摘要
为很好地解决汽轮发电机组经常同时发生多种振动故障的问题,考虑到模糊c划分极易陷入局部最优的缺点以及遗传算法全局寻优的优点,提出了结合两者的机组振动多故障诊断新方法。它将模糊c划分的求极小值问题转化为遗传算法的求最大值问题,并用分类系数和平均模糊熵加以判别。经大量诊断实例分析证明,该方法既可正确判断汽轮发电机机组的单一故障,又可有效诊断振动多故障,诊断的可靠性和实用性较高。
The turbogenerator vibration faults have the character of variety. Many faults often occur synchronously. Considering the shortcoming of fuzzy c and the advantage of genetic algorithm, this paper introduces a new method based on fuzzy c and genetic algorithm. It converts the problem about minimum for fuzzy c to the problem about maximum for genetic algorithm, which is distinguished by the class coefficient and the average fuzzy entropy. From the practice, the new method can diagnosis the single fault and multi fault of turbogenerator and has higher reliability and practicability.
出处
《高电压技术》
EI
CAS
CSCD
北大核心
2005年第6期3-5,共3页
High Voltage Engineering
