| Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 102 === This thesis presents the trajectory tracking approach for a wall-climbing robot by using adaptive control schemes. The robot platform is configured with 2 active wheel motors, a suction motor, a motion controller, a battery set, and sensors. The platform's diameter is 25 cm, and its weight is 1.5 kg. The primary sensors are encoders and an accelerometer sensor, the sensors are used for measuring the robot's spatial orientation. The most important consideration for controlling the wall-climbing robot is to make sure that the wheels can be always well contacted to the wall regardless of the slope conditions without sacrificing robot’s mobility. To consider different slope conditions of the wall, this thesis proposes an adaptive gain sliding control schemes to alter the vacuum force so that different gravity effects can be properly dealt with. Moreover, the proposed vacuum force control approach can be desired to avoid the slips of wheels, as well as to reduce the power consumptions of wheel motors and suction motor. Practically, encoder and accelerometer sensors provide the spatial posture information for realizing adaptive control schemes. The sensors are connected to a field-programmable gate array (FPGA) based onboard motion controller to generate control inputs for wheel motors and suction motor according to a specific trajectory. Finally, MATLAB simulations and real tests for dealing with different surface slope conditions were performed with the trajectories of circle, triangle and rectangle. The results were evaluated according to the measurement of the accuracy of trajectory and the power consumptions of the wall climbing robots.
|
|---|
