Data processing: measuring – calibrating – or testing – Measurement system – Performance or efficiency evaluation
Reexamination Certificate
2001-02-14
2004-06-15
Assouad, Patrick J (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Performance or efficiency evaluation
C700S051000, C700S052000, C702S179000
Reexamination Certificate
active
06751575
ABSTRACT:
BACKGROUND
The invention relates to a system and method for monitoring, controlling, and diagnosing process machines. Particularly, the invention relates to a system of data mining and comparison, using a similarity search engine to identify the degree of similarity between process-related objects and a process control machine.
The process and factory automation industry is well-known for its control of processes that are common among such facilities as plants, refineries, power generation, water and waste management, food and beverage production, etc. Control of these processes may comprise a multiplicity of process control loops that control process pressures, flows, temperatures, levels, alkalinity and acidity, composition, packaging, positions, and the like. A control system may comprise a single-loop system, multi-loop system, Digital Control System (DCS), Programmable Logic Controller (PLC), Supervisory Control and Data Acquisition (SCADA) system, or even a simple data acquisition system used for strictly monitoring a process.
Accurate control of a process is predicated on obtaining accurate information from a sensor or transmitter. These devices may fail or drift, resulting in false or inaccurate readings being posted to the controller. In these cases, the controller is unable to issue control actions to the control device to achieve the desired set point. Thus, there is a need for a control system that can determine whether false or inaccurate readings are being given, can monitor a process machine's performance and health, and can diagnose problems with process machines.
SUMMARY
The current invention provides a method of data mining for monitoring and controlling a process via a process controller, comprising the steps of collecting and storing process attribute information in a plurality of databases, receiving at least one process measurement from a measurement device, similarity searching the at least one process measurement against the process attribute information stored in the databases, assigning a similarity score to the process measurement, and comparing the similarity score to a match tolerance level. The invention further includes computing a process action for at least one process machine, via an algorithm having a process variable, which comprises replacing the process variable in the algorithm with the process measurement where the similarity score is equal to or greater than the match tolerance level, and replacing the process variable in the algorithm with a set point where the similarity score is lower than the match tolerance level. The process action is then communicated to a process machine.
The process attribute information may comprise the process impact of at least one process machine, process machine performance ranges, process machine conditions, process set points, past process measurements, and any combination of these. The at least one database may reside on a process controller, or may communicate with a process controller, via a network. The network may consist of local area networks, wide area networks, global communication networks, intranet, and Ethernet. The measurement device from which the process measurement is received may comprise a process sensor. The process measurement may be received from the measurement device, via an input/output device.
The step of similarity searching may be performed via a similarity search engine, which may reside on a process controller or communicate with the process controller via a network. This network may be chosen from a group consisting of local area networks, wide area networks, global communication networks, intranet, and Ethernet. The similarity search score may be assigned via a process controller. The process action may be computed via a process controller, and the process action is communicated via a process controller. The current invention is also directed to a software program embodied on a computer-readable medium incorporating the invented method.
In another embodiment of the present invention, a computer-implemented method for monitoring and controlling a process comprises collecting and storing process attribute information in a plurality of databases, receiving at least one process measurement from a measurement device onto a process controller via an input/output device, similarity searching the at least one process measurement against the process attribute information stored in the databases via a similarity search engine, assigning a similarity score to the process measurement via the process controller, comparing the similarity score to a match tolerance level, the process controller computing a process action for at least one process machine via an algorithm having a process variable comprising replacing the process variable in the algorithm with the process measurement where the similarity score is equal to or greater than the match tolerance level, replacing the process variable in the algorithm with a set point where the similarity score is lower than the match tolerance level, and the process controller communicating the process action to a process machine via an input/output device.
In another embodiment of the current invention, a system for monitoring and controlling a process comprises a plurality of databases for storing process attribute information, a means for receiving a process measurement from a measurement device, a similarity search engine for similarity searching the measurement against the process attribute information stored in the databases, a means for assigning a similarity search score to the measurement, a means for comparing the similarity search score to a match tolerance level, a process controller for computing a process action, and a means for communicating the process action to a process machine.
In another embodiment of the current invention, a method for identifying an unidentified object having at least one process attribute comprises collecting information about at least one process attribute of the unidentified object, converting the collected attribute information into a language independent format, arranging the collected information in language independent format in a predetermined sequence, comparing the language independent collected attribute information with information related to a plurality of known objects, wherein the known objects may include a different number and type of attributes than the unidentified object, assigning a value to each of the known objects that indicate the degree of similarity each known object has to the unidentified object, providing a list of known objects most closely matching the unknown object based on the assigned similarity values, and replacing the attribute of the unidentified object with the known object of highest similarity.
In another embodiment of the present invention, a method of data mining objects having attributes for one or more process control loops comprises collecting information about at least one process attribute from a process variable database, converting the collected attribute information into a language independent format, arranging the language independent collected information in a predetermined sequence, comparing the language independent collected attribute information with information related to a plurality of known objects, wherein the known objects may include different numbers and types of attributes than the collected process attribute information, assigning a similarity value to each of the known objects for indicating the degree of similarity each object has to the collected process variable attribute, and providing a list of the known objects that most closely match the object having the collected attribute based on the similarity value.
The similarity searching method used in accordance with the present invention may comprise any suitable similarity searching method or technique. The method may comprise the similarity search method in U.S. Pat. No. 5,666,442 by Wheeler, which is incorporated by reference herein.
REFERENCES:
patent: 5067099 (1991-11-01), McC
Lenz Gary A.
Wheeler David B.
Assouad Patrick J
Infoglide Corporation
Taylor Russell & Russell P.C.
LandOfFree
System and method for monitoring and control of processes... 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 monitoring and control of processes..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for monitoring and control of processes... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3293694