A survey on statistical indexes applied on frequency response analysis of electric machinery and a trend based approach for more reliable results

Sant’Ana, Wilson Cesar; Salomon, Camila Paes; Lambert‐Torres, Germano; Silva, Luiz Eduardo Borges da; Bonaldi, Erik Leandro; Oliveira, Levy de Lacerda de Ely; Silva, Jonas Guedes Borges da

doi:10.1016/j.epsr.2016.03.044

Cited by 38 publications

(28 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Difficulties in the comparison of FRA results of two cases were found to get reduced when statistical parameters were employed for extracting the features from the FRA magnitudes on the sub-band basis [19]- [21], [28]. Hence, the comparison was further carried out using three statistical parameters: Absolute Difference (DABS), Min-Max ratio (MM) and Comparative Standard Deviation (CSD).…”

Section: Procedures Followed For the Analysismentioning

confidence: 99%

“…DABS gives the overall absolute difference between two sets of data compared and a result of '0' suggests no difference and the numerical value shows the severity of deviation [19]- [21]…”

Section: Procedures Followed For the Analysismentioning

confidence: 99%

“…Min Max Ratio gives the ratio of minimum value to the maximum value of the two data sets under comparison. A value of '0' or closer to 0 indicates maximum deviation and value of '1' or closer to 1 indicates the closeness of the values [19]- [21]…”

Section: Procedures Followed For the Analysismentioning

confidence: 99%

“…Comparative Standard Deviation (CSD) value needs to be 0 for a complete match between the two data sets compared. Initially, the variation of each data point with respect to its mean is calculated, which is further used to compute the Comparative Standard Deviation [19]- [21]…”

Section: Procedures Followed For the Analysismentioning

confidence: 99%

“…Though FRA approaches for electrical and mechanical fault diagnosis were demonstrated successfully in literature, interpretation of FRA results were found to be complicated, inviting expert service to avoid misinterpretations and to reach subjective conclusions. Such challenges were addressed and, the usefulness of feature extractions from FRA results in the form of statistical tools was successfully demonstrated in recent researches [19]- [21]. Jiangnan Liu et al combined the Support Vector Machine with FRA to diagnose the transformer faults and, discussed the usefulness of the parameter optimisation algorithms [22].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Influence of the Load on the Impulse Frequency Response Approach Based Diagnosis of Transformer’s Inter-Turn Short-Circuit

2020

View full text Add to dashboard Cite

Frequency Response Analysis (FRA) is an Off-line condition monitoring technique for the transformers and is successful, in diagnosing mechanical faults and inter-turn short-circuits. Only limited literature is available for On-line FRA works, and the effect of the transformer's load on the fault diagnosing capability of FRA has not yet been analysed. Motivated by this research gap, an experimental investigation was done to analyze the effect of load on IFRA based inter-turn short-circuit diagnosis. A single-phase 1 kVA, 240V/240 V transformer supplying R and R-L loads and a three-phase 5 kVA, 440 V/440 V transformer supplying an induction motor at various load levels were tested. Inter-turn short-circuits were emulated in their windings and, the effectiveness of IFRA in diagnosing the same was investigated. Comparisons were made between the frequency response of the loaded transformers in their healthier conditions and emulated fault conditions, based on the transfer function plots and the statistical parameters. Investigations reveal that the loads alter the useful spectrum of IFRA. The frequency response of the healthier transformer itself varies based on the transformer's load and necessitates a careful focus on the sensitive frequency spectra for effective inter-turn short-circuit diagnosis in loaded transformers.

show abstract

Section: Procedures Followed For the Analysismentioning

confidence: 99%

“…DABS gives the overall absolute difference between two sets of data compared and a result of '0' suggests no difference and the numerical value shows the severity of deviation [19]- [21]…”

Section: Procedures Followed For the Analysismentioning

confidence: 99%