Research On Micro Wire Electrical Discharge Machining Constant Tension Control System

Micro wire electrical discharge machining(MWEDM) technology is themaching method of using the high temperature generated by the high-frequencypulse between the electrode wire and workpiece to corrode and machine micro smallparts. With the growing demand for high-precision micro parts, the accuracy ofMWEDM becomes one of the most important issues in the current research.Therefore, as the key part of MWEDM, the electrode wire tension control system isestablished in this research. Since wire constant tension control system is astrongcoupling, nonlinear and time-varying system. The design of the electrode wiretension control system become an intractable control problem.The mechanical design, hardware architecture and the control algorithms ofMWEDM tension control system are given in this paper, repectively. According tothe practical requirements of MWEDM, the wire drive mechanism of the electrodewire tension control system is designed, the corresponding servo hardware and theinterface design of control software are set up initially.An optimal dual-motor constant tension control system is developed. In orderto eliminate the tension measurement noise caused by the vibration of wire electrodeetc., Kalman filtering algorithm is proposed to fuse tension and speed sensor. Theaccurate dual input dual output mathematical model of wire drawing and tensionwheel is established, and least square method is used to determine the discreteparameter identification model. The minimum performance index which is the sumof the wire speed and tension squared error is achieved by optimal control, whichmake the accuracy of the tension improved greatly. To overcome the input dead zoneand interference effect the disturbance observer and robust adaptive control methodis employed for the actual processing environment. The simulation results show thatthe expected control performances are achieved.On the basis of the control algorithm simulation, servo motor performance testat very low speed and Kalman filtering effect test are implemented, the experimentalresults show the usefulness of the hardware selection and filtering algorithm. Theeffectiveness and superiority of the proposed control algorithm is demonstrated byseveral comparative experiments under different control algorithms.