Development Of New Hydraulic System And Method Study On Fault Diagnosis In Filter Press

The content of this thesis is selected from the renovation project of hydraulic system for an acrylon plank filter press. On the base of analyzing the development of control system for filter press internationally, and combining the applicant development status of fault diagnosis for hydraulic system. factors which affect the capability of the whole system are found out and a series of schemes for the industry automatization renovation for filter press are put out according to factory demands, aimed at the phenomenon of heavily leak, much failure and low ability of keeping pressure.By the analysis for design limitations, fault phenomenon and the special aatlysis for technics flow, quickly rigid orientation hydraulic loop and automate unload loop are designed by using simplified design, energy-saving design and other modem design methods comprehensively. Furthermore, industry automatization and energy-saving for filter press are implemented by combining electric technology with hydraulic technology.Through special analysis for all components of hydraulic system in filter press, the mathematic model of the whole hydraulic system is built up by the method of bond-graph, and hydraulic system is simulated by MATLAB program.On the basis of faults of hydraulic system, fault modes and fault mechanisms are analyzed for filter press, which is the premise of fault diagnosis by neural network in hydraulic system. The points of fault signal are confirmed, and a series of signal disposal methods are adopted, by the analysis for status messages of hydraulic system.Based on the study on ART mode neural network, mathematic model process is analyzed, structure of neural network for fault diagnosis is studied, and the program for fault diagnosis in hydraulic system is made by MATLAB. Further, Sample system of fault diagnosis is built for hydraulic system in filter press.