基于综合赋权法与模糊理论的水电机组转轮碰磨故障判定  

Determination of the Rotor Grinding Fault of Hydropower Unit Based on Combined Weighting Method and Fuzzy Theory

作　　者：刘东 李友平[1] 任文锋[1] 孔丽君[1] 肖志怀[2] 王云鹤 LIU Dong;LI You-ping;REN Wen-feng;KONG Li-jun;XIAO Zhi-huai;WANG Yun-he(China Yangtze River Electric Power Co.,Ltd,Wuhan Hubei 430000,Hubei Province,China;School of Power and Mechanical Engineering,Wuhan University,Wuhan 430072,Hubei Province,China)

机构地区：[1]中国长江电力股份有限公司,湖北武汉430000 [2]武汉大学动力与机械学院,湖北武汉430072

出　　处：《中国农村水利水电》2024年第12期156-161,共6页China Rural Water and Hydropower

基　　金：中国长江电力股份有限公司资助“融合多源故障征兆指标的水电机组轴系故障诊断关键技术研究”(1521020003)。

摘　　要：为了提高水电站机组运行的安全性和稳定性,实时监控机组的运行状态,准确有效判定水电机组转轮碰磨故障发生概率,提出了一种基于主客观综合赋权与模糊理论相结合的故障判定模型。首先提取转轮碰磨故障对应的征兆指标,构建“故障-征兆指标”的判定结构体系;其次通过层次分析法(Analytic Hierarchy Process,AHP)和优劣解距离法(Technique for Order Preference by Similarity to an Ideal Solution,TOPSIS)获得各征兆指标主、客观权重,基于极差最大化法计算征兆指标综合权重;采用高斯阈值法确定振摆类征兆指标限值,同时采用行业规范确定温度类征兆指标限值;基于征兆指标限值确定水电机组征兆指标偏离度,构建“偏离度-隶属度”映射函数,计算获得隶属度判定矩阵;将隶属度判定矩阵与综合权重相结合得到故障判定矩阵,根据故障判定矩阵的最大隶属度反映该故障发生概率。以国内SK水电站故障案例实际运行数据为例,对转轮碰磨故障判定模型进行验证,判定结果与实际故障情况相符。In order to improve the safety and stability of hydropower unit operation,monitor the unit′s operation status in real time,and accu⁃rately and efficiently determine the probability of the rotor grinding fault of hydropower units,a fault determination model based on the combi⁃nation of subjective-objective combined weighting method and fuzzy theory is proposed.Firstly,the symptom indexes corresponding to the ro⁃tor grinding fault are extracted,and a“fault-symptom index”judgment structure system is constructed.Secondly,the Analytic Hierarchy Process(AHP)and Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS)are used to obtain the subjective and objec⁃tive weights of each symptom index,and the comprehensive weights are calculated based on the maximum extreme difference combination method.The Gauss threshold method is used to determine the limit value of oscillating signs,and the industry standard is used to determine the limit value of temperature signs.The deviation degree of hydropower units is determined based on the limit of signs,and the mapping function of“deviation degree-membership degree”is constructed,and the membership degree judgment matrix is obtained.The fault judg⁃ment matrix is obtained by combining the membership decision matrix with the comprehensive weight,and the fault probability is reflected according to the maximum membership degree of the fault judgment matrix.Based on the actual operation data of the SK hydropower station,the fault judgment model of rotor grinding fault is verified,and the judgment results are consistent with the actual fault situation.

关 键 词：故障诊断 模糊层次分析 综合权重法 优劣解距离法 最大极差组合法 

分 类 号：TK730.8[交通运输工程—轮机工程]