Journal of pplied Science and Engineering Innovation, Vol.5, No., 8, pp. 7-79 ISSN (Print): -96 ISSN (Online): -97 Research on Fault Location lgorithm with Traveling Wave Based on Determinant of Matrix Lianjun Hu, Ling Tang, Penggao Wen, Hong Song College of Mechanical Engineering, Sichuan University of Science & Engineering, Zigong, China. College of utomation & Information Engineering, Sichuan University of Science & Engineering, Zigong, China. bstract: In view of the dead judgment on the in the location with traveling wave based on the existing T circuit, a location algorithm with traveling wave based on determinant of matrix is proposed. This algorithm does not need to judge the separately, the matrix is constructed by the arrival time of traveling wave and wave beam, line length obtained by three ports on T circuit after ing, the and location can be obtained directly by comparing the size of the determinant of matrix. Through simulation analysis using MTLB, the algorithm with clear analysis and high precision meets the requirements of location. Keywords T circuit; Matrix ; Traveling wave; Fault location INTRODUCTION long with the structure distribution of the power grid, the growth of regional power load and the increase of the density of the power grid, T circuits have been more and more used in high voltage and extra high voltage power system. The power system is connected to large power plants and substations, once the occurs, it is need to find the location as soon as possible and recover the power supply. The location of the T circuit consists of two parts: one is the judgment of the ; the other is the judgment of the location. t present, the location algorithms with traveling wave for T circuit are single ended and double ended location methods. For single ended method, the literature [YU et al., 7] injected the signal at one end of the line when the occurs. By recording the start time of the injection and collecting the time when the signal reached the point, the location can be fixed using distance measurement formula. Thus the method has many problems, such as great difficulties in the data processing; signal attenuation to traveling wave in transmission process; inaccurate back time captured; how to identified in the multi network and so on. For double ended method, three endpoints in T circuit are proposed as the reference point in the literature [Li et al., ], select a fixed reference point, and then convert the into two separate lines with double end location method respectively. The method can only be used for different power lines, otherwise it will cause the. Literature [Li et al., 4] combining the single ended and double ended traveling wave method for location, but the traveling wave attenuation can not be avoided in the single end positioning, which causes the traveling wave head difficult to be identified. In the literature [LU et al., ], three es of the T circuit are transformed into two lines for location with double ended location method, but when the occurs in the vicinity of the node, there may be a dead zone which leads to the wrong point. The new algorithm constructs the matrix with the arrival time of traveling wave and wave beam line length obtained by three ports on T circuit after ing, the and location can be get directly by comparing the determinant of matrix. LGORITHM BSED ON THE DETERMINNT OF MTRIX The schematics of T circuit C l T Figure T circuit simulation diagram s shown in Figure, B C are the three ports in T circuit,,, are the measurement points, T is the connection point, with the corresponding line length l l l. When the occurs in one point, the traveling wave will spread at B C ports. The corresponding arrival time t t t can be measured at points respectively. lgorithm principle Consider the three es in T circuit with l =l =l l <l =l l >l =l l <l <l. When a occurs at any time, the time series l l B 7
