Abstract
In this thesis, we design and construct a low-level servo control system for a biped walking robot. This system consists of a 32-bit master microcontroller (ARM9) and four 16-bit slave microcontrollers (dsPIC30F4012). Furthermore, we also construct a CAN-based（Controller Area Network-based） microcontroller network, composed of the master and slave microcontrollers. The main controller (ARM9) is responsible for the real-time transmission of the 12 joint-command trajectories to the 4 slave controllers, and the slave controllers (dsPIC30F4012) are responsible for the servo control of the joint actuators (DC motors) and for the sensor feedback signals transmission back to the master controller. In the low-level servo control of each joint axis, we use fuzzy PID controller. The PID parameters Kp, Ki, Td are tuned online by their fuzzy systems. Friction compensation is integrated in the fuzzy PID controller for improving the control performance. In order to compensate for the effects of elastic deformation of the timing belts, a simple command-shift method is used to reduce the tracking errors. Since the feedback signals from the analog position sensors may contain serious noises, in each servo system, we design a fifth-order Butterworth analog low-pass anti-aliasing filter and a 20th-order digital low-pass filter for the filtering of possible high-frequency noises to improve the quality of the feedback signal and thus the control performance.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
