GO法进行有共因失效系统可靠性分析的新算法研究

New algorithm for the system reliability analysis with the common cause failures via the GO methodology

共因失效可能使系统的可靠性大大降低,是系统可靠性分析中的重点和难点。GO法可以对含共因失效的系统进行可靠性分析。为了选取已有算法合适的参数,更为了提高有多个共因失效组系统的GO法可靠性分析效率,进行了相关研究。对关键参数公式进行了理论分析,并就几个算例系统进行了计算试验;对已有算法进行了深入研究,为了改进已有算法对共因失效组逐个处理,提出了将共因失效组中的部件均由本组等效共因失效部件代替的新算法。进而,应用分组法和新算法对一个简单示例系统和外电源备用系统进行了共因失效分析。给出了对已有算法参数选取的建议;通过对比可知分组法与新算法计算结果相近。通过对已有算法参数的合理化选取,可以提高计算结果的准确度;新算法可对所有共因失效组一次性处理,是共因失效分析的有效方法。

The given paper intends to make an analysis of the reliabilities of the method known as CCFs( short for the common cause failure) by using the so-called GO methodology. To solve the problem,the paper has to choose a rational preference from the 2 applicable algorithms: the new one and the old one,actually for the two purposes. One is to choose rational preferences of the old algorithm,whereas the other is inclined to heighten the efficiency for the system reliability analysis with the multiple common cause component groups( CCCGs). For the said purpose,it would be necessary to make a theoretical analysis of the formula of the key parameters involved in the old algorithm by using a few examples for the test. Actually,the fault probability of the system with CCFs involves 2 parts,one of which is without CCFs while the other involves the CCFs uniquely. However,a new algorithm can adopt in case to replace the equivalent CCF components in a CCCG for the components are to be used to gain a new quantified GO chart. As to the fault probability method of the system with CCFs,it should be pointed out that the method here consists of 2 parts,one of which is used without CCFs,whereas the other is particularly needed to be adopted along with CCFs. For instance,when resolving the fault probability of the system with CCFs,the old algorithm can only actually solve the CCCGs one by one along with the new one,for the equivalent CCF component in the CCCG can be used to replace the components in it to gain a new quantified GO chart. Therefore,the above-mentioned 2 algorithms can be used to analyze CCFs in a simple exemplary system and an offsite power standby system.Thus,by using Markov model in the offsite power standby system,the state transfer process of the 2 repairable components( each component with 2 fault modes) can be outlined,and then the relatively exact state probability of the system can be educed.Furthermore,the approximate formula can also be derived,for the CCF probability of the 2 components can be applied to the GO methodology. Choosing preferences for the old algorithm is suggested rationally. The above comparison indicates that the 2 algorithms can help to arrive at the very close results with hardly any discrepancy. By rational preferences of the old algorithm,it would be possible to enhance the veracities of the calculation approaches. Therefore, since the new algorithm can solve all CCCGs together,it should be regarded as an effective method for the system reliability analysis with CCFs.