| Compared to the established aeroengines condition mornitoring and fault diagnosis & prognosis system, the aviation hydraulic pumps’counterpart system is still developing. According to the characteristics of the aviation hydraulic system and its pumps, this paper is aimed at building an aviation hydraulic pump health prognostic simulation platform with the application of computer aided testing technique, pattern recognition and status parameter prediction methods.The platform can be used for the research on condition based maintenance and health management of aviation hydraulic pump. The construction process can be used as a reference for the design and development of similar platforms.The pressure compensated constant pressure variable pump is selected as the research object, and its model is established in AMESim simulation software. The plunger pumps’ typical faults has been investigated. After analyzing the friction pairs of plunger pumps, and compared with other fault injection methods, a scheme to inject internal leakage into the plunger pump is put forward. The feasibility of internal leakage fault injection is verified by simulation experiment.According to the computer aided test (CAT) technology, the hydraulic circuit, data acquisition module circuit, electrical control, process control program of the simulation platform for the health state prediction of the plunger pump has been designed. The platform supports are designed in CATIA. The process control program of the simulation platform is designed by using the LabVIEW software. The simulation model of the hydraulic circuit is built in AMESim to verify the rationality and validity of the hydraulic circuit design of the platform, and to provide the basis and reference for the selection of components. Through the software simulation, the non destructive internal leakage fault injection method of the simulation platform plunger pump is determined. The process control program, data acquisition and display program of the simulation platform is designed by using the LabVIEW program development software. Simulation platform can automatically monitoring the output of the plunger pump circuit, accomplishing the data display and storage.On the simulation platform, the normal state of the plunger pump and the output pressure and flow data of the fault state are collected.The fault diagnosis of plunger pump is carried out by using BP neural network, continuous wavelet transform, empirical mode decomposition and radial basis function network.The3 kinds of fault diagnosis algorithms can be used to diagnose the leakage fault of plunger pump.Using the support vector regression machine, the output flow and pressure of the plunger pump in the fault state and the normal state are predicted. Support vector machine parameter prediction algorithm results show that flow and pressure prediction error is not more than 1%. |
PDF Full Text Request