A new algorithm for mixed Weibull analysis of partial discharge amplitude distributions

In this paper, an heuristic algorithm is presented for the evaluation of the parameters of additive-Weibull distributions having more than two elementary functions, which is a necessary condition when testing complex insulating structures with different partial discharge (PD) sources. The algorithm has, in principle, no limit in the number of component Weibull functions. Furthermore, it should be emphasized that it is not required to have a very close initial guess of the parameters for solution convergence, unlike other algorithms used at present. In order to check the validity of the proposed algorithm in finding the elementary components of a mixed Weibull function, many PD experimental tests have been performed on some lumped capacitance specimens in order to compare the experimental cumulative probability of pulse amplitude distributions (PAD) with the ones obtained from the Weibull analysis. In this aim, the results of the Cramer Von Mises (CVM) test performed on the experimental results are reported. Some considerations are also given in the case of testing commercial HV components having complex PAD.

[1]  E. Gulski,et al.  Computer-aided recognition of discharge sources , 1992 .

[2]  E Gulski,et al.  Computer-aided analysis of discharge patterns , 1990 .

[3]  Rosario Schifani,et al.  A computer‐aided instrument for pattern recognition of partial discharges , 1998 .

[4]  E. Kay,et al.  Methods for statistical analysis of reliability and life data , 1974 .

[5]  D. Kececioglu,et al.  Graphical representation of two mixed-Weibull distributions , 1992 .

[6]  John H. K. Kao.  A Graphical Estimation of Mixed Weibull Parameters in Life-Testing of Electron Tubes , 1959 .

[7]  Alfredo Contin,et al.  Use of a mixed-Weibull distribution for the identification of PD phenomena [corrected version] , 1995 .

[8]  John C. Fothergill,et al.  Electrical degradation and breakdown in polymers , 1992 .

[9]  R. Schifani,et al.  Measurement of partial discharge in solid dielectrics with a microprocessor-based system , 1992 .

[10]  Wayne Nelson,et al.  Applied life data analysis , 1983 .

[11]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[12]  Gian Carlo Montanari,et al.  Stochastic analysis of partial discharge patterns deriving from electrical tree measurements , 1995, Proceedings of 1995 IEEE 5th International Conference on Conduction and Breakdown in Solid Dielectrics.

[13]  R. J. Brunt,et al.  Stochastic properties of partial-discharge phenomena , 1991 .

[14]  G. C. Montanari,et al.  Use of a mixed-weibull distribution for the identification of PD phenomena , 1995, IEEE Transactions on Dielectrics and Electrical Insulation.

[15]  G. Cran,et al.  Graphical estimation methods for Weibull distributions , 1976 .

[16]  A. Contin,et al.  PD analysis of rotating AC machines , 1993 .

[17]  Gian Carlo Montanari,et al.  In search of convenient techniques for reducing bias in the estimation of Weibull parameters for uncensored tests , 1997 .