| Summary: | 碩士 === 國立交通大學 === 電機與控制工程系所 === 94 === Digital video cameras (DVC) have be more and more popular in recent years. Many technologies which can improve the qualities of image and video are rapidly been developed, such as anti-shocking (AS) and auto-focusing (AF). Conventionally, auto-focusing system is usually applied to high-level cameras and stepper-motor is mostly used to be the driving plant of the AF system with the simple proportion-integral (PI) controller or proportion-differentiation (PD) controller. As for some manufacturers of DVCs in Japan, they adopt high-precision processes in manufacturing the apparatus of the AF system and that high-precision processes can make the friction of the AF apparatus be more uniform; these methods can simplify the complexity of the control system and reduce the influence of the friction on the AF apparatus.
Stepper motor which is used to be the plant of the AF system is the convenient and simple design. However, the responsed time of the conventional AF system needs 300∼500 millisecond (ms); for taking pictures, it needs more focusing time in order to capture a pellucid image. Moreover, the slow focusing is a very unfavorable defect for motion-video recording.
Therefore, the voice-coil motor (VCM) with faster response time and high energy efficiency is used to this study. Because the electrical apparatus consumes about 30% power energy of the total supplying electric energy of a DVC, so a low-power and high energy efficiency VCM design is the first topic of this study. In the other hand, the friction of the AF apparatus brings many non-ideal control problem such as larger steady-state error and moving vibration. In order to overcome the effect of the friction, the disturbance observer with PI controller feedback system is applied to implement this AF system control. Using this mothod can also reduce the position disturbance which is caused by the change of inclination. Besides, the mini magnetic scale, mini magneto-resistive (MR) sensor and positioning estimating algorithm instead of using expensive interpolation chip are used for cost-down and mini-size; that positioning estimating algorithm can be used to calculate the current position and velocity.
The results of this study, the responsed time needs about 100∼120 ms with the 10% steady-state error; the 10% steady-state error is caused form the defects of the magnetic scale such as the variations of the amplitudes of the MR sensor signals. The disturbance observer offers the better compensation and the effects of the friction and weight force have been abated. The maximum current consumption of the VCM in this study is about 30 milliampere (mA), this advantage can save more battery energy of the DVC.
|
|---|
