Data processing: measuring – calibrating – or testing – Measurement system – Statistical measurement
Reexamination Certificate
1999-08-09
2002-10-15
Bui, Bryan (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system
Statistical measurement
C702S105000, C702S041000, C700S282000, C073S168000
Reexamination Certificate
active
06466893
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to process control networks and, more particularly, to a method of and an apparatus for statistically determining an estimate of a process control loop parameter while the process control loop is connected on-line within a process environment.
BACKGROUND ART
Large scale commercial manufacturing and refining processes typically use a process controller to control the operation of one or more process control devices such as valve mechanisms. In turn, each process control device controls one or more process variables comprising, for example, fluid flow, temperature, pressure, etc., within a process. Generally, process control devices such as valve positioners have an actuator controlled by a positioner that moves an associated control element, such as a valve plug, a damper, or some other alterable opening member, in response to a command signal generated by the process controller. The control element of a process control device may, for example, move in response to changing fluid pressure on a spring biased diaphragm or in response to the rotation of a shaft, each of which may be caused by a change in the command signal. In one standard valve mechanism, a command signal with a magnitude varying in the range of 4 to 20 mA (milliamperes) causes a positioner to alter the amount of fluid and thus, the fluid pressure, within a pressure chamber in proportion to the magnitude of the command signal. Changing fluid pressure in the pressure chamber causes a diaphragm to move against a bias spring which, in turn, causes movement of a valve plug coupled to the diaphragm.
Process control devices usually develop or produce a feedback signal, indicative of the response of the device to the command signal, and provide this feedback signal (or response indication) to the process controller for use in controlling a process. For example, valve mechanisms typically produce a feedback signal indicative of the position (e.g., travel) of a valve plug, the pressure within a fluid chamber of the valve or the value of some other phenomena related to the actual position of the valve plug.
While a process controller generally uses these feedback signals, along with other signals, as inputs to a highly tuned, centralized control algorithm that effects overall control of a process, it has been discovered that poor control loop performance may still be caused by poor operating conditions of the individual control devices connected within the control loop. In many cases, problems associated with one or more of the individual process control devices cannot be tuned out of the control loop by the process controller and, as a result, the poorly performing control loops are placed in manual or are detuned to the point where they are effectively in manual. The processes associated with these control loops require constant supervision by one or more experienced human operators, which is undesirable.
Poor control loop performance can usually be overcome by monitoring the operational condition or the “health” of each of the process control devices connected within the loop, or at least the most critical process control devices connected within the loop, and repairing or replacing the poorly performing process control devices. The health of a process control device can be determined by measuring one or more parameters associated with the process control device and determining if the one or more parameters is outside of an acceptable range.
One process control device or loop parameter that may be used to determine, and that is indicative of the health of a process control device such as a valve is a friction measurement. In general, the friction measurement relates to the amount of force or pressure that must be applied to a moving part of the device to overcome the force of friction, e.g., to begin initial movement of a valve plug. If the friction measurement of a device becomes greater than a set amount, it may mean that the valve plug is sticking for some reason and that, therefore, the device may need repair or replacement.
Another process control device parameter that may be used to determine, and that is indicative of the health of a process control device is dead band. Generally speaking, in process instrumentation, dead band is the range through which an input signal may be varied, upon reversal of direction, without initiating an observable change in an output signal. Dead band, which may be caused by the physical play between mechanically interconnected components, friction, and/or other known physical phenomena, is best observed when a command signal causes a reversal in the direction of movement of a moveable element of a process control device. During this reversal, the command signal undergoes a discrete amount of change (dead band) before the movable element of the process control device actually exhibits movement in the new direction. Put another way, the difference between the value of the command signal (or other control signal) at which movement of the process control device element in a first direction last occurred and the value of the command signal (or other control signal) at which the movement of the process control device element first occurs in a second and different direction is a measure of the dead band of the process control device.
Still other device parameters that may be used to determine the health of a process control device or loop are dead time, response time, oscillation, and shaft windup. Dead time is associated with, and may be considered to be a measurement of the amount of time it takes the process control device to actually begin moving a moveable element in response to a change in a control signal. Response time is the amount of time it takes the moveable element of a process control device to reach a certain percentage, for example 63 percent, of its final value in response to a change in a control signal. Oscillation is a measurement that determines if a signal associated with a process control device or loop is periodic and, therefore, has some oscillatory behavior, which is typically undesirable. Shaft windup is the actuator travel that occurs (typically in a rotary valve) before the applied pressure reaches an amount that overcomes the friction inherent in the mechanism.
If the friction, dead band, dead time, or some other process control loop parameter increases a significant amount over its calibrated value, it may be necessary to repair or replace a process control device to establish adequate control within the process control loop. However, it is not usually very easy to measure process control loop parameters, such as friction, dead band and dead time to monitor the health of process control devices when those devices are connected on-line within a control loop.
In the past, operators have had to remove a process control device from a control loop to bench test the device or, alternatively, control loops have been provided with bypass valves and redundant process control devices to make it possible to bypass a particular process control device to thereby test that device while the process is operating. Alternatively, operators have had to wait until a process is halted or is undergoing a scheduled shut-down to test the individual process control devices within the process. Each of these options is time consuming, expensive, and still only provides intermittent measurement of the parameters of the individual process control devices required to determine the health of those control devices.
There have been some attempts to collect data from a process control device on-line and to obtain an indication of characteristics of a device therefrom. For example, U.S. Pat. No. 5,687,098 to Grumstrup et al. discloses a system that collects device data and constructs and displays the response characteristic of the device. Likewise, application Ser. No. 08/939,364 filed Sep. 29, 1997 entitled “Method of and Apparatus for Nonobtrusively Obtaining On-Line Measurements of a Process Control Device Parameter,” upon which this a
Junk Kenneth W.
Latwesen Annette L.
Bui Bryan
Fisher Controls International Inc.
Marshall Gerstein & Borun.
Vo Hien
LandOfFree
Statistical determination of estimates of process control... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Statistical determination of estimates of process control..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Statistical determination of estimates of process control... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2980060