engleski

Using exponential penalty functions in optimizing bearing support dynamic characteristics

This paper deals with the optimization of the support dynamic characteristic of a simple rotor bearing system by finding the values of support stiffness and damping at which a chosen objective function has its minimum and at the same time satisfying the limitations of the maximum allowable vibration amplitudes and forces transmitted to the bearings and to the foundation. Penalty function method is used in the way that constraints are taken into account by extending the original object function with an exponential penalty function for each limitation. The minimum is found by a numerical algorithm that solves a set of non-linear equations directly obtained from the necessary condition for a local minimum whose gradient of the function is equal to zero.

rotor-bearing-support system ; support dynamic characteristic ; optimization ; exponential penalty functions

nije evidentirano