Data processing: generic control systems or specific application – Generic control system – apparatus or process – Optimization or adaptive control
Reexamination Certificate
1998-09-09
2001-07-31
Grant, William (Department: 2121)
Data processing: generic control systems or specific application
Generic control system, apparatus or process
Optimization or adaptive control
C709S241000
Reexamination Certificate
active
06269274
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention is directed, in general, to computing and processing systems and, more specifically, to a system and method for balancing and distributing control algorithm processing load and a real time plant control system that employs the system or the method.
BACKGROUND OF THE INVENTION
Automated process control systems include a comprehensive set of control algorithms, or software-definable process control routines, to control and monitor various processes within, for instance, a manufacturing facility. The control systems can be tailored to satisfy a wide range of process requirements globally or within specified portions of the facility. Conventionally, the control systems include a variety of modules, each having its own processor or firmware, linked together by communication buses to result in a distributed process control system. The distributed nature of the system affords high performance with the capability to expand the system incrementally to satisfy growth or modifications in the facility.
To allow a user to tailor a process control system to a particular processing environment, it is important to provide such systems with highly-configurable and adaptable control systems. Process control systems generally provide a means by which users can create custom control strategies, e.g., software-definable process control routines. In object-oriented programming environments, a complete control strategy may be built from smaller components, such as “blocks” and “parameters.” A block is a software construct, or “object,” used to encapsulate the data and the algorithms of elemental control computations; parameters define the interface to individual pieces of data within a block.
Digital process controllers (DPCs) are generally required to meet a series of requirements on the manner in which they process control algorithms. For example, they must execute the control algorithms in a periodic fashion. The control algorithms must be executed at a sample rate, where the sample rate “jitter” is sufficiently low compared to the tolerance of process dynamics. DPCs may be required to execute additional algorithms whose functions are essential for the overall process control mission but are not periodic and predictable, e.g., communication with supervisory display and control stations, communication with peer controllers, event distribution and housekeeping tasks. DPCs must execute non-periodic algorithms in a way which does not interfere with deterministic execution of periodic control algorithms as long as DPCs are configured within a defined operating range.
To meet these above-described requirements, DPCs are required to balance the control algorithm processing load and distribute it over the time interval of the DPCs' fundamental repetition rate. Conventional DPCs, such as Honeywell Inc. TDC3000 controllers known as “Process Manager,” “Advanced Process Manager” and “High Performance Process Manager,” pre-qualify a configuration and accept or reject the configuration on the basis of predicted potential for overload. These DPCs, however, do not support manual balancing, balancing a subset of algorithms while other algorithms maintain control, user measurement of execution time for a particular algorithm or associating execution time with an algorithm as a characteristic attribute.
On the other hand, other conventional DPCs, such as Honeywell's “Alcont Process Manager,” provide only for manual balancing of the algorithm processing load. These DPCs, however, do not support qualification of an algorithm load on the basis of a model computation or rejection of a load which is too high. Still other DPCs, such as Honeywell's “TPS Networks Safety Manager,” provide for computing execution time from characteristics of a particular algorithm configuration, but do not support balancing of subset algorithms while other algorithms maintain control.
Therefore, what is needed in the art is an improved method for managing algorithm processing load within a process controller that incorporates all the above-described features into one controller. More particularly, what is needed in the art is a method that provides flexible management of algorithm processing load.
SUMMARY OF THE INVENTION
To address the above-discussed deficiencies of the prior art, it is a primary object of the present invention to provide a far more flexible and powerful way to balance and distribute control algorithm processing loads.
In the attainment of the above primary object, the present invention provides, for use in a digital process controller (DPC), a system for, and method of, balancing control algorithm processing load and distributing the load over a given time interval and a real time plant control system employing the system or the method. In one embodiment, the system includes: (1) a model for estimating execution time of at least one control algorithm based on the configuration data applied to that algorithm, the DPC capable of balancing and distributing said load based on the estimate generated by the model and (2) a data entry process, associated with the algorithm model, that allows an empirically-determined execution time to be associated with at least one control algorithm, the DPC capable of balancing and distributing the load based on the time in lieu of the model based estimate and balancing and distributing the load upon configuration of at least one control algorithm.
The present invention therefore introduces a model-based control algorithm balancing and distribution system in which data pertaining to each control algorithm to be balanced are employed to derive an estimated execution time, but to which the user may optionally supply overriding, empirically-determined (perhaps actually measured, perhaps formulaically analyzed, perhaps intuitively guessed) execution times.
The system can run the model at the time that particular control algorithms are configured to allow balancing and distributing to occur iteratively and interactively, if desired. Alternatively, in one embodiment of the present invention, the data entry process allows the user to balance and distribute the load under the user own control.
In one embodiment of the present invention, the model allows at least one control algorithm to be configured and the load to be balanced and distributed without interrupting operation of other control algorithms. Alternatively, the system can interrupt execution of the other control algorithms to pause or stop while balancing and distribution take place.
In one embodiment of the present invention, a selected one of the data or the time is replicated in the model for each instance of at least one control algorithm. Preferably, the user does not have to associate the data or the time with each occurrence of a given control algorithm. In some embodiments, however, the user can perform such manual association.
In one embodiment of the present invention, the configuration data includes an algorithm type, an algorithm execution time, an algorithm block count and a data flow connection count. Those skilled in the art are aware that other configuration data types may also be advantageously employed with the present invention.
In one embodiment of the present invention, at least one control algorithm forms a portion of a control module having configuration parameters selected from the group consisting of: (1) period, (2) estimated weight and (3) phase. The model employs the configuration parameters to balance and distribute processing load. Of course, additional or other parameters may be desirable in a given application.
In one embodiment of the present invention, the model is capable of providing an actual execution time of at least one control algorithm to a user. This allows the user to evaluate at least one control algorithm's performance and make decisions based thereon.
The foregoing has outlined rather broadly the features and technical advantages of the present invention so that those skilled in the art may better understand the detailed
Himmer Richard P.
Radomy Irina
Steinman Jethro F.
Grant William
Hitt Chwang & Gaines
Honeywell International Inc
Rapp Chad
LandOfFree
System and method for balancing and distributing 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 System and method for balancing and distributing control..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for balancing and distributing control... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2521319