摘要
针对模糊C-均值算法在汽轮机故障诊断中的不足,提出了粒子群优化加权模糊聚类分析的方法。首先,采用基于样本相似度的特征加权方法对样本特征及样本进行加权,以适应各种复杂分布的样本;然后,利用粒子群算法优化加权模糊聚类的特征权值和聚类目标函数,并依据聚类有效性指标自适应确定最佳聚类数及聚类结果。试验结果表明,该方法具有收敛速度快和全局收敛的特点,有效降低了汽轮机故障诊断的误分类率,诊断结果可靠。
Aimed at the disadvantages of fuzzy C-means in fault diagnosis of steam turbine set,a weighted fuzzy clustering method based on particle swarm optimization is put forward.Firstly,the method adopts similarity based weighting method to assign feature weights and sample weights in order to handle the variety of samples with complicated distribution.Then,the particle swarm optimization with compression factor is used to optimize feature weights and clustering target function of weighted fuzzy clustering.Finally,the best clustering num and clustering result are adaptively obtained by clustering validity function.Application results show that the method reduces the misclassification rate in fault diagnosis of steam turbine set with the features of fast convergence and global convergence.It can diagnose single fault and compound faults with high reliability and practicality.
出处
《振动.测试与诊断》
EI
CSCD
北大核心
2011年第5期574-577,662-663,共4页
Journal of Vibration,Measurement & Diagnosis
基金
重庆市自然科学基金资助项目(编号:2008BB3179)
关键词
汽轮机
故障诊断
模糊C-均值
加权模糊聚类
粒子群优化
steam turbine fault diagnosis fuzzy C-means weighted fuzzy clustering particle swarm optimization