Palletizing Simulator Using Optimized Pattern and Trajectory Generation Algorithm

Abstract

Collision avoidance and robot path planning problems have emerged as a potential domain of robotics research of late because of their indispensable requirements in the field of manufacturing vis-a-vis material handling, such as picking and placing an object and loading/unloading a component to/from a machine or storage bins. This chapter focuses on palletization, a form of unitization in which a uniform load is stacked on a wooden pallet using a predetermined case pattern sequence and a given number of layers. In many kinds of proposed C-space construction approaches, several algorithms deal with the boundary of the C-obstacle analytically. Lozano Perez proposed the fundamentals of the C-space approach. When both the robot and obstacles have the shape of convex polygons, the Cobstacle boundary for an n-DOF manipulator is approximated by sets of n-1 dimensional slice planes, which are made from a one-dimensional slice plane. C. Zhao and his colleague proposed an algorithm to describe the C-obstacle as a set of parametric equations formulated from the mapping of the boundaries of the obstacles in a workspace. They use inverse pseudo kinematics to convert the obstacles in a workspace into a C-space. Debanik Roy studied path planning algorithms and their heuristics using the concept of visibility graph, and he presented an overview of the case study of robot path planning in an industrial environment in real time. Xiaojun and his colleague proposed a two-phase approach for C-obstacle construction and the collision detection of manipulators. This method is applicable to manipulators with various types of kinematic structures and geometric shapes. M. Pettersson and his coworkers proposed trajectory optimization method considering fatigue and thermal load of real robot. They also referred that the proposed method could be directly adapted to palletizing system. The issues of these papers, however, are for the operation of real industrial robot, and it is a part of entire palletizing system. Many other latest researches solely oriented towards path planning or modified apparatuses to improve the specified handling task have also been conducted. Studies on the total robot palletizing system, however, which integrates loading pattern optimization, robot OLP simulation, and path optimization, have yet to be systematically conducted. This study was dedicated to the development of OLP Simulation S/W for a robot palletizing system (Non-vision system), which means that this study prioritized the reflection of the 17