J. of ppl. Sci. and Eng. Inno., Vol.5 No. 8, pp. 7-79 t t t t of traveling waves propagating to the three ends are obtained. The observation point is the reference point. Based on the principle of double end distance measurement, the distance between point on line B and C and observation point is calculated respectively. ssume the distance is x, and the distance to the node T is x ', as shown in Figure : C l T x l B For the different situations on the es, the determinant of matrix can be used to identify the. x ' is the distance between the and T node in equation(7), the corresponding points can be obtained when getting the es. The occurred at T, as shown in Figure : C l T x l B l x x l Figure The occurred at T Figure Simulation diagram after the occurs ccording to the same principle of traveling wave velocity, the formula () can be obtained: x l l x t t () v v so: l l vt t x () ' ' x l x l t t () v v then l l vt t l (4) and ' l l vt t l l vt t x x l (5) thus x' l vt l vt (6) ln tn Consider matrix n, where l n is the v length of the in circuit T, v is the velocity of traveling wave, t n is the arrival time. So l t, v x' x' When x' l v t, then (7),. occurs at T. x ', occurs at BT. x ', occurs at CT., 74 In this case the value of x'. can be calculated. Thus we can only get the without the location of the, at this point it is need to change the reference point. ssume the conversion reference point is, calculate x', where x ' is still the distance between the and T node, and the location can be obtained by the conversion. Flow chart of the algorithm In the actual power system, the length of power line has been determined. Figure 4 is the flow chart of location algorithm based on determinant of matrix. x=l +x Start Input t t t v l l l Calculate,, x = - < Fault at CT x = - >. Fault at BT Distance x=l -x End Fault at T x = - x=l -x Figure 4 The flow chart of location algorithm based on determinant of matrix SIMULTIONS ND FULT NLYSIS In order to verify the proposed algorithm, three end power system model with KV voltage is
J. of ppl. Sci. and Eng. Inno., Vol.5 No. 8, pp. 7-79 established using MTLB/Simulink as shown in Figure. In the simulation, wavelet transform is used to find the corresponding time after the modulus maximum of the current extracted by the linear model transform. Consider the simulation time is.s, the sampling frequency is MHZ, the start time of the is.s, the line parameters are: R =.7Ω/km; R =.864Ω/km; L =.97 H / km; L =4.64 H / km; C =.74 9 F/km; C =7.75 9 F/km; the velocity probability value of traveling wave is V / L C 8994km / s, simulation results [Silvio et al., 5] showed as follows:. Consider T, BT and CT with length of km, km, km, the simulation results of current extracting the modulus maxima after linear model transform and wavelet transform are shown in Figure 5. 5 -.5.5 x 4 - -.5.5 x 4-5 -.5.5 x 4.5.5.5.5.5 x 4.5.5 x 4.5.5 x 4 Figure 5 Results with at T BT CT Table ~ show the time series, matrix and solutions of ' x to determine the, and with the different types when the occurs the location and the absolute error of the point. at T, BT, CT respectively. Table 4~6 give the Table The measured time series under single phase earthing-short T BT 85 CT 5 (.46.74.74) (.6.95.6) (.8.8.79) 8994 8994 8994.46.95.8 8994 8994 8994.74.6.79 Table The measured time series under single short circuit between two phases T 85 BT 5 CT 5 (.64.88.88) (.778.74.778) (.468.468.5) 8994 8994 8994.64.74.468 8994 8994 8994.88.778.5 Table The measured time series under phase-to-phase grounding short-circuit T BT 5 (.7.89.89) (.99..99) 75 8994 8994.7. 8994 8994.89.99
J. of ppl. Sci. and Eng. Inno., Vol.5 No. 8, pp. 7-79 CT 5 (...98) 8994. 8994.98 Table4 The judgement of single phase earthing-short x ' x ' Fault Fault Error(m) 4.7.4 T 99.96-4 69.94 4.97 BT 85. -9.87 4.94 CT 5.6 6 Table5 The judgement of single short circuit beteen two phases x ' x ' Fault Fault Error(m) 69.88 4.94 T 85.6 6.5 5.75 BT 4.9-7 -4. 7 5.4 CT 4.86-4 Table6 The judgement of phase-to-phase grounding short-circuit x ' x ' Fault Fault Error(m).9 6.96 T.4 4 7.6 85. BT 4.87 -. 5. CT 4.89. Consider T, BT and CT with length of 6km, km, km, the simulation results of current extracting the modulus maxima after linear model transform and wavelet transform are shown in Figure 6. - - - -.5.5 x 4 -.5.5 x 4-4.5.5 x 4.5.5.5.5.5 x 4 Table 7~9 show the time series, matrix and with the different types when the occurs at T, BT, CT respectively. Table ~ give the.5.5 x 4 Figure 6 Results with at T BT CT Table7 The measured time series under single phase earthing-short.5.5 x 4 solutions of ' x to determine the, and the location and the absolute error of the point. T 6 BT CT 58 (.8.6.6) (.9.6.47) (.698.86.546) 6 8994 8994 8994.8.6.86 8994 8994 8994.6.47.546 76
