| Summary: | 博士 === 國立高雄第一科技大學 === 系統資訊與控制研究所 === 98 === This thesis discusses the cleaning simulation of a domestic cleaning robot in a virtual environment. The cleaning path is automatically generated from the cleaning strategy that controls the interacting response of the robot with a simulated domestic environment. Collision, drop and wall-following sensors are established in the robot to detect obstacles. The coverage and cleaning efficiency for the cleaning strategy can be analyzed in the simulation system. Typical cleaning modes include random bouncing, spiral, wall following, and zigzag movements. This research applies Automatic Dynamic Analysis of Mechanical System, ADAMS, to construct a virtual prototyping environement that combines sensors, simulation scripts and macros to determine the response for different obstacles. The cleaning path for a specified cleaning strategy can be derived and transformed to the coverage rate and the cleaning efficiency through a analysis program written in Visual C#. The proposed methodology adopts a soft prototyping to evaluate the cleaning strategies of a cleaning robot that can greatly reduce experimental costs and increase development efficiency.
|
|---|
