A Method for Online Identification of Oscillatory Type Machine Model Including Nonlinear Friction (1'st Report)

—Development of a Practical Method for On-line Identification—

Abstract

The linear actuator driven by a ball screw mechanism is widely used in the machine tools, since it can raise the machine speed higher. For the purpose of the stability in continuous path control of this type mechanism, it is important to identify the machine parameters, such as inertia, spring stiffness and dominant nonlinear friction. This paper presents a practical method for identifying the nonlinear parameters of the table driven by a ball screw mechanism. In the proposed method, the mechanical system is modeled as two-inertia system. Then, the two-inertia system is represented as an inverse model and the inertia and Coulomb friction of the table and the spring stiffness are identified by using the steepest descent method with a dead band. From the results of the computer simulation, the identified mechanical parameters quickly approach to the true mechanical parameters without measuring the velocity of the table.