Abstract:
Machine automation requires the robotic machine to be at least as productive as a manually operated machine. To increase robot productivity robot motion speed should be improved. A feasible approach to improving the motion speed is to minimize the motion time needed to perform a given task subject to actuator constraints. This work addresses the problem of optimal trajectory planning for heavy-duty hydraulic manipulators. These manipulators have the following characteristics: they are powered by a single engine mounted on the machine and they are under-powered even during normal operations resulting in dynamic power redistribution to the actuators. For the hydraulic manipulators, the actuator characteristics are very significant and complex due to high nonliniarities in the hydraulic system and power coupling between the actuators. The method developed in this thesis focuses on utilizing advantageously the actuator capabilities to minimize the time needed to move the manipulator end-effector along a specified path. To perform the search for the minimum motion time along the specified path, a downhill simplex technique is implemented. The method is applied to a Caterpillar 215B excavator-based log loader in a typical pick and place task. The main contributions of thi thesis are the incorporation of the complex actuator characteristics in the optimal trajectory planning and the implementation of an optimization algorithm (downhill simplex method), which shows effective results for solving the optimal trajectory planning problem.
