5. EXPERIMENTAL RESULTS

This section describes the experiments that were used to refine and validate the algorithms presented in Chapter 4. In Section 5.1, the details of the experiment hardware and procedures are described, along with an overview of the facilities and required support. Section 5.2 describes the tuning and initial testing of the PD controller and evaluates its step response performance. Section 5.3 describes the initial testing of the quaternion based PD plus nonlinear compensation controller, and the methods employed to compensate for the physical limitations and unmodeled dynamics inherent to the hardware. The nonlinear dynamics in the controller uses the same values for the physical parameters that are used in the vehicle dynamic simulation. Single axis tracking results are then examined. Section 0 describes the performance of the automatic balancing algorithm. Finally in Section 5.5 the general capability and limitations of the adaptation algorithm and illustrated along with methods used to improve its performance. Then the adaptive controller follows a tumbling sinusoidal trajectory starting with no knowledge of the system parameters. As the trajectory is followed, the controller builds and continuously refines its estimate of the system parameters. This data set gives an example of the level of improvement that the feedforward nonlinear dynamics can provide when used in conjunction with the quaternion based PD controller. Finally, Section 5.6 compares the performance of each of the controllers.