Application of voiceprint monitoring technology in the construction of smart pumping stations

Abstract In this paper the aim is to realize the real-time monitoring of the operating status of the equipment in the smart pumping station based on the principles of voiceprint monitoring and the characteristics of the pumping stations. To achieve this aim, the data of pumping stations were firstly collected, preprocessed and the feature values were extracted. Then, the data were used to train a convolutional neural network (CNN) deep learning recognition model. Finally, a voiceprint samples database and a model library of the operating status were established and a voiceprint monitoring system for the vibration state of the pumping station equipment was constructed based on the CNN deep learning recognition model. A case study of a mixed-flow pumping station was carried out. In this case study, a voiceprint detection system was arranged on site and the operation status of the equipment can be monitored and evaluated through practical application of the system. Further, the health status of the equipment can be effectively diagnosed and evaluated, and abnormal faults can be analyzed and predicted. The results indicate that the reliability of equipment operation has improved through the application of voiceprint monitoring technology which provides an effective technical means for the construction of smart pumping stations.