22 5 2018 5 Electri c Machines and Control Vol. 22 No. 5 May 2018 1 1 2 1 1 1 1. 300130 2. 318020 :,,,, c - : DOI: 10. 15938 /j. emc. 2018. 05. 012 TM 501 A 1007-449X 2018 05-0094- 08 Arc fault detection based on cluster analysis and electromagnetic radiation LI Kui 1 CHEN Zhao 1 ZHANG Yang-zi 2 WANG Yao 1 NIU Feng 1 DAI Yi-hua 1 1. Province-Ministry Joint Key Laboratory of Electromagnetic Field and Electrical Apparatus Reliability Hebei University of Technology Tianjin 300130 China 2. State Grid Electric Power Company of Taizhou Huangyan District Taizhou 318020 China Abstract Nowadays the current non-linear loads in the power system increase and the main drawback of fault arc detection method based on arc current characteristics is that there is only one information source which leads to misjudgment. A method of series arc fault diagnosis was presented based on the method of cluster analysis and the electromagnetic signals. On the basis of analying the theory of electromagnetic signal of arc the signal was filtered to reduce noise then combined with the research on electromagnetic behaviors of arcing fault the feature vector of the electromagnetic signals of arc fault was extracted. The fault is diagnosed by the fuzzy c-means clustering method. The result of test shows that the accuracy of the method for the identification of series arc fault meets the requirements. Keywords arc fault electromagnetic signals cluster analysis non-linear loads 2016-06 - 30 : ( QN2014148 ) ( E2015202143 ) ( 201545) ( LJRC003) : ( 1965 ),,,,, : ( 1990 ),,, ( 1992 ),,, ( 1981 ),,, ( 1986 ),,, ( 1990 ),,,
5 95 0 1-5 000 ~ 15 000 2 2 ~ 10 A 2 000 ~ 4 000 3-6 c - 0. 5 A 1. 1 7 1 3 1 8 1 10-4 cm 10 5 ~ 10 6 V /cm 9-13 17-18 12-13 Charles J. Kim 120 V 2 A 14 cluster analysis 16 1 Fig. 1 Schematic diagram of arc area c 1. 2 3 18
96 22 2 2 10-6 ~ 3 10-6 s 10 6 A /cm 2 10 8 2. 1 A /cm 2 18 3 UL1699 10 9 Hz 6. 4 mm 10. 0 mm 1. 3 19 x 2 R 220 V 50 Hz σ x E rad E rad R t = sinθ 4πε 0 Rc 2 δ 0 di dt dx 1 θ i R c P rad ω =. E rad t e -jωt dt 2 = 0 k 2 I 2 arc 0 ω 2 ω 2 0 - ω 2 + 4α 2 ω 2 2 k ω 0 ω 2 k 2 I 2 arc 0 ω 2 3 Fig. 3 Point contact arc test rig Fig. 2 2 x Electromagnetic field radiated by a current component oriented along x-axis 50 Ω
5 97 30 MHz 20 khz 15 0 ~ 250 khz 0 ~ 250 khz 4 25 MHz 1 1 m Fig. 4 4 Schematic diagram of experimental circuit 2. 2 6 Fig. 6 Waveform of measured nonlinear load current 4 MHZ 5 6 Fig. 7 7 Measured resistive load magnetic field signal waveform Fig. 5 5 Measured resistive load current signal waveform 7 8 8 Fig. 8 Measured nonlinear load magnetic field signal waveform 3 3. 2 3. 1
98 22 SIMULINK 2 3. 3 5 khz 9 2 2 1 Table 1 Table 2 Results of resistive load operation Number of electromagnetic signal extracted from test 1 1. 509 3 13 1. 451 5 1 13 2 14 3 15 4 \ 2 1. 375 3 14 2. 422 9 3 0. 094 250 15 0. 093 125 4 1. 434 3 16 2. 572 9 5 1. 772 8 17 0. 096 250 6 0. 096 250 18 0. 093 125 7 0. 096 250 19 2. 133 4 8 2. 946 8 20 1. 730 8 9 1. 791 6 21 0. 093 125 3 m 5 16 3 m 6 17 7 18 8 19 9 20 10 21 11 22 12 23 24 28 25 29 26 30 27 31 10 0. 096 250 22 0. 096 250 11 0. 096 250 23 0. 177 50 12 0. 323 37 / / / 3 3 Table 3 Results of nonlinear load operation 24 4. 650 6 28 4. 905 8 25 0. 033 703 29 0. 036 475 26 2. 397 3 30 3. 485 8 27 0. 025 460 31 0. 048 571 2 3 Z Fig. 9 9 Schematic diagram of filtering 10 11 C - FCM 4 4. 1 C - n FCM C -
