| US 7,587,442 B2 | ||
| Method of determining the derivative of an input signal | ||
| William R. Cawthorne, Milford, Mich. (US); and Jy-Jen F. Sah, West Bloomfield, Mich. (US) | ||
| Assigned to GM Global Technology Operations, Inc., Detroit, Mich. (US) | ||
| Filed on May 14, 2004, as Appl. No. 10/846,018. | ||
| Prior Publication US 2005/0256919 A1, Nov. 17, 2005 | ||
| Int. Cl. G06F 7/00 (2006.01) | ||
| U.S. Cl. 708—443 [327/335] | 4 Claims |
| 1. A method of determining a phase-corrected derivative of a noise-containing input signal, comprising the step of:
periodically sampling the input signal to form a data stream of successive input signal values;
determining a normal derivative of the input signal based on a rate of change with respect to time between successive input
signal values separated by a time interval Δt, and updating said normal derivative at the time interval of Δt to form a first
data stream of normal derivative values;
determining an aliased derivative of the input signal based on a rate of change with respect to time between successive input
signal values separated by an increased interval of an integer R times the time interval Δt, and updating said aliased derivative
at the increased interval to form a second data stream of aliased derivative values;
low-pass filtering said first data stream of normal derivative values to form a derivative output; and
periodically resetting said derivative output to a reset value based in part on said second data stream of aliased derivative
values, wherein the resetting removes a phase delay in the first data stream introduced by the low-pass filtering to create
the phase-corrected derivative.
|