OFF-LINE PROGRAMMING FOR MODEL BASED SENSOR CONTROLLED ROBOTS (TASK PLAN, WORLD MODEL, OBJECT FEATURES)
Abstract
This thesis develops a consistent method for combining an arbitrary number of sensor observations in real time. The method includes a planning phase where each sensor system examines the task plan and the world model and then determines the features it expects to detect during task execution. It is not necessary to know the number or the type of sensors available to the robot manipulator at the time the task plan is developed. We also developed a method to correct the manipulator's trajectory, in real time, to account for changes in the target object's location.
Degree
Ph.D.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.