Real-Time Vibration Control of Rotor-Bearing System Based on Artificial Neural Networks and Active Support Stiffness

Abstract

A real-time dynamic response of the rotor-bearing system is controlled through an active support stiffness designed and constructed for this purpose. It consists mainly of a variable, flexible beam length. A stepper motor with a screw is implemented to manipulate the beam length to the required optimum position. Hence, it works as active spring stiffness, which minimizes the vibration response for the system. Stiff support is fixed at a specified position on the beam, and it is provided with a small ball bearing at the point of contact with the rotor at the other end. An artificial neural network has been used to control the dynamic system response. Response simulation with real-time LabVIEW is conducted to play a role as an interface to deal with the required sensors' records and the rotation of the stepper motor. The results show that the controlled system is efficient in obtaining the optimum support stiffness for different rotational speeds of the driven motor, which gives, in turn, the optimum system dynamic response.