Abstract
Disturbance/vibration reduction is critical in many applications using machine vision. The off-focusing or blurring error caused by vibration degrades the machine performance. In line with this, real-time disturbance estimation and avoidance are proposed in this study instead of going with a more familiar approach, such as the vibration absorber. The instantaneous motion caused by the disturbance is sensed by an attitude heading reference system module. A periodic vibration modeling is conducted to provide a better performance. The algorithm for vibration avoidance is described according to the vibration modeling. The vibration occurrence function is also proposed, and its parameters are determined using the genetic algorithm. The proposed algorithm is experimentally tested for its effectiveness in the vision inspection system.