5 99 x j w i u ij 0 1 u ij = X = x 1 x 2 1 x n FCM c V = v 1 v 2 v c v i w i l 1.d 2 ij = x j - v i T A x j - v i x j J min U V = N c u l ijd 2 ij j = 1 i = 1 U = u ij c N 3 u ij w i A = 1 Fig. 10 10 Magnetic field signal diagram of microwave oven Fig. 11 11 Magnetic field signal diagram of Vacuum cleaner FCM δ δ 12 3 4 Fig. 12 Results of cluster analysis for resistive load 12 13 12 23
100 22 1 1 17 22 6 11 7 18 21 15 3 23 12 5 9 20 4 13 1 2 14 16 19 8 2 8 14 16 19 3 8 14 16 19 5 9 20 4 13 1 2 3 17 22 6 11 7 18 21 15 3 23 12 8 14 16 19 5 9 20 4 13 1 2 2 14 Fig. 14 Results of cluster analysis for all the experimental data resistance and nonlinearity 300 13 Fig. 13 Results of cluster analysis for nonlinear load 4 1 2 6 4 8 1 5 3 7 2 3 7 3 1 5 3 7 2 6 4 8 3 Table 4 2 6 4 8 1 5 3 7 unknown loads 3 14 14 3 3 24 28 99 100 4. 2 4 Arc fault detecting test result of the 100 100 99 /%
5 101 5. IEEE Transactions on Industry Applica- emerging productj tions 1998 34 5928. et al. Series arc fault i- dentification in low voltage system based on autoregressive parame- 1 YONG Jing GUIXiaozhi NIU Liangliang ter modelj 2011 26 8213. 2 c - 10 D. 2013. 11 c - J. 2007 3 194. 98% 3 2007. 13 1ZHANG Ruixiang SONG Zhengxiang. Arc Fault Detection Method Based on Signal Energy Distribution in Frequency BandC. 2012 Asia-Pacific Power and Energy Engineering Conference. 27-29 March 2012 Shanghai China 1-4. 2IEC 62606 ed1. 0 General requirements for arc fault detection devicess 2013. 3Underwriter Laboratories Inc. UL standard forsafety for arc-fault circuit-interrupters UL 1699 J. 2008. 4. J. 2011 1 18. JIA Ning LI HongwenWEI Youping et al. Research of a novel arc fault circuit interrupterj. Low Voltage Apparatus 2011 1 18. - 28. 5. J. 19MichelMardiguian. M. 2007 3 7. 2008 45-47. CHEN Degui. The arc fault interrupter a new low voltage protection apparatus J. Low Voltage Apparatus 2007 3 7. 6CHAI Sim Por KUAN Lee Choo LIM You Jian. A study of arc fault current in low voltage switchboardc. 2012 IEEE Conference on Sustainable Utilization and Development in Engineering and Technology. 6-9 Oct. 2012 Kuala LumpurMalaysia 52-56. 7GREGORY G D SCOTT G W. The arc-fault circuit interrupter an 8. J.. 2011 26 8213.. Transactions of China Electrotechnical Society 9. D. 2010... YANG Jianhong ZHANG Rencheng DU Jianhua. Study on application of arcing protective system faults based on multiple information fusion J. High Voltage Apparatus 2007 3 194. 12. D.. J. 2012 7 36. 14KIM C J. Electromagnetic Radiation Behavior of Low-Voltage Arcing FaultJ. IEEE Transactions on Power Delivery 2009 24 1416. 15Micha el Rabla Patrick Schweitzer Etienne Tisserand. Method to Design Arc Fault Detection Algorithm Using FPGAC. The 57 rd IEEE Holm Conference on Electrical Contacts 11-14 Sept. 2011 Minneapolis MN. 16. J. 2014 19 68. PENG Xiangang LAI Jiawen CHEN Yi. Application of clustering analysis in typical power consumption profile analysisj. Power System Protection and Control 2014 19 68. 17. M. 1996 45-47. 18. M. 1991 25 ( : )