Design and control of an automatic high-viscosity sealant filling system

Abstract

In this article, a novel vision-based automatic liquid filling control system is proposed. The method can be used for many high-viscosity liquid filling applications, especially for packaging of fuel tank sealant in the aerospace industry. Due to the high viscosity and highly adhesive properties, filling and packaging this type of liquid can have challenges in terms of time and cost. A new control scheme based on motion synchronization with vision feedback is proposed in this article, which enables a new filling method for this type of liquid. The method can effectively resolve the challenges and enable an efficient filling process. Considering the time delay caused by image processing of the vision system, the Smith delay compensator is employed with an iterative learning control framework to improve the reliability of the feedback control under disturbances. The hardware implementation and experimental results demonstrate the convergence of tracking errors over the learning iterations, proving the effectiveness of the proposed control algorithm.