【摘要】 提出了一种基于模糊聚类的支持向量机算法,构建了适用于故障诊断的模糊支持向量机分类器,并把该算法运用到汽轮发电机组振动故障诊断中。对实验数据进行故障模式识别验证,结果表明该方法与传统的SVM方法相比,在保证分类器性能的前提下,可有效提高训练正确率。更多还原

【Abstract】 Combined the fuzzy clustering technology with support vector machine method,the new algorithm has applied to the vibration fault diagnosis.The fault diagnosis of fuzzy support vector machines algorithm has reduced the structure and complexity of support vector machine classifier,and has also reduced the number of 2-class classifier.The simulation results have showed that the diagnosis accuracy was increased by the algorithm with the guarantee of classifier’s performance.更多还原