J.C.S. Souza

Fault location in electrical power systems using intelligent systems techniques

In this work an artificial neural network based methodology is proposed for power systems fault location. Several artificial neural networks are employed, each of them being responsible for detecting faults involving a limited number of components. The proposed methodology is tested using a test system and a real Brazilian system. Indexes based on the Hamming distance are also proposed for feature selection and analysis.

Online topology determination and bad data suppression in power system operation using artificial neural networks

The correct assessment of network topology and system operating state in the presence of corrupted data is one of the most challenging problems during real-time power system monitoring, particularly when both topological (branch or bus misconfigurations) and analogical errors are considered. This paper proposes a new method that is capable of distinguishing between topological and analogical errors, and also of identifying which are the misconfigured elements or the bad measurements. The method explores the discrimination capability of the normalized innovations, which are used as input variables to an artificial neural network whose output is the identified anomaly. Data projection techniques are also employed to visualize and confirm the discrimination capability of the normalized innovations. The method is tested using the IEEE 118-bus test system and a configuration of a Brazilian utility.

Data debugging for real-time power system monitoring based on pattern analysis

This paper presents a new method for solving bad data acquisition problems in power system state estimation. The normalized innovations, available in the pre-filtering stage of a forecasting-aided state estimator, are used as input variables to a constructive artificial neural network (ANN). The ANN performs a pattern analysis in order to identify both topological and analogical errors. The performance of the method is evaluated and discussed for different types of error and operating conditions using the IEEE-24 bus system.

Enhanced Bad Data Processing by Phasor-Aided State Estimation

The utilization of all the available data sources is a permanent objective of state estimation (SE). In this sense, forecasting-aided state estimation (FASE) is an important alternative to conventional state estimation, especially regarding its expeditious data validation scheme, which comprises bad data (BD) smearing effect elimination, block identification, and adequate replacement. Also, with the availability of synchrophasor measurements, there has been a growing interest in building a phasor-aided state estimation (PHASE) process. This paper presents a novel way of processing BD, whose features are similar to those found in the data validation routines of FASE. The proposed PHASE approach has the advantage of leaving the existing SE application software intact, complementing it with an extra estimation module, capable of processing phasor measurements separately and judging whether the measurement set contains anomalies. The results of a proof of concept study performed on the IEEE 14-bus benchmark system demonstrate the application of the proposed methodology. Also, PMU-observability issues are addressed and illustrated through simulation studies conducted on the IEEE 118-bus system.

Fast contingency selection through a pattern analysis approach

Abstract This paper presents a method for automatic contingency selection and static security evaluation of electrical power systems. The method employs multi-layer perceptron neural networks whose inputs are power flows and injections, while the outputs compute performance indexes associated with post-contingency scenarios. Contingency ranking and selection are performed based on the artificial neural networks responses. Classifications of system operating state with respect to static security are also provided. The performance of the method is evaluated for different operating conditions using the IEEE 24-bus test system.

Quantifying Observability in State Estimation

Observability means the aptitude for estimating the system state in its entirety from data currently available. As such, the problem of quantifying this aptitude assumes practical interest. Dealing with observable grids, this paper concentrates on proposing indicators capable of establishing unobservability risks. These indicators are based on measurement criticality analyses (with the aid of Venn diagrams) for a given network configuration. They are obtained in terms of the probability of unobservability, assuming that an event has occurred, such as the unavailability of: a single measurement; one pair of measurements; one k-tuple of measurements; a single metering unit; a single network branch; one pair of network branches. The potential/practical application of the proposed indicators is illustrated by the introduction of metric patterns, capable of reducing the information on unobservability risks to a single quantity. Numerical results obtained with the IEEE 14- and 118-bus test systems exemplify the computation of the proposed indicators.

Data Compression in Smart Distribution Systems via Singular Value Decomposition

Electrical distribution systems have been experiencing many changes in recent times. Advances in metering system infrastructure and the deployment of a large number of smart meters in the grid will produce a big volume of data that will be required for many different applications. Despite the significant investments taking place in the communications infrastructure, this remains a bottleneck for the implementation of some applications. This paper presents a methodology for lossy data compression in smart distribution systems using the singular value decomposition technique. The proposed method is capable of significantly reducing the volume of data to be transmitted through the communications network and accurately reconstructing the original data. These features are illustrated by results from tests carried out using real data collected from metering devices at many different substations.

Revealing gross errors in critical measurements and sets via forecasting-aided state estimators

Abstract State estimators are important monitoring tools which process real-time data in power system control centers. The capability of detecting and identifying bad data depends on the redundancy level of the information to be processed. Network changes or a temporary malfunction of the data acquisition system reduce data redundancy for state estimation. Measurement redundancy deterioration can be characterized by the presence of critical measurements and sets. For the vast majority of data validation algorithms, it is impossible to process gross errors in critical measurements and sets. This paper proposes an algorithm for detecting, identifying and removing bad data in critical measurements and sets through forecasting-aided state estimators. Using the IEEE-14 bus test system, the performance of the proposed algorithm is evaluated and discussed through the simulation of different levels of data redundancy degradation.

A genetic-based methodology for evaluating requested outages of power network elements

The evaluation process of requested outages in power grids comprises a harmonious adjustment of maintenance needs placed by market participants, in which priorities and operational constraints are taken into account. It is a critical process for independent system operators (ISOs) since they were conceived for handling a huge amount of transactions, without affecting system reliability and market prices. This paper presents a methodology for the evaluation of requested outages, viewed as a constrained optimization problem due to its combinatorial nature. A genetic algorithm technique is adopted to obtain high-quality solutions. Simulation results with the IEEE 14-bus test system and part of a Brazilian system are presented to illustrate the proposed methodology.

Automatic contingency selection based on a pattern analysis approach

Summary form only given. This paper presents a method for power systems automatic contingency selection and static security evaluation. It is possible to identify potentially harmful contingencies in a very short computational time, being the risk of false alarms and contingency misses. The method is tested for many different operating conditions and simulated with the IEEE 24 bus test system. Classification rates for contingency selection and static security evaluation are also provided.