摘要
汽轮机故障频率存在非线性、诊断复杂、相关参数属性集庞大的特点,应用决策树C4.5算法进行分类,速度快,准确性高。首先对汽轮机故障频率的连续性进行离散化,对决策表属性约简,得到分类规则,做为实际应用的依据。以Matlab为开发平台构建汽轮机振动故障诊断系统,准确率达到86.7%,具有实际应用价值。
Steam turbine fault frequency has the characteristics of nonlinear,complex diagnosis and large set of related parameter attributes.The application of decision tree C4.5 algorithm for classification is fast and accurate.Firstly,the continuity of turbine fault frequency is discretized,the attributes of decision table are reduced,and the classification rules are obtained as the basis for practical application.Using Matlab as the development platform,the turbine vibration fault diagnosis system is constructed with the accuracy of 86.7%,which has practical application value.
作者
侯国安
肖东彩
田舟祺
HOU Guoan;XIAO Dongcai;TIAN Zhouqi(Yinchuan Energy College,Yinchuan 750015)
出处
《计算机与数字工程》
2022年第4期898-902,共5页
Computer & Digital Engineering
基金
国家自然科学基金项目“基于跨层安全机制的异构Mesh微电网分布式协同控制研究”(编号:62173256)
银川能源学院校级重点科研项目(编号:2019-KY-Z-01)资助。
