Search

Power system security assessment is an important and challenging problem. Large variations in loads and power generation present increased risks to the secure operation of power systems. This study proposes a distributed deep network structure for power system security knowledge discovery based on multitask learning to monitor and control power grids more properly and effectively. First, a deep neural network structure based on the deep belief network (DBN) is designed to non‐linearly extract deep and abstract features layer‐by‐layer for total transfer capability (TTC) regression tasks. Then, a distributed training algorithm for the deep structure is developed to accelerate the training process. Furthermore, multitask learning is adopted by grouping and training‐related tasks together to improve the task performance. Finally, the accuracy and efficiency of the deep structure are evaluated using the Guangdong Power Grid in China. The simulation results demonstrate that the proposed model can outperform the existing shallow models in terms of accuracy and stability and can meet the requirements of online computing efficiency. [ABSTRACT FROM AUTHOR]

Short‐term voltage stability (SVS) is a serious issue in modern power systems. In China, the East China Power Network is especially vulnerable to short‐term voltage instability due to its increasing dependence on electrical power from external through high‐voltage direct current (HVDC) transmission lines. To study the SVS, a criterion/index is first required to evaluate the SVS of power systems. However, the currently used practical criteria cannot effectively evaluate the influence of controlling strategies (such as regulating dynamic VAR reservation) of power systems. Therefore, a practical and continuous SVS index (SVSI) based on voltage curves is proposed in this study. The proposed SVSI can quantify the extent of SVS in power systems, thus can be used in the optimisation target of preventive control for improving the SVS of power systems. Moreover, the SVSI consists of three components, and each component reflects a feature of the voltage curve in terms of SVS. The SVSI is essentially a model‐free method for assessing the SVS of power systems, so it is robust. Three verifying cases and one application case are presented to show the validity of the SVSI and the feasibility of the SVSI in practical applications. [ABSTRACT FROM AUTHOR]

To prevent the wind units from cascading trip faults, the concept of an autonomous voltage security region (AVSR) has been discussed in previous studies. However, the AVSR is sometimes not sufficiently large due to the insufficient dynamic reactive power reserve (DRPR) under normal conditions, resulting in secure operation problems and a large amount of wind curtailment. Therefore, reservation of an appropriate capacity of dynamic reactive power is a crucial issue in wind farms, so a DRPR optimisation model to enlarge the AVSR is proposed in this work. The computation of the optimal DRPR includes six steps and involves the exchange of information between wind farms and the control centre. The computation is rapid and can be applied to real wind systems online. The proposed method demonstrated good performance in a simple test case and a real system in Northern China. [ABSTRACT FROM AUTHOR]

A multivariate statistical analysis (MSA)‐based power‐grid‐partitioning method is proposed. Considering the effectiveness of automatic voltage regulation of generators, network modelling is based on the quasi‐steady‐state sensitivity of voltage with respect to reactive power. The MSA‐based power‐grid‐partitioning method contains two stages: a quantitative method for determining the appropriate number of zones and allocation of nodes into a number of zones. The proposed method is applied to an automatic voltage control system and an N − 1 robust pilot node selection method is presented. Simulation studies on IEEE 14‐bus, 39‐bus systems and a real provincial power system in China are described to illustrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

With the growing application of electric vehicles (EVs), it is of great significance to have a deep understanding of EV users driving and charging patterns for charging forecasting. However, the rapid growth scale of EV taxis with charging patterns that are closely coupled with human behaviours of temporal–spatial charging choices was not compatible with most previous coordinated strategies. Unlike the majority of existing approaches, a large volume of second‐level EV global positioning system (GPS) data was used to study the behaviour patterns of EV users. In practise, a mobile crowdsensing system that records GPS data and transmits information to the server was deployed in a fleet of electric‐taxi cabs in Shenzhen, China, making it possible to record the exact behaviour of each vehicle. Travelling and charging statuses of EVs were recorded and analysed into different characteristics of behaviour for each user. The load forecast methods proved to be more effective with more knowledge of both history and real‐time data. [ABSTRACT FROM AUTHOR]

This study presents the distributed model predictive control (D‐MPC) of a wind farm equipped with fast and short‐term energy storage system (ESS) for optimal active power control using the fast gradient method via dual decomposition. The primary objective of the D‐MPC control of the wind farm is power reference tracking from system operators. Besides, by optimal distribution of the power references to individual wind turbines (WTs) and the ESS unit, the WT mechanical loads are alleviated. With the fast gradient method, the convergence rate of the D‐MPC is significantly improved which leads to a reduction of the iteration number. Accordingly, the communication burden is reduced. Case studies demonstrate that the additional ESS unit can lead to a larger WT load reduction, compared with the conventional wind farm control without ESS. Moreover, the efficiency of the developed D‐MPC algorithm is independent from the wind farm size and is suitable for the real‐time control of the wind farm with ESS. [ABSTRACT FROM AUTHOR]

This study describes a solution to discriminate the original sources of bad data which are concerned by the substation maintainers. Bad data caused by defects of monitoring system, such as mal‐setting of monitoring devices or time skew, are concerned here. These bad data can be identified mostly in electric power control centre by bad data identification (BDI) methods including robust state estimation; however, these methods can hardly explain the relevant original sources of bad data in substations. Here, the authors propose a data‐driven method named abductive identification (AbI) to discriminate the original sources of bad data already identified by BDI methods. In the AbI method, feature patterns of bad data, based on their residuals, are investigated. A feature pattern selection method for bad data is then proposed, in which both the absolute residual and relative residual are considered. Based on this feature selection approach, a Bayesian classifier is developed to identify the original sources of bad data. A two‐level architecture for the AbI approach is introduced to deal with diverse bad data patterns of wide‐area networks. Simulation and field test results demonstrate the applicability of the proposed method. [ABSTRACT FROM AUTHOR]