Research And Development Of Remote Condition Monitoring And Fault Diagnosis System For Pump Sets Conditions Of Pumping Station

For the large interbasin water transfer project, the pump sets of pumping station need long time running and high reliability. The pump sets failure may cause great economic losses and serious social impact. Thus, the research about remote condition monitoring and fault diagnosis technology of pump sets has a significant benefit which will ensure the safety of engineering, economic operation, promote the maintenance and management level of pumping station, as well as express the overall effectiveness in engineering. This paper focused on the analysis of the key technology of pumping station supervision and fault diagnosis technology of pump sets. A man-machine cooperation fault diagnosis method is put forward. And the implementation method of remote condition monitoring and fault diagnosis is explored.Firstly, the functional requirement of the remote condition monitoring and fault diagnosis system of pump sets of pumping station are put forward based on the review of current situation and trend about machine set condition monitor and fault diagnosis in domestic and abroad. In the second Chapter, the system structure is presented. Besides, the hardware composition and software architecture are also described.In the third Chapter, the features of typical faults such as Unbalance, Oil whirl, Oil whip, Misalignment, action a friction is listed based on analyzing and classifying some significant documents; The simulation results of typical faults of Unbalance, Oil whirl, Oil whip, Misalignment in the laboratory, validated the correspondence between fault causes and fault characters in the form of a relation diagram.Since the traditional fault diagnosis method based on vibration signal spectrum possess some disadvantages of low credibility, incomplete information, a man-machine cooperation fault diagnosis method based on multi-information fusion is presented in the Chapter4. The man-machine cooperation fault diagnosis method combined three methods of BP neural network, fuzzy operation and the man-machine cooperation. Firstly, it uses BP neural network or fuzzy operation to reason possible fault causes. Then, the waveform features and the axis path shape information will be obtained by man-machine dialogue. Finally, the expert system is used to "reverse chain" reason credible fault causes based on the preliminary result, phase characteristic, waveform features and the axis path shape information. In this chapter, the fault diagnosis method based on BP neural network is introduced in detailed firstly. Then, this method is used to diagnose some typical simulative faults. The experimental results show that this method can recognize the typical rotating machinery fault such as rotor imbalance and misalignment. The fault diagnosis method which combining the fuzzy operation and expert system are introduced in succession. Then, a man-machine cooperation fault diagnosis method based on multi-information fusion is described in detail. It integrates multiplex information such as signal spectrum, wave character, axis path shape, phase position, tendency and so on. As well as the workflow of man-machine cooperation fault diagnosis is introduced briefly. In the end, the update method of diagnosis knowledge is discussed.The field condition monitoring system implemented with LabWindows/CVI is introduced in Chapter5. Several interfaces which include main interface, system configuration interface, system condition monitoring interface, time-filed analyses interface are described. The remote condition monitoring agent server is also described briefly.In the Chapter6, the remote supervision and fault diagnosis system of pump sets implemented with MeasurementStudio is introduced. Several interfaces which include main interface, condition monitoring interface, time-filed analyses interface are described. Several fault diagnosis interfaces with man-machine cooperation method are described in detailed.