Research On Key Technologies Of Wireless Monitoring And Evaluation Of Pump Group Based On Internet Of Things

Centrifugal pump is an indispensable key equipment in petrochemical industry and other production fields.During the operation of centrifugal pump,due to different reasons,there will be various kinds of faults.The light of these faults will cause the centrifugal pump to stop running,and the heavy will cause major accidents and serious economic losses.Effectively monitor and evaluate the operation status of the centrifugal pump,promptly remind the staff to carry out maintenance when the centrifugal pump fails,which can avoid accidents and provide technical guarantee for the safe and stable operation of the centrifugal pump.In this paper,the centrifugal pump as the research object,in view of the shortcomings of the existing state monitoring and health evaluation methods,a wireless sensor for the centrifugal pump wireless monitoring is designed,which realizes the wireless perception of the monitoring system,and enhances the convenience and usability of the system on-site implementation.This paper puts forward a method of centrifugal pump health evaluation,which can not only evaluate the overall health status of centrifugal pump,but also monitor and analyze the change of the data reflecting the fault symptoms of centrifugal pump through index decomposition,so as to avoid the occurrence of missing and false alarm.A wireless monitoring and evaluation system of pump group based on the Internet of things is established.The health evaluation model of centrifugal pump is built by the way of collaborative calculation of edge side and server cloud.According to the common failure forms and characteristics of centrifugal pump,the frequency band and alarm threshold are divided adaptively,and the status evaluation of overall and single failure indexes is carried out.The main research work and achievements of this paper are as follows:(1)An adaptive health evaluation method of centrifugal pump based on frequency band feature extraction is proposed.According to the common failure modes of centrifugal pump,the vibration spectrum of centrifugal pump is divided into several characteristic frequency bands,and the failure characteristics are extracted.According to the characteristics of water hammer fault,a method of rapid identification of water hammer fault is proposed: firstly,the vibration signal is processed by high pass filter,then the dimensionless indexes such as kurtosis and margin of filtered signal are calculated and analyzed to identify water hammer fault quickly and effectively.Both of them make up the whole centrifugal pump and single fault evaluation model.(2)The wireless sensor nodes are designed,and the system design is optimized by edge computing.In the way of edge cloud collaboration,the fault features of centrifugal pump are extracted at the sensor node end,the health assessment of centrifugal pump operation state is carried out at the gateway end,the probability model of equipment state is established at the cloud server end,and the alarm threshold required for health assessment is divided adaptively;the system is optimized to reduce the energy consumption and transmission cost of wireless sensor node.(3)The self-adaptive health evaluation method of centrifugal pump and water hammer fault monitoring and diagnosis method are verified by experiments.The monitoring and evaluation system of centrifugal pump based on the Internet of things is established.The correctness and reliability of the proposed monitoring and selfadaptive evaluation method for centrifugal pumps are verified by the experiments of normal state,unbalanced fault,misalignment fault,impeller defect,bearing fault and water hammer fault.