Multi-Labeled Recognition of Distribution System Conditions by a Waveform Feature Learning Model

Abstract

A waveform contains recognizable feature patterns. To extract the features of various equipment disturbance conditions from a waveform, this paper presents a practical model to estimate distribution line (DL) conditions by means of a multi-label extreme learning machine. The motivation for the waveform learning is to develop device-embedded models which are capable of detecting and classifying abnormal operations on the DLs. In waveform analysis, power quality waveform modeling criteria are adopted for pattern classification. Typical disturbance waveforms are applied as class models, and the formula-generated waveform features are compared with field-acquired waveforms for disturbance classification. In particular, filtered symmetrical components of the modified varying window scale are applied for feature extraction. The proposed model interacts suitably with the parameter update method in classifying the waveforms in real network situations. The classification result showed disturbance features on model with the real DL waveform data holds a potential for determining additional DL conditions and improving its classification performance through the update mechanism of the learning machine.

Event Sequence Model Application for Prioritization and Detection of Pre-Fault Waveforms on Power Distribution Lines

“…In terms of a distribution disturbance and the fault circuit status, waveform analysis is conventionally applied to classify the event occurrence. The classification for event waveforms has been applied to propose a practical disturbance classifier for empirical distribution monitoring devices [21]. The previous study correspondingly shows that features of the waveforms and of the classifier could potentially be implemented for waveform shape identification, to be transformed to event classes.…”

“…The arranged event value using the ESM structure is described in this section to construct a timesequential matrix of the classified events. Each obtained and classified waveform is defined as a class C having its own classification number based on the machine-learning classifier [21]. Therefore, C is annotated with time and DL indices to comprise a sequence matrix: Waveform data acquired from several distribution monitoring systems were used to build a substation scale event matrix that has time-series sequences proposed as the ESM level-3 matrix as follows:…”

“…For examining massive measurement waveform data, sequenceoriented signal feature extraction was conducted in a prior stage, as was learning the data feature patterns needed to obtain event classes. The method employs zero sequence current signals to discover transient waveforms by machine learning-based pattern classification models [21]. Distinct patterns of fault-related phenomena were discovered by the data-driven model.…”

A New Clustering Approach for Automatic Oscillographic Records Segmentation

“…As mentioned in the abstract, these oscillographic databases are immense and the scientific community has been working on this topic to find solutions that allow a better use them for knowledge generation purposes. Works [1][2][3] are prominent examples in the literature.…”

Hybrid Machine Learning Models for Classifying Power Quality Disturbances: A Comparative Study

“…The selection of suitable feature remains a key challenge that requires developing tools in areas such as statistical analysis, machine learning, or data mining [14]. Valuable efforts have been made in this sense and some techniques are used for a precise selection of features including the principal component analysis [15], K-means-based apriori algorithm [16], classification and regression tree algorithm [17], multi-label extreme learning machine [18], random forest model [19], sequential forward selection [20], and bionic algorithms. This latter group has also been successfully used in classification rule discovery.…”