Pulse test of digital control system

Claim 1

In a process in which a process variable is manipulated by a control means in response to a measurement of a process condition in accordance with a mathematical model in the time domain X(t) which is representative of the expression Tau 1 Tau 2 c(t) + ( Tau 1 + Tau 2) c(t) + c(t) KX(t- Tau d) where t is time, c is said measurement, c is the first derivative of c with respect to time, c is the second derivative of c with respect to time, Tau 1 and Tau 2 are time constants, Tau d is dead time, and K is a constant; the method of obtaining data which can be employed to revise the mathematical model to compensate for changes in the process, which method comprises: removing the control means from control of the process; applying a pulse of preselected amplitude to said process variable to cause a change in said process condition; measuring said process condition and establishing a first signal representative thereof; establishing a second signal which is representative of the first derivative of said first signal with respect to time; removing said pulse from said process variable when said second signal reaches a maximum value; and establishing a third signal which is representative of the second derivative of said first signal with respect to time, said established signals providing information which can be employed to update said mathematical model.