Research On Control And Modeling Of Voice Coil Motor Under High Frequency Response

Voice Coil Motor (VCM) is a special kind of electromagnetic device, which generate linear motion or swing motion directly without any intermediate devices. It has many features such as high frequency response, high speed, high acceleration, high accuracy, simple structure, small size, light weight, strength characteristics, and easy control. In the semiconductor wire bonding, VCM is also widely adopted for its high positioning accuracy such as micron, submicron, and even nanometers. In order to improve efficiency, VCM works at high frequency, for example, the first-generation bonding machine work in the 10 lines/sec; the second-generation bonding machine works at 25 lines/sec and the third-generation bonding machine will reach 40 lines/sec, but many experiments show that the VCM positioning inaccuracy for complex hysteresis under 40Hz.Research on control and modeling of VCM under high frequency response is supported by the National Natural Science Foundation and other projects. The main research contents of this paper are as follows:1) The history of VCM is reviewed, the structure of VCM is described and the research situation of VCM is introduced. The paper points out that there exists a problem of complex hysteresis for VCM in the case of high-frequency, high-speed, and this complex hysteresis is non-monotonic phenomenon.2) The characteristics of complex hysteresis are analyzed and the C-S hysteresis model is improved to adapt the non-monotonic hysteresis, then based on which, the C-S hybrid neural network hysteresis model is proposed. The modeling accuracy is less than 0.44% which can meet the requirements of modeling for complex hysteresis.3) In order to improve modeling method for complex hysteresis, the dynamic complex hysteresis model based on neural network is proposed. Neural network can accurately approximate any complex nonlinear curve, but it can’t achieve on multi-valued hysteresis function. Thus, the previous input value of current input value is used as an auxiliary input of neural network. The experimental results show that the model also has high accuracy for complex hysteresis modeling.4) The complex hysteresis inverse model is proposed based on neural network dynamic complex hysteresis model for the inverse of complex hysteresis is a complex hysteresis phenomenon. The complex hysteresis inverse model controller is designed based on neural network dynamic complex hysteresis inverse model.5) The control algorithm is simulated on the platform of MATLAB. The simulation results show that the complex hysteresis is eliminated. Complex hysteresis inverse model controller is verified on the platform of cSPACE, and the control method is proved to be effective.Inverse model control is a kind of control strategy and control technology that is advanced. VCM complex hysteresis inverse model controller is designed in this paper, by adopting the theory of inverse model control. The experiments show that the VCM complex hysteresis is eliminated by the complex hysteresis inverse model controller and the system’s tracking performance is improved significantly. Compared with other control methods, the complex hysteresis inverse model controller can improve the control effect of VCM, which has the advantages of simple structure and easy to obtain for parameters.