An inverse prediction method considering the constraints of flow-path structure parameters was proposed to satisfy the designing requirements of aero-engine intershaft bearing lubrication system while obtaining feasible structure parameters corresponding to the given bearing lubrication efficiency. Lubrication efficiency function was constructed to demonstrate the fitting relation between structure parameters and lubrication efficiency according to the neural network model. By virtue of structure parameters optimization

the distraction between lubrication efficiency function values and the given lubrication efficiency was mininized

the constraints of each parameter were considered to obtain

the flow-path structure parameters satisfying the given lubrication efficiency(minimized distraction). In comparison with the present analytical methods

the proposed inverse prediction method could improve the prediction precision by 439%

shorten the average calculation time by 175 minutes

and predict multiple parameters simultaneously

providing an approach to the design of intershaft bearing lubrication system.