Intelligent Control of 3-Dimensional Autonomous Mobile Cranes

本文完成了桥式起重机系统、传感器、模拟器和控制器的硬件设计。本文的目的是提出一种建立自主移动的起重机系统的方法，该方法考虑到障碍物识别和离线路径规划，并抑制摆动的传输系统。该系统由三部分组成：视觉系统、路径规划系统和控制系统。在视觉系统中，使用一对CCD图像传感器和激光线标记器通过滑动光线投影方法生成环境地图。在规划系统中，使用所获得的环境地图来导出从起点到目标位置的最佳转移路径，诸如最小距离路径或最小能量成本路径或无摆动路径。该路径是通过使用自动生成算法，使用基于三维扩散方程的势方法生成的。在控制系统中，采用具有指定极点配置的变结构控制器，实现了对被控对象的位置控制和摇摆抑制。最后通过仿真和实验验证了该方法的有效性。为了完善自主式移动的起重机系统，在线障碍物识别与避障的研究是必不可少的，也是下一步需要解决的问题。

In the present research, the hardware of overhead traveling crane systems, sensors, simulators and controllers parts has been completed. The purpose of this paper is to present a method for establishing an autonomous mobile crane system which takes into account obstacle recognition and path planning by off-line, and suppression of swinging for the transfer system. The system has three different parts : a visual system, a path planning system and a control system. In the visual system, an environment map is generated by the Slid-Ray Projection method using a pair of CCD image sensors and a laser-line marker. In he planning system, the obtained environment map is used to derive an optimal transfer path such as a minimum distance path or a minimum energy cost path or swinging-free path from the start to the goal position. The path is generated through the use of auto-generation algorithm using a potential method based on the three dimensional diffusion equations. In the control system, the variable-controller with the specified-pole assignments is used to achieve position control and the suppression of sway for the transfer objects. Finally, the validity of the proposed method is confirmed by simulatoions and experiments. The studies of on-line obstacle recognition and avoidance are now indispensable in order to complete a perfect autonomous mobile crane systems, and therefore it will be a next problem to be solved very soon.