This paper presents a methodology aimed at extracting features to obtain information that will highlight disturbances related to the field of power quality. Due to the concept of smart grids, it is clear that the classification of the disturbances should be undertaken using smart meters, so that a large amount of data corresponding to the voltage and current waveforms are not exchanged using the communication channels, i.e., between smart meter and Utility's database server. Thus, it is necessary to ensure a balance between computational effort (arising from the implementation of algorithms on hardware) and the satisfactory performance of the algorithm for the classification of disturbances. Based on the assumption that the classification task is onerous, this paper proposes a step of feature extraction that may be calculated and analyzed offline using synthetic waveforms/signals, which are subsequently validated using field measurements. It should be noted that this offline analysis is required to determine the most relevant features for the process of classifying each disturbance. However, in order to establish the effectiveness of the feature extraction step, the response of decision trees of the C4.5 type and of artificial neural networks of the multilayer perceptron type were verified during the phase of disturbance classification. In short, good results were obtained that corroborate the hypothesis that the feature extraction step is necessary to classify disturbances effectively and with low computational effort.

Demand Response (DR) aims to motivate end consumers to change their energy consumption patterns in response to changes in electricity prices or when the reliability of the electrical power system (EPS) is compromised. Most of the proposals found in the literature only aim at reducing the cost for end consumers. However, this article proposes a home energy management system (HEMS) that aims to schedule the use of each home appliance based on the price of electricity in real-time (RTP) and on the consumer satisfaction/comfort level in order to guarantee the stability and the safety of the EPS. Thus, this paper presents a multi-objective DR optimization model which was formulated as a multi-objective nonlinear programming problem subjected to a set of constraints and was solved using the Non-Dominated Sorted Genetic Algorithm (NSGA-II), in order to determine the scheduling of home appliances for the time horizon. The multi-objective DR optimization model not only to minimize the cost of electricity consumption but also to reduce the level of inconvenience for residential consumers. Moreover, a priori, it is expected to obtain a more uniform demand with fewer peaks in the system and, potentially, achieving a more reliable and safer EPS operation. Thus, the energy management controller (EMC) within the HEMS determines an optimized schedule for each home appliance through the multi-objective DR model presented in this article, and ensures a more economic scenario for end consumers. In this paper, a performance evaluation of HEMS in 15 Brazilian families between 1 January and 31 December 2016 is presented with different electric energy consumption patterns in the cities of Belém—PA, Teresina—PI, Cuiabá—MT, Florianópolis—SC and São Paulo—SP, with three families per city, located in the regions north, northeast, central west, south and the southeast of Brazil, respectively. In addition, a total of 425 home appliances were used in the simulations. The results show that the HEMS achieved reductions in the cost of electricity for all the Scenarios used while minimally affecting the satisfaction/comfort of the end consumers as well as taking into account all the restrictions. The largest reduction in the total cost of electricity occurred for the couple without children, resident in the city of Teresina—PI; with a drop from US$ 99.31 to US$ 90.72 totaling 8.65% savings in the electricity bill. Therefore, the results confirm that the proposed HEMS effectively improves the operating efficiency of home appliances and reduces electricity costs for end consumers.

The Electrical Power Quality (PQ) studies are commonly related to disturbances that alter the sinusoidal voltage features and/or current wave shapes. The classification approaches of electrical power quality disturbances found in the literature mainly consist of three steps: 1) signal analysis and feature extraction, 2) feature selection and 3) disturbances classification. However, there are some problems inherent in disturbances classification. The manual extraction of features is an imprecise and complex process, which can influence the resuits and, therefore, does not deal well with noisy signals. This paper proposes an approach based on Deep Learning using the raw data, without pre-processing, manual extraction or manual feature selection of the PQ disturbances signals for the classification of fifteen electrical power quality disturbances. A deep network is used, which consists of a hybrid architecture, composed by convolutional layers, a pooling layer, an LSTM layer, and batch normalization to extract features automatically. We adopted a 1-D convolution to adapt the input. The extracted features are used as input to fully connected layers, the last one being a SoftMax layer. The results are compared with state of the art methods based on the three steps, showing that the proposed approach had satisfactory performance even with noisy data.

The correct location of the source of voltage sags is not a trivial task due to the short duration of these events and their rapid propagation in the distribution feeder. This paper proposes a method based on an ensemble method of the type Adaboost, with neural networks as base classifiers to determine the area where the voltage sag source is located. A voltage sag at a bus affects all other feeders, i.e. this disturbance is propagated in the whole system. The data management from smart meters installed in distribution feeders and decision support tools can become a viable alternative. In this sense, the smart meters could extract feature of the voltage sag and send it to the utility. At the utility, the AdaBoost performs location of the region by measuring the input’s features similarity to samples from the training set. For this purpose, it was necessary to analyze the relevance of each feature extracted from smart meters’ voltage signals to establish the structure which best represents the propagation of the disturbance in the system. The AdaBoost with Neural Network was tested in different scenarios of the 13-bus IEEE test feeder and was able to estimate the region with a good accuracy.