归一化算法在混流泵温度监测评估中的应用

Application of Normalization Algorithm in Mixed Flow Pump's Temperature Monitoring and Evaluation

大型变频混流泵是城市原水供应中最为关键的设备,长期工作于频繁变化的工况下,常常造成故障的错判和误判,存在较大隐患。针对该问题,分析了大量同类设备在不同工况下实测温度及其影响因素,提出了一套针对大型混流泵的温度监测的归一化评价方案。该方案针对无故障机泵在典型工况下温度测量值,采用二元函数平方逼近算法,确定测量温度相对转速和环境温度的曲面拟合二元函数,利用该函数对任意转速和环境下的温度值进行处理得到归一化温度值,并建立归一化评价标准。归一化温度值剔除了环境因素和转速因素的影响,较好地反映了设备运行的劣化状况。该评价方案在机泵监测和健康评估过程中提高了准确性和预知维修水平。

Large-scale mixed flow pump is the most critical equipment in raw water supply of city. Working in a condition changes frequently for a long time, it often gives some wrong judgments and bring large hidden dangers. In order to solve this question, by analyzing the real temperature of vast similar equipment in different working condition, a normalization assessment scheme is put forward which aims at the temperature monitoring to large-scale mixed flow pumps. Specific to the temperature measurement of the normal machines and pumps in typical working condition, dual function square approximation algorithm and curved surface fitting are adopted to assure the binary function of rotating speed and ambient, then, the speed and the environment temperature measurement can be normalized. The gotten measurement eliminates the influences of environment and speed, which better reflects the degradation condition of the working equipment. This scheme improves the accuracy of the pump monitoring and the health assessment system, and predicts maintenance level equipment.