improved wind power uncertainty is established. Third, the ADMM nested C&CG algorithm is employed to solve the MMG dispatch model. Finally, examples are analyzed to verify that the proposed model can increase the accuracy of wind power uncertainty description and improve the economic and low-carbon performance of microgrid operation.
Limited by the resource endowment of the power system, the bottleneck of renewable energy absorption capacity has always restricted the development of renewable energy. In recent years, integrated energy systems have been widely used owing to their multi-energy coupling property. They can transfer the surplus power of the power system to other energy subsystems such as gas/heat/cold for consumption, providing a new idea for the consumption of renewable energy. However, owing to the complex multi-energy coupling, large scale, high dimension, and strong non-convex nonlinearity of the integrated energy system, accurately evaluating its renewable energy absorption capacity is difficult. In this study, considering the physical characteristics of energy subsystems such as electricity/gas/heat/cold, a renewable energy absorption model of the integrated energy system was developed to evaluate the limit of the system’s renewable energy absorption capacity under the multi-energy coupling effect. Then, a convex relaxation and approximation method was developed for transforming the original mixed-integer nonlinear programming problem into a mixed-integer conic programming problem, for solving the model. Finally, on the basis of the improved IEEE-39 node power system and the Belgium 20 node natural gas system, a numerical example of an electricity/gas/hydrogen/hydrothermal/wind/optical coupling integrated energy system was constructed. The model is used to demonstrate the enhancement of renewable energy integration through multi-energy coupling and to thoroughly investigate the variations in the capabilities of different nodes to accommodate renewable energy.
