Multi-fault diagnosis of electro hydrostatic actuator based on signed bond graph
Aiming at the problem of multiple fault diagnosis of electro hydrostatic actuator(EHA),a signed bond graph(SBG)fault diagnosis method considering the contributions of parameter was proposed.Firstly,based on the EHA working principle and bond graph theory,a bond graph model was constructed.System parameter values and distribution standard deviations were set,and then the bond graph model was used to derive its analytical redundancy relations(ARR)and construct the fault signature matrix.The bond graph model was transformed into a signed bond graph model,and possible fault sets for the EHA were inferred based on the causal relationships and consistent paths among the signed bond graph elements.Then,the contribution of EHA system parameters was defined and calculated.High-contribution system parameters were selected,low-contribution system parameters that were difficult to cause changes in analytical redundant residual data in possible fault sets were removed,and the range of possible fault parameters was narrowed down.Finally,the bond graph simulation software 20-sim was employed,and multiple fault simulation experiments were conducted by changing the values of the Q-axis resistance and motor magnetic flux chain in EHA.The proposed multi-fault diagnosis method was validated.The research results show that when multiple faults occur in EHA,the combination of the possible fault set that eliminates the low contribution system parameters with sensor data and analytical redundancy residual data can accurately diagnose the two preset high contribution fault parameters.Comparing with the possible fault set of conventional SBG fault diagnosis,the EHA system fault parameters can be locked more accurately.
electromechanical system reliabilityelectro hydrostatic actuator(EHA)signed bond graph(SBG)analytical redundancy relations(ARR)possible fault setmultiple fault diagnosisparameter contribution
NETL
NSTL
万方数据
