Study On Multi-source Information Fusion Method Of Low-voltage Switchgear

The low-voltage switchgear is used in some production sites, public places, residential buildings and other sites as well. Improving the reliability of low-voltage switchgear has great significance for production and life. But traditional low-voltage switchgear reliability evaluation method has long test period, expensive test price and few test data. For solving this problem, low-voltage switchgear reliability evaluation method which is based on Bayes theory attracts the attention of researchers. This paper presents low-voltage switchgear reliability evaluation method which is based on Bayes theory. And the main content of this paper is multi-source information fusion which is the key part of the method.Firstly, after deeply studied Bayes theory, low-voltage switchgear reliability evaluation method which is based on Bayes theory is presented. The sales information and maintenance information of low-voltage switchgear is collected from a factory as prior information. To prove the prior information and field test information of low voltage switchgear obey the same type of distribution, the prior information makes credible analysis and compatibility test.Secondly, prior information of low-voltage switchgear is proved obeying Weibull distribution by using Weibull distribution test method. Then the prior distribution of low-voltage switchgear which is the result of multi-source information fusion is obtained by using maximum entropy multi-source information fusion method and non-parametric Bootstrap method dealing with prior information of low- voltage switchgear in MATLAB.Finally, the robustness of the prior distribution of low-voltage switchgear is analyzed. And the low-voltage switchgear equipment reliability(mean time between failures for the first time) is evaluated which is based on the posterior distribution of low-voltage switchgear.