Design of a disturbance observer applied to a SCARA robotic arm

Many applications in motion control require continuous improvement in basic performance (higher accuracy and higher acceleration) without the higher cost generally associated with such improvements. This is especially difficult in the presence of internal and external disturbances. The addition of a Disturbance Observer (DOB) to a typical motion control feedback loop can significantly and economically improve system performance without requiring any additional sensors. Moreover, the additional computational burden is very light. The DOB algorithms presented in this thesis are developed in continuous time and converted to discrete time using a trapezoidal algorithm with prewarping at the gain crossover frequency. Four DOB algorithms are implemented on the elbow link of an existing Selectively Compliant Assembly Robotic Arm (SCARA). Experimental results demonstrate the effect of DOB order on performance using second, third, and fourth order DOB algorithms.