Control of Dynamic Performance in Step Motor Systems

106 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1983.The step motor is an electromagnetic incremental motion control actuator. Step motors of both linear and rotary types have been constructed. When energized by direct currents in a programmed manner the step motor indexes in fixed linear or angular increments. In general, the motor does not follow the drive command instantaneously.A lack of dynamic damping becomes apparent in a single-step response. The motor rapidly passes the one-step position, overshoots, and settles in an oscillatory manner. In addition to the lack of dynamic damping during final positioning, step motors frequently exhibit a dynamic instability at high stepping rates. This lack of dynamic performance has restricted the application of step motors to the low-performance segment of the incremental motion control market. It would be desirable to improve the dynamic performance of step motors without adding velocity or position measurement devices to the shaft of the motor. Such devices add significantly to system cost. It would be preferable to achieve improved dynamic performance by measuring and controlling only the electrical variables. The solution to this control problem is complicated by uncertainties in phase resistances and inductances, load inertias and viscous frictions.An analysis of the high-speed operation of the permanent-magnet step motor is given. Nonlinear state equations are linearized about a constant speed trajectory. The resulting periodically time-varying system is then transformed into a time-invariant system for which a characteristic equation can be found. Root loci for a resistance-limited sinusoidal-voltage drive are shown.Various means of improving the zero-speed damping of the step motor are considered. The simplest technique involves optimizing the stator resistance and steady-state phase current. Significant differences in the damping of permanent-magnet and variable-reluctance type step motors are found to exist. Other techniques involve compensation of the electrical system to promote natural damping and construction of a state-observer with state-feedback to place the eigenvalues. The performance of these techniques degrades when parameter variations take place. Lastly, a new technique has been developed which measures motor velocity at one motor phase and injects a damping current at a second motor phase. A single gain element is used to adjust the system damping and account for most of the parameter variations. The sensitivity to the remaining parameter variations can be made arbitrarily small.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD