| With the rapid development of information technology,the operation and management of large-scale pump units have become more digitized and intelligent.In order to better monitor and manage the operation status of pump units,improve their safety and stability,the construction of intelligent pump stations has emerged.Intelligent pump stations utilize advanced measurement,Internet of Things,big data analysis,artificial intelligence,and other new generation information technologies to achieve intelligent monitoring,optimized scheduling and operation,intelligent safety monitoring,intelligent maintenance,intelligent inspection,intelligent management,data analysis and visualization,as well as mobile pump station applications.In accordance with the goals of "comprehensive perception,interconnection and integration,and intelligent application" proposed by intelligent pump station construction,and in combination with the operation,maintenance,and management needs of the Gao Gang Pump Station on the Yangtze River,this paper focuses on the construction of a unified data platform for the pump station and the research of vibration signal fault analysis for water pump units.Based on this,a pump station equipment management mobile app has been developed to achieve status monitoring of major equipment in the pump station and vibration fault analysis.In chapter 1,the overview of the research status and development trends of pump station computer monitoring systems,condition monitoring and fault diagnosis technology of pump units,smart water conservancy,and digital twin technologies.In chapter 2,the architecture of the intelligent pump station system is introduced firstly,followed by an overview of the research tasks and research roadmap of this paper,building upon the existing research in the laboratory.In chapter 3,aiming at the issue of lack of collaborative and data-sharing capabilities among multiple systems in pump stations firstly,this research focuses on constructing a unified data platform for pump stations.The overall architecture of the pump station data platform is introduced.It includes four main parts:data modeling,integration of heterogeneous system data,intelligent data storage,and data sharing and publishing.Data modeling involves classifying and coding pump station equipment based on equipment type and component division,and further dividing the state parameters of different components.Unique identification codes are then assigned to each equipment,component,and state parameter based on the classification results for better management and tracking.Integration of heterogeneous system data involves the fusion of diverse and heterogeneous data using different data service interfaces.Intelligent data storage is achieved through various storage strategies to enable smart storage of pump station data.Finally,data sharing and publishing are implemented through data interface services for information exchange and sharing.In chapter 4,an analysis and research on the fault diagnosis method for water pump units based on Principal Component Analysis(PCA)was presented.The establishment of the algorithm model,determination of statistical quantities and control limits,and fault identification based on contribution plots of statistical variables were analyzed.The algorithm utilizes vibration data from normal operating conditions of the pump unit to construct a PCA model,and then uses the T2 and Q statistics for sensor fault detection.Based on the contribution plots of statistical variables,the faulty sensor causing the fault can be located.Experimental results showed that the model was able to identify fault signals,but there were cases where the statistical quantities exceeded the limits during normal operating conditions,indicating accuracy issues in the process monitoring and fault analysis of pump unit vibration signals using the PCA algorithm.Then,the autocorrelation analysis of vibration data shows that the vibration signal of the unit is dynamic,so the structured dynamic principal component analysis algorithm is proposed.In chapter 5,the structured dynamic principal component analysis algorithm is introduced for constructing a fault analysis model.The algorithm extracts dynamic and static features from the vibration data of pump unit with dynamic characteristics,and monitoring according to the constructed Td2、Ts2 and Q statistics.Finally,the simulation experiment proves that the proposed algorithm has a good monitoring effect on fault analysis of pump unit vibration signals.Compared with static PCA algorithm,this algorithm reduces the false alarm rate during normal operating conditions,and effectively identifies fault signals.In chapter 6,the hardware components of the intelligent pump station system based on supOS is introduced firstly,followed by the process of developing a pump station equipment management APP based on supOS industrial operating system.The pump station data modeling module and the pump station operation and maintenance management module are developed,which can achieve real-time monitoring of pump units’ operational status,data integration between different systems,historical data analysis of vibration signals,and realtime monitoring and fault analysis of list data. |
