Tracking Control Of Micro Hard Disk Servo System

Micro hard disk is developing toward small volume and big capability, which results in high requisition for the positing precision of slider head. Therefore, it is necessary to investigate tracking control of micro hard disk servo system so as to enhance its control performance. This urges us to study two important control problems: single-stage control and dual-stage control, in which their theoretical analysis and numerical experiments are considered. Main research work is summed up as follows:A. In order to decrease the positing time of the slider head in the micro hard disk servo system, a tracking control scheme is built upon the existing time-variant linear system to describe the servo system. In this research, a reduced order observer and its corresponding controller are designed in terms of Lyapunov stability theory and linear matrix inequality, and accordingly, the stability of the system is achieved. The control performance is compared to those of other relevant controllers through numerical experiments. The results show that the proposed controller can step up the slider head to track the given reference signal.B. Study the problem of disturbance attenuation for the micro hard disk servo system associated to the existing time-variant linear system with stochastic disturbance. This is solved through designing a nonlinear observer and corresponding feedback controller with saturation function. Depending upon nonlinear stability theory, the system is globally uniformly ultimate bound. Experimentally, the controller is examined by using multiple disturbance functions. Numerical experiments show that with the controller the slider head can rapidly track the reference signal and maintain it within an admissible scope of the specific track, while disturbance attenuation can be achieved rapidly.C. Single-stage control is difficult to make the slider head achieve high read/write precision, due to severe resonance of E-block in the micro hard disk. Consequently, dual-stage control is studied for the micro hard servo system in which the main work is to design an observer and corresponding controller. In design of the controller, an adjustable control parameter is introduced to ensure that the output of the microactuator's controller can effectively compensate the tracking error of the voice coil motor. In terms of Lyapunov stability and linear matrix inequality, the stability of the system is satisfied. Numerical experiments show that, while compared to the control effect in the case of single-stage control, the total output of the slider head can rapidly track the given reference signal with high positing precision and strong noisy attenuation.D. considering that neural network has the capabilities of nonlinear approximation and online control, the dissertation attempts to utilize neural networks to cope with dual-stage control of the micro hard disk servo system. The mathematical models of the actuators, microactuator and voice coil motor, are identified by means of the neural networks, and further the controller for each of the actuators is designed by the PID controller network for which initial weights and its learning rate are determined by the genetic algorithm with elitism. On the other hand, the controllers are applied to the micro disk dual-stage servo system through numerical experiments. The results show that the output of the controller for the microactuator can compensate the track-following error of the voice coil motor, which also demonstrates the reasonability of the designed control scheme.