Summary: | 碩士 === 國立中正大學 === 電機工程研究所 === 100 === In recent year, more and more robot competitions attract many countries to join, whether the robot is humanoid robot or wheeled robot, and there are many types of competition, such as football competitions, cross-country competitions, and the intelligent service robot competitions. In this thesis, we use a wheeled robot. The main content of my thesis is that we attempted to create the software modules for a service robot, and the module can be successfully implemented. There are two main modules:
The first one is Simultaneous localization and mapping (SLAM) module. The SLAM module requires input from several sensors. The first one is the laser range finder (laser scanner) to get the range sensor data of the environment, the second one is encoder data from the robot’s differential wheels which will be treat as the reference data to generate fixed amount of particles for particle filters. Particle filters are used to match the observation with the current map, and to find the exact location in the map, which will fix accumulate error from encoder. The third one is joystick command, which will control many functions, such as draw map on the screen, save map in computer, start to use the particle filter.
The second one is navigation, when giving the robot a target position, the robot can autonomously go to the target position. While the robot is moving, it will get the range sensor data from laser range finder and the encoder data from the robot, which will feed as the input data of navigation module. We use the artificial potential field method to calculate the current desired moving direction of robot. This method can also be used to avoid obstacles while moving when the robot’s path is occupied by obstacles, such as human, chairs or tables etc... .
|