J. of ppl. Sci. and Eng. Inno., Vol.5 No. 8, pp. 7-79 Table8 The measured time series under single short circuit between two phases T 75 BT 5 CT 89 (.6.985.985) (.57.88.95) (.59.79.65) 6 8994 8994 8994.6.88.79 8994 8994 8994.985.95.65 Table9 The measured time series under phase-to-phase grounding short-circuit T BT 7. CT 5 (.46.89.89) (.994.5.) (.6.64.9) 6 8994 8994 8994.46.5.64 8994 8994 8994.89..9 Table The judgement of single phase earthing-short x ' x ' Fault Fault.7.4 T 59.96-4 8. 9.7 BT 9.94-6 -84.8 4.4 CT 57.96-4 Table The judgement of phase-to-phase grounding short-circuit x ' x ' Fault Fault 7. 85. T 74.89-49.96 74.98 BT 5. -.4. CT 88.98 - Table The judgement of phase-to-phase grounding short-circuit x ' x ' Fault Fault 74.95 6.97 T. 55.44 7.7 BT 7.8-89.97 94.98 CT 5.. Consider T, BT and CT with length of 8km, 5km, 5km, the simulation results of current extracting the modulus maxima after linear - -.5.5 x 4 5-5 model transform and wavelet transform are shown in Figure 7. -.5.5 x 4 - -.5.5 x 4.5.5.5.5.5 x 4.5.5 x 4 Figure 7 Results with at T BT CT 77.5.5 x 4
J. of ppl. Sci. and Eng. Inno., Vol.5 No. 8, pp. 7-79 Table ~5 show the time series, matrix and with the different types when the occurs at T, BT, CT respectively. Table 6~8 give the solutions of ' x to determine the, and the location and the absolute error of the point. Table The measured time series under single phase earthing-short T 6 BT CT (.8.9.88) (.54.46.6) (.85.96.69) 8 8994 5 8994 5 8994.8.46.96 5 8994 5 8994 5 8994.9.6.69 Table4 The measured time series under single short circuit between two phases T 75 BT 5 CT 5 (.6.4.7) (.88.59.) (.6.788.864) 8 8994 5 8994 5 8994.6.59.788 5 8994 5 8994 5 8994.4..864 Table5 The measured time series under phase-to-phase grounding short-circuit T 5 BT 8 CT 65 (.59.88.978) (..77.75) (.7.46.6) 8 8994 5 8994 5 8994.59.77.46 5 8994 5 8994 5 8994.88.75.6 Table6 The judgement of single phase earthing-short x ' x ' Fault Fault 4.. T 59.99-5.4 5.7 BT 99.8 7-5.99 5.99 CT. Table7 The judgement of phase-to-phase grounding short-circuit x ' x ' Fault Fault.5 5.7 T 74.9-7 5. 75. BT 74.98 - -5.96.98 CT 5. 78
J. of ppl. Sci. and Eng. Inno., Vol.5 No. 8, pp. 7-79 Table8 The judgement of phase-to-phase grounding short-circuit x ' x ' Fault Fault 59.96 9.98 T 5. 9.6 45.8 BT 79.8 8-75.9 87.97 CT 65. The simulation results show that the location algorithm based on determinant of matrix has high precision and feasibility. With the same two es, the maximum absolute error is m; When the es are not the same, the maximum absolute error is 8m; otherwise the error maintains within m, which satisfy the engineering application. In addition the algorithm does not need to know the time at which occurred, only to know the arrival time of the traveling wave for location. CONCLUSIONS In this paper, a location algorithm based on determinant of matrix is proposed to avoid the judgment of the es in T circuit. With the calculation of the two lines, we can directly determine the, thus reducing the actual calculations and the errors. In practical application, the wave velocity can be determined directly by the line parameters, and the arrival time accuracy of current traveling wave is improved by the wavelet transform. When the occurs the can be accurately judged by the arrival time of the traveling wave with no misjudgement, and the measurement errors meet the needs of practical engineering. CKNOWLEDGMENT This work was supported by projects of rtificial Intelligence Key Laboratory of Sichuan Province (4RYY5,5RYY), and projects of Sichuan University of Science & Engineering (PY8). REFERENCES Cai Xiuwen Tan Weipu Yang Yihan,Single-Terminal Traveling Wave Fault Location for Distribution Network Based on Mathematics Morphology Theory[J],MODERN ELECTRIC POWER,6, (6): 5-9. Evrenosoglu C Y, bur li. Travelling wave based location for teed circuit[j]. IEEE Trans on PWRD, 5, (): 5-. Li Chuan-bing, TN Bo-xue, GO Peng et al. location method for T-connection lines based on D-type travelling wave theory[j], Power System Protection and Control,,4(8):78-8. LI Zhi-bin WU Bao-xing XU Yun-hui, pplied research of improved Hilbert-Huang transform in location of T-type transmission line[j], Chinese Journal of Power Sources,4, 8(): 9-96. LIN Fuhong, ZENG huimin, One-Terminal Fault Location of Signal-Phase to Earth Fault Based on Distributed Parameter Model of HV Transmission Line[J], Power System Technology,,5(4):-5. LU Yi, HN Zhi-kun, WNG Yi-chun, RO Shu-yong, Improved location algorithm based on traveling waves for Teed-circuits[J].Power System Protection and Control,, 9(5):7-,6. Silvio Giuseppe Di Santo, Carlos Eduardo de Morais Pereira. Fault location method applied to transmission lines of general configuration[c]. International Journal of Electrical Power & Energy Systems, Volume 69, July 5, Pages 87-94. YU Sheng-nan, YNG Yi-han, BO Hai, Study on location in distribution network based on C-type travelling-wave scheme[j], Relay, 7,5(), pp:- 4, 79