Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Mechanical measurement system
Reexamination Certificate
2000-06-16
2003-07-15
Assouad, Patrick (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Mechanical measurement system
C700S032000, C702S179000
Reexamination Certificate
active
06594593
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS
Patents: Performance assessment of model predictive controllers, Ser. No. 09/1,340,531 (pending)
Performance assessment of non-deadtime compensated controllers, Ser. No. 09/501,129 (pending) Both these patent applications by Alan Hugo are for performance assessment of nonintegrating processes.
STATEMENT REGARDING FEDERALLY SPONSORED REASEARH AND DEVELOPMENT
N/A
REFERENCE TO MICROFICHE APPENDIX
N/A
BACKGROUND—FIELD OF INVENTION
This invention relates to a technique to accurately assess the performance of controllers applied to integrating processes. A common example of an integrating process is the level of a liquid surge vessel. Specifically, the technique compares the performance of the current control to what would be stained if an optimal controller for integrating processes was applied to the process.
BACKGROUND—DISCUSSION OF PRIOR ART
An integrating process is one where any imbalance causes the system to continue changing until some constraint is reached. The most common example of this is a liquid surge vessel as shown in FIG.
1
. This is called an integrating process because any discrepancy between the inlet and outlet flow will cause the liquid level in the vessel to rise or fall until the vessel if either full or empty (the process integrates the discrepancy over time). The purpose of this surge vessel is to dampen out the variation in the inlet flow by allowing the amount of liquid in the vessel to vary between upper and lower limits. A significant fraction of industrial processes are integrating, and controllers must be applied to these processes to keep them within constraints.
A major difference between the objectives of a controller on a non-integrating process and a controller on an integrating process is that the objective of the first controller is usually to maintain some measured value as close to a desired value at all times. In U.S. Pat. No. 5,838,561, a method is described for evaluating controllers where the purpose of the controller is to keep a process measurement at a desired value. In contrast, often the purpose of a controller on an integrating process is to use the capacity of the unit to dampen disturbances. For instance, the purpose of the surge vessel level controller in
FIG. 1
is to let the level in the vessel rise and fall to attenuate short-term variations in the inlet flow, thus maintaining the outlet flowrate as smooth as possible. In more formal terms, the objective of the surge vessel level controller is to minimize the maximum rate of change of the outlet flow, subject to the level being maintained between specified upper and lower bounds.
This patent is concerned with measuring the effectiveness of controllers applied to integrating processes. There are techniques currently available for measuring controller performance on nonintegrating processes (Harris & Desborough, 1992; Harris, et al., 1999), but these are not applicable for measuring the effectiveness of controllers on integrating processes, as the objectives of these two types of controllers are very different. There are no patents or prior works addressing the measurement of performance for controllers applied to integrating processes.
Harris and Desborough (
Performance Assessment for Univariate Feedback Controllers,
Cdn. J. Chem. Eng., non-integrating process to a theoretical Minimum Variance controller (i.e., the best physically realizable linear controller for non-integrating processes), and can be calculated from routine operating data. This measure has become a standard in industry, and is termed the Harris Index. However, the objective of the Minimum Variance Controller (to keep the measurement (PV) as close as possible to a desired value (SP)) is wrong for this case, and the Harris Index does not provide the proper measure of controller performance for integrating processes.
The simplest and most direct technique of assessing controller performance is to calculate the variance between the measurement (PV) and the setpoint (SP), i.e.,
σ
2
=
1
n
∑
i
=
1
n
(
PV
-
SP
)
2
(
1
)
For surge vessels, the variance may be calculated for both the level and outlet flow. However, this measure is not much of practical use, as there is little guidance as to what are desirable variances. The variances are, moreover, almost totally dependent on the inlet flow variance which is in general not constant over different time periods.
The correct performance measure is a comparison between the actual outlet flow variance, and the outlet flow variance that would occur if some “optimal” controller were applied to the system. Mathematically, this is expressed similarly to the Harris Index as:
η
=
1
-
σ
opt
2
σ
act
2
(
2
)
The term &sgr;hd act
2
is the actual flow variance, while the optimum variance &sgr;
opt
2
is, for the purposes of this patent, the flow variance that would result if the optimal level control algorithm of McDonald et al. (
Optimal Averaging Level Control,
AIChE J., 1986), or similar variants, such as proposed by Wang and Cluett (
Tuning PID controllers for integrating processes,
IEE Proc.-Control Theory Appl, 1997) were applied to the system.
BRIEF SUMMARY OF THE INVENTION
Of concern in the present invention is to determine the performance of a controller on an integrating process using normal closed-loop operating data. Also of concern in the present invention is for the performance measure to reflect the specific objectives of controllers for integrating processes—i.e., the objectives of minimizing the rate of change of the manipulated variable (outlet flow in the above example) while keeping the controlled variable (level) between specified limits.
It is, therefore, a feature of the present invention to provide a method to determine the performance of a controller for integrating processes.
It is, also, a feature of the present invention that it only requires normal operating data to determine the controller performance.
Yet another feature of the present invention is that the performance measure adequately reflects the objectives of controllers for integrating processes.
It is also a feature of the present invention that it has the same range and interpretation as the Harris Controller Performance Index.
Additional features and advantages of the invention will be set forth in part inthe description which follows, and will in part be apparent from the description, or may be learned from practice of the invention.
The features and advantages of the invention may be realized by means of the combinations and steps pointed out in the appended claims.
REFERENCES:
patent: 5796606 (1998-08-01), Spring
patent: 5838561 (1998-11-01), Owen
patent: 6459939 (2002-10-01), Hugo
Hugo, “Performance Assessment of DMC Controllers”, IEEE, Jun. 1999.*
Swanda et al., “Controller Performance Assessment Based on Setpoint Response Data”, IEEE, Jun. 1999.*
Panagopoulos et al., “Design of PiD Controllers Based on Constrained Optimization”, IEEE, Jun. 1999.*
Hugo, “Process Controller Performance Monitoring and Assessment”, Internet Web page of Control Arts, Inc., date unknown.*
Hugo, “Limitations of Model Predictive Controllers”, Internet Web Page of Control Arts, Inc., date unknown.*
Desborough & Harris Cdn J. Chem Eng Performance Assessment for Univariate Feedback Controllers 70, 1992.
McDonald, McAvog, & Tits Optimal Averaging Level Control AIChEJ, 32, 1986.
Wang & Cluett IEE Proc—Control Tuning PID controllers for Integrating Processes Theory Appl , 144, No. 5, 1997.
Harris, Seppala, Desborough A review of performance monitoring & Assessment techniques for univariate & Multivariate control systems J. Proc Control 9 1999, pl-17.
U.S. patent application Ser. No. 09340531, Performance Assessment of Model Predictive Controllers, filed Jun. 29, 1999.
Desborough & Harris Performance Assessment for Univariate Feedback Controllers Cdn J. Chem Eng 70, 1992, p 1186-1197.
Harris, Seppala, Desborough A review of performance monitoring & assessment techniques for univariate & multivariate control system J. of Pro
Alan J. Hugo
Assouad Patrick
LandOfFree
Performance assessment of controllers applied to integrating... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Performance assessment of controllers applied to integrating..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Performance assessment of controllers applied to integrating... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3102525