Frequency dynamics during primary frequency control is closely related to pre-disturbance operating conditions, the disturbance, and equipment characteristics. How to rigorously express such a relation in a frequency stability constrained optimal dispatch problem is still an open question. This paper presents an optimal power flow (OPF) model for generation redispatch that considers dynamic frequency response constraints to ensure frequency stability during primary frequency regulation. The distinct feature of the model is that the dynamic frequency response is formulated as a set of differential algebraic equations (DAEs), which allows considering the transient behavior of frequency and mechanical/electromagnetic power for each generator during primary frequency control period. Based on the solution of the optimization problem, a definition of the primary reserve for each unit is proposed according to the maximum of incremental mechanical power. This definition helps meet the adequacy requirement of primary frequency response and achieve partial frequency restoration. Simulation results on WSCC 3-machine 9-bus system, New England 10-machine 39-bus system and the modified IEEE 54-machine 118-bus system validate the effectiveness of the proposed model and reveal the strong coupling between frequency dynamics and the predisturbance generation.

Line loss rate plays an essential role in evaluating the economic operation of power systems. However, in a low voltage (LV) distribution network, calculating line loss rate has become more cumbersome due to poor configuration of the measuring and detecting device, the difficulty in collecting operational data, and the excessive number of components and nodes. Most previous studies mainly focused on the approaches to calculate or predict line loss rate, but rarely involve the evaluation of the prediction results. In this paper, we propose an approach based on a gradient boosting decision tree (GBDT), to predict line loss rate. GBDT inherits the advantages of both statistical models and AI approaches, and can identify the complex and nonlinear relationship while computing the relative importance among variables. An empirical study on a data set in a city demonstrates that our proposed approach performs well in predicting line loss rate, given a large number of unlabeled examples. Experiments and analysis also confirmed the effectiveness of our proposed approach in anomaly detection and practical project management.

Conventional dielectric response measurement techniques, for instance, recovery voltage measurement (RVM), frequency domain spectroscopy (FDS) and polarization–depolarization current (PDC) are effective nondestructive insulation monitoring techniques for oil-impregnated power transformers. Previous studies have focused mainly on some single type of dielectric measurement method. However, the condition of oil paper insulation in transformer is affected by many factors, so it is difficult to predict the insulation status by means of a single method. In this paper, the insulation condition assessment is performed by grey relational analysis (GRA) technique after carefully investigating different dielectric response measurement data. The insulation condition sensitive parameters of samples with unknown insulation status are extracted from different dielectric response measurement data and then these are used to contrast with the standard insulation state vector models established in controlled laboratory conditions by using GRA technique for predicting insulation condition. The performance of the proposed approach is tested using both the laboratory samples and a power transformer to demonstrate that it can provide reliable and effective insulation diagnosis.

This paper presents a framework of power dispatching automation system (PDAS) based on virtual private cloud, which takes advantages of technologies such as virtual private cloud, virtual private network, https protocol, etc. Unlike existing on-premise or public cloud hosted PDAS which are vulnerable and may subject to blackouts due to extreme conditions such as natural disasters or cyberattacks, the proposed framework can operate more safely and securely under those conditions, while at the same time also meeting the reliability requirements for power systems. This paper presents the theory, architectural design, characteristics and implementation process of the system, as well as its reliability and cyber security measurements. The annual failure time of the proposed system is reduced from 61.2 minutes to about 0.43 minutes, about 142 times less than existing PDAS. The framework is implemented in Lipu power system in China since May 2016. Results show that the proposed system is reliable, safe and cost-effective, with a bright prospect for future applications.

Polarization-depolarization current (PDC) measurements are now being used as a diagnosis tool to predict the ageing condition of transformer oil-paper insulation. Unfortunately, it is somewhat difficult to obtain the ageing condition of transformer cellulose insulation using the PDC technique due to the variation in transformer insulation geometry. In this literature, to quantify the ageing condition of transformer cellulose insulation using the PDC technique, we firstly designed a series of experiments under controlled laboratory conditions, and then obtained the branch parameters of an extended Debye model using the technique of curve fitting the PDC data. Finally, the ageing effect and water effect on the parameters of large time constant branches were systematically investigated. In the present paper, it is observed that there is a good exponential correlation between large time constants and degree of polymerization (DP). Therefore, the authors believe that the large time constants may be regard as a sensitive ageing indicator and the nice correlations might be utilized for the quantitative assessment of ageing condition in transformer cellulose insulation in the future due to the geometry independence of large time constants. In addition, it is found that the water in cellulose pressboards has a predominant effect on large time constants.

This paper presents a Dynamic Adam and dropout based deep neural network (DADDNN) for fault diagnosis of oil-immersed power transformers. To solve the problem of incomplete extraction of hidden information with data driven, the gradient first-order moment estimate and second-order moment estimate are used to calculate the different learning rates for all parameters with stable gradient scaling. Meanwhile, the learning rate is dynamically attenuated according to the optimal interval. To prevent over-fitted, we exploit dropout technique to randomly reset some neurons and strengthen the information exchange between indirectly-linked neurons. Our proposed approach was utilized on four datasets to learn the faults diagnosis of oil-immersed power transformers. Besides, four benchmark cases in other fields were also utilized to illustrate its scalability. The simulation results show that the average diagnosis accuracies on the four datasets of our proposed method were 37.9%, 25.5%, 14.6%, 18.9%, and 11.2%, higher than international electro technical commission (IEC), Duval Triangle, stacked autoencoders (SAE), deep belief networks (DBN), and grid search support vector machines (GSSVM), respectively.