Predictions of renewable energy (RE) generation and electricity load are critical to smart grid operation. However, the prediction task remains challenging due to the intermittent and chaotic character of RE sources, and the diverse user behavior and power consumers. This article presents a novel method for the prediction of RE generation and electricity load using improved stacked gated recurrent unit-recurrent neural network (GRU-RNN) for both univariate and multivariate scenarios. First, multiple sensitive monitoring parameters or historical electricity consumption data are selected according to the correlation analysis to form the input data. Second, a stacked GRU-RNN using a simplified GRU is constructed with improved training algorithm based on AdaGrad and adjustable momentum. The modified GRU-RNN structure and improved training method enhance training efficiency and robustness. Third, the stacked GRU-RNN is used to establish an accurate mapping between the selected variables and RE generation or electricity load due to its self-feedback connections and improved training mechanism. The proposed method is verified by using two experiments: prediction of wind power generation using multiple weather parameters and prediction of electricity load with historical energy consumption data. The experimental results demonstrate that the proposed method outperforms state-of-the-art methods of machine learning or deep learning in achieving an accurate energy prediction for effective smart grid operation.

Presents a discussion of some important and unresolved issues related to previous work to provide a more comprehensive understanding of the practicability of wavelet-based denoising.

The objective of the paper is to discuss a tool which is proving extremely efficient in partial discharge measurement studies. Though the technique itself is not new, its application to partial discharge studies is. It will be demonstrated in this paper that it has tremendous power and this accounts for its rapid growth as an application in this field. The paper begins with the description of the fundamentals of wavelet analysis, wavelet categories and the properties of the associated wavelet transforms. PD pulses as acquired from detectors composed of different detection circuits are investigated and numerically simulated, and a method on how to select optimally the wavelet corresponding to the representative forms of PD pulse is then presented. Finally, applications of wavelet analysis to partial discharge studies are explored. The paper demonstrates that the wavelet based denoising method proposed in the paper can be employed in separating PD pulses from electrical noise successfully and can be used in pulse propagation studies of partial discharge in distributed impedance plant to provide enhanced information and further infer the original site of the PD pulse.

An effective condition monitoring system of wind turbines generally requires installation of a high number of sensors and use of a high sampling frequency in particular for monitoring of the electrical components within a turbine, resulting in a large amount of data. This can become a burden for condition monitoring and fault detection systems. This paper aims to develop algorithms that will allow a reduced dataset to be used in wind turbine fault detection. The paper firstly proposes a variable selection algorithm based on principal component analysis (PCA) with multiple selection criteria in order to select a set of variables to target fault signals while still preserving the variation of data in the original dataset. With the selected variables, the paper then describes fault detection and identification algorithms, which can identify faults, determine the corresponding time and location where the fault occurs, and estimate its severity. The proposed algorithms are evaluated with simulation data from PSCAD/EMTDC, SCADA (Supervisory control and data acquisition) data from an operational wind farm, and experimental data from a wind turbine test rig. Results show that the proposed methods can select a reduced set of variables with minimal information lost whilst detecting faults efficiently and effectively.

The role of electricity theft detection (ETD) is critical to maintain cost-efficiency in smart grids. However, existing methods for theft detection can struggle to handle large electricity consumption datasets because of missing values, data variance and nonlinear data relationship problems, and there is a lack of integrated infrastructure for coordinating electricity load data analysis procedures. To help address these problems, a simple yet effective ETD model is developed. Three modules are combined into the proposed model. The first module deploys a combination of data imputation, outlier handling, normalization and class balancing algorithms, to enhance the time series characteristics and generate better quality data for improved training and learning by the classifiers. Three different machine learning (ML) methods, which are uncorrelated and skillful on the problem in different ways, are employed as the base learning model. Finally, a recently developed deep learning approach, namely a temporal convolutional network (TCN), is used to ensemble the outputs of the ML algorithms for improved classification accuracy. Experimental results confirm that the proposed framework yields a highly-accurate, robust classification performance, in comparison to other well-established machine and deep learning models and thus can be a practical tool for electricity theft detection in industrial applications.