Le Xu

A classification approach for power distribution systems fault cause identification

Power distribution systems play an important role in modern society. When distribution system outages occur, fast and proper restorations are crucial to improve the quality of services and customer satisfaction. Proper usages of outage root cause identification tools are often essential for effective outage restorations. This paper reports on the investigation and results of two popular classification methods: logistic regression (LR) and artificial neural network (ANN) applied on power distribution fault cause identification. LR is seldom used in power distribution fault diagnosis, while ANN has been extensively used in power system reliability researches. This paper discusses the practical application problems, including data insufficiency, imbalanced data constitution, and threshold setting that are often faced in power distribution fault cause identification problems. Two major distribution fault types, tree and animal contact, are used to illustrate the characteristics and effectiveness of the investigated techniques.

Power Distribution Fault Cause Identification With Imbalanced Data Using the Data Mining-Based Fuzzy Classification

Power distribution systems have been significantly affected by many outage-causing events. Good fault cause identification can help expedite the restoration procedure and improve the system reliability. However, the data imbalance issue in many real-world data sets often degrades the fault cause identification performance. In this paper, the E-algorithm, which is extended from the fuzzy classification algorithm by Ishibuchi to alleviate the effect of imbalanced data constitution, is applied to Duke Energy outage data for distribution fault cause identification. Three major outage causes (tree, animal, and lightning) are used as prototypes. The performance of E-algorithm on real-world imbalanced data is compared with artificial neural network. The results show that the E-algorithm can greatly improve the performance when the data are imbalanced

E

Power distribution systems have been significantly affected by many fault causing events. Effective outage cause identification can help expedite the restoration procedure and improve the system reliability. However, the data imbalance issue in many real-world data often degrades the outage cause identification performance. In this paper, artificial immune recognition system (AIRS), an immune-inspired algorithm for supervised classification task is applied to the Duke Energy outage data for outage cause identification using three major causes (tree, animal, and lightning) as prototypes. The performance of AIRS on these real-world imbalanced data is compared with an artificial neural network (ANN). The results show that AIRS can greatly improve the performance by as much as 163% when the data are imbalanced and achieve comparable performance with ANN for relatively balanced data

-Algorithm

Power distribution systems play an important role in modern society. Proper outage root cause identification is often essential for effective restorations when outages occur. This paper reports on the investigation and results of two classification methods: logistic regression and neural network applied in power distribution fault cause classifier. Logistic regression is seldom used in power distribution fault diagnosis, while neural network, has been extensively used in power system reliability researches. Evaluation criteria of the goodness of the classifier includes: correct classification rate, true positive rate, true negative rate, and geometric mean. Two major distribution faults, tree and animal contact, are used to illustrate the characteristics and effectiveness of the investigated techniques.

Power Distribution Outage Cause Identification With Imbalanced Data Using Artificial Immune Recognition System (AIRS) Algorithm

The reliability and quality of power distribution systems are affected by different distribution faults. Trees are one of the major fault causes. In this paper, four different measures: actual measure, normalized measure, relative measure, and likelihood measure are used to data mine the Duke Energy Distribution Outage Database for meaningful data features and to analyze the characteristics of tree-caused distribution faults. This paper also applies statistical techniques to analyze tree-caused faults with respect to several selected influential factors. The results can be used to assist power distribution engineers to provide a more effective fault restoration system and design a more effective tree-fault prevention strategy

Comparisons of logistic regression and artificial neural network on power distribution systems fault cause identification

The elegant fuzzy classification algorithm proposed by Ishibuchi et al. (I-algorithm) has achieved satisfactory performance on many well-known test data sets that have usually been carefully preprocessed. However, the algorithm does not provide satisfactory performance for the problems with imbalanced data that are often encountered in real-world applications. This paper presents an extension of the I-algorithm to E-algorithm to alleviate the effect of data imbalance. Both the I-algorithm and the E-algorithm are applied to Duke Energy outage data for power distribution systems fault cause identification. Their performance on this real-world imbalanced data set is presented, compared, and analyzed to demonstrate the improvement of the extended algorithm.

Data Mining and Analysis of Tree-Caused Faults in Power Distribution Systems

Power distribution systems reliability is significantly affected by many outage causing events; good outage cause identification can help expedite the restoration procedure. However, the data imbalance issue encountered in many real-world data affects the performance of fault cause identification. The elegant fuzzy classification algorithm, I-algorithm, proposed by Ishibuchi et al. achieves satisfactory performance on many carefully preprocessed data sets but not on the imbalanced data, I-algorithm, an extension of the I-algorithm, is developed in this paper to alleviate the effect of imbalanced data constitution. Both the I- and E-algorithms are applied to Duke Energy outage data for power distribution systems fault cause identification. Their performance on this real-world imbalanced data set is presented, compared, and analyzed to demonstrate the improvement achieved by the extended algorithm

Data Mining Based Fuzzy Classification Algorithm for Imbalanced Data

Imbalanced data are often encountered in real-world applications, they may incline the performance of classification to be biased. The immune-based algorithm artificial immune recognition system (AIRS) is applied to Duke Energy distribution systems outage data and we investigate its capability to classify imbalanced data. The performance of AIRS is compared with an artificial neural network (ANN). Two major distribution fault causes, tree and lightning strike, are used as prototypes and a tailor-made measure for imbalanced data, g-mean, is used as the major performance measure. The results indicate that AIRS is able to achieve a more balanced performance on imbalanced data than ANN.

Using the Data Mining Based Fuzzy Classification Algorithm for Power Distribution Fault Cause Identification with Imbalanced Data

Power distribution systems have been significantly affected by many events. Effective outage cause identification can help expedite the restoration procedure and improve the system reliability. Fuzzy classification E-algorithm and biological immune system based AIRS algorithm have demonstrated good capability in outage cause identification especially with imbalanced data, E-algorithm can produce inference rules but is computational demanding; AIRS has the quick searching capability but is lack of rule extraction capability. In this paper, fuzzy artificial immune recognition system (FAIRS) has been proposed to utilize the advantages of both E-algorithm and AIRS. FAIRS is applied to Duke Energy outage data for cause identification using three major causes (tree, animal, and lightning) as prototypes. It is compared with both E-algorithm and AIRS, and the results show that FAIRS achieves comparable performance while being able to extract linguistic rules with rule length flexibility to explain the inference with significantly reduced computing time than E-algorithm.

On the Investigation of Artificial Immune Systems on Imbalanced Data Classification for Power Distribution System Fault Cause Identification

Many conventional fault diagnosis techniques do not effectively and efficiently use the available information and cannot achieve a satisfactory diagnosis in high dimensional real-world problems. In this paper, the fault diagnosis method using hierarchical clustering (HC) and small world (SW) networks stratification has been proposed to utilize the available information and trace up/downward based on event hierarchy and up/downstream along the physical network. As such, one can determine if certain diagnosis is applicable globally or more depends on the nature of events or locations; consequently the diagnostic uncertainty can be reduced. Duke energy distribution outage data are used to generate examples for the purpose of illustrating the motivation, necessity, implementation planning, and potential benefits of HC-SW stratification for power distribution system outage cause identification