Chapters authored
Kinematics: On Direction Cosine Matrices
Motion mechanics (dynamics) comprises kinetics to describe the implications of applied forces and torques; and also kinematics (phoronomics). Developed in the 1700s, kinematics describes mathematical translations from one basis of measurement to another using common kinematic measurement variables like quaternions, Euler angles, and direction cosine matrices. Two ubiquitous rotation sequences are unquestionably adopted for developing modern direction cosine matrices from among the 12 potential options, stemming from applicability to aerospace systems, accuracy, and computation burden. This chapter provides a comprehensive reevaluation of all 12 options yielding a menu of options for accuracy and computational burdens, with the results illustrated compared to the ubiquitous two modernly adopted choices, broken into two rotational groups: symmetric rotations and nonsymmetric rotations. Validation will be provided by critical analysis of integration using step size to illustrate correlated minimal accuracy. No single rotational sequence is universally superior with respect to all figures of merit, enabling trade-space analysis between rotational sequences. One interesting revelation of one of the two ubiquitous sequences (the 3-1-3 symmetric sequence) is illustrated to have relatively less accuracy but lower computational burden than the other (the 3-2-1 nonsymmetric sequence). Meanwhile, a relatively unknown “2-3-1” rotational sequence is shown to have similar computational burden and accuracy.
Part of the book: Advances in Spacecraft Attitude Control
Modern Control System Learning
This manuscript will explore and analyze the effects of different controllers in an overall spacecraft’s attitude determination and control system (ADCS). The experimental setup will include comparing an ideal nonlinear feedforward controller, a feedback controller, and a combined ideal nonlinear feedforward + feedback controller within a Simulink simulation. A custom proportional, derivative, integral controller was implemented in the feedback control, adding an additional term to account for the nonlinear coupled motion. Consistent proportional, derivative, and integral gains were used throughout the duration of the experiment. The simulated results will show that the ideal nonlinear feedforward controller lacked an error correction mechanism and took extra time to execute, the feedback controller faired only slightly better, but the combined ideal nonlinear feedforward controller with feedback correction yielded the highest accuracy with the lowest execution time. This highlights the potential effectiveness for a learning control system.
Part of the book: Deterministic Artificial Intelligence