Data processing: generic control systems or specific application – Generic control system – apparatus or process – Optimization or adaptive control
Reexamination Certificate
1999-12-30
2002-07-23
Black, Thomas (Department: 2121)
Data processing: generic control systems or specific application
Generic control system, apparatus or process
Optimization or adaptive control
C700S008000, C700S033000, C318S610000
Reexamination Certificate
active
06424873
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention is directed, in general, to process control systems and, more specifically, to a process control system containing Proportional, Integral, Derivative (“PID”) controllers.
BACKGROUND OF THE INVENTION
Many process facilities (e.g., a manufacturing plant, a mineral or crude oil refinery, etc.) are managed using distributed control systems. Typical contemporary control systems include numerous modules tailored to monitor and/or control various processes of the facility. Conventional means link these modules together to produce the distributed nature of the control system. This affords increased performance and a capability to expand or reduce the control system to satisfy changing facility needs.
Industrial control systems often employ feedback controllers for controlling the operation of one or more operating units of the system such as a heater, a pump, a motor, a valve, or a similar item of equipment. In a feedback controller a command is sent to the feedback controller that represents a desired value or setpoint (“SP”) for a process variable (e.g., a desired pressure, a desired temperature, a desired flow rate). A feedback signal is also sent to the feedback controller that indicates the actual value of the process variable (“PV”) (e.g., the actual pressure, the actual temperature, the actual rate of flow). An error signal is calculated utilizing the difference between the setpoint (“SP”) command and the feedback signal that indicates the actual value of the process variable.
From the error signal, the feedback controller calculates a change command to change the current setting of the operational unit. For example, if the operational unit is a motor, the change command would cause the speed of the motor to change (either increase or decrease) in order to cause the actual value of the process variable to more closely approach the desired setpoint value for the process variable.
In a simple feedback controller, the change command is proportional to the error signal. In more complex feedback controllers, the change command may be a more complex function of the error signal. The relationship between the error signal and the change command greatly affects the characteristics of the control system. These characteristics include (a) the “response time” of the system (i.e., how fast the operational unit responds to the new change command); (b) the “overshoot” of the system (i.e., how much the operational unit initially exceeds its new setting); and (c) the “damping ratio” of the system (i.e., how long the output values of the operational unit oscillate before eventually stabilizing at the new setting).
Industrial control systems often employ a type of feedback controller known as a Proportional, Integral, Derivative (“PID”) controller. PID controllers are capable of calculating a variety of functional relationships between an error signal and a change command signal in a feedback control system.
A PID controller may be used to calculate a functional relationship between an error signal and a change command signal that minimizes the time that the control system takes to reach a stable state following a change command signal. PID controllers are capable of operating in three modes. The modes are the Proportional mode, the Integral mode, and the Differential mode. PID controllers generate a proportional-integral-differential function that is the sum of (a) the error signal times a proportional gain factor (“P gain”), and (b) the integral of the error signal times an integral gain factor (“I gain”), and (c) the derivative of the error signal times the derivative gain factor (“D gain”). An appropriate selection of the three gain factors (“P”, “I” and “D”) must be made to calculate a transfer function that will result in a desirable system response. Selecting the three gain factors is sometimes referred to as “tuning” the PID controller.
In a PID controller the integral mode will continue to integrate the error as long as the error is not zero. This can cause the output of the PID controller to increase well beyond the acceptable output limits of the PID controller. When this occurs, the PID controller is said to be “wound up” or is said to be in a “wind up” state. A “wound up” PID controller can no longer affect the value of the process variable because the output of the PID controller is outside the operating range of the operational unit. For example, a valve may be fully open but the “wound up” PID controller is asking for the valve to be five hundred percent (500%) open. For an additional example, a motor may be operating at is maximum speed of five hundred revolutions per minute (500 RPM) but the “wound up” PID controller is asking for the motor to run at three thousand revolutions per minute (3,000 RPM).
When the sign of the error changes, the PID controller must “unwind” (i.e., cease causing an excessive output signal) before the output of the PID controller returns into the proper operating range. The process of “unwinding” may result in “overshoots” in the value of the process variable or may result in significant oscillations in the value of the process variable.
To prevent a PID processor from entering the “wound up” state it is possible to limit the contribution of the integral value when it is determined that the integral value contribution would cause the output signal to increase in the direction that will cause violation of the output limits. Implementing integral value limits in a PID controller is relatively simple because the upper and lower output limits are known, and the PID controller is able to determine whether the sum of the proportional value contribution (the “P contribution”) and the derivative value contribution (the “D contribution”) violates the output limits. If the sum of the P and D contributions do not violate the output limits, then a portion of (or all of) the integral value contribution (the “I contribution”) may be included in the output signal up the level of the output limit. As will now be explained, this method is not sufficient in cases involving two coupled PID controllers.
Two PID controllers may be coupled to operate in a cascade structure. In such an arrangement, the primary PID controller sends an output signal to an input of the secondary PID controller. The primary PID controller also receives a feedback signal from the secondary PID controller. The primary PID controller performs a PID calculation to determine the output signal that it transfers to the secondary PID controller. The secondary PID controller is capable of determining that the output signal of the primary PID controller has exceeded an output limit for output signals that the secondary PID controller will transfer.
The method of limiting the integral value contribution described above for the case of a single PID controller is not sufficient in the case of two coupled PID controllers because (1) the output limits in the secondary PID controller are not available to the primary PID controller, and (2) the secondary PID controller may have two different types of output limits. Specifically, the secondary PID controller may have either setpoint limits or output limits (or both types of limits). It is possible to transfer setpoint limits from the secondary PID controller to the primary PID controller as constant values. But it is not possible to transfer the output limits of the secondary PID controller as constant values. In general, when integral value calculations are involved, the PID calculation algorithm of the primary PID controller cannot determine the output limits of the secondary PID controller without complete knowledge of the past history of the input values.
One prior art method limits the integral value contribution in a primary PID controller (that is coupled to a secondary PID controller) by including or excluding the integral value contribution in response to information received from the secondary PID controller via limit flags. This prior art method causes the secondary PID controller to set an Integra
Barnes Crystal J
Black Thomas
Honeywell Inc.
Novakov Davis & Munck
LandOfFree
Systems and methods for limiting integral calculation... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Systems and methods for limiting integral calculation..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Systems and methods for limiting integral calculation... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2901148