Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
1999-02-17
2002-12-24
Beausoleil, Robert (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S724000, C709S224000, C399S026000
Reexamination Certificate
active
06499114
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention generally relates to a method and apparatus for locally collecting data and diagnosing malfunctions. More particularly, the present invention is related to a system and a method that enable a combination of on-site anomaly detection and remote data analysis.
Safety and efficiency considerations require that operating conditions of various processes be continuously monitored to obtain an early indication of possible malfunctions. Early indications of possible malfunctions allow corrective measures to be taken more quickly, thereby reducing waste and preventing more serious and potentially hazardous conditions.
Some industrial processes, such as, the generation of electricity, the operation of an aircraft engine or locomotive, the manufacturing of steel or paper, or other processes implemented by industrial drives or equipment, utilize hundreds or thousands of sensors to obtain on-line indications of operating conditions and parameters. In one particular example of an industrial process, turbine generator equipment in a power plant can be remotely monitored and diagnosed to improve its availability and performance. Certain turbine generator equipment can be utilized with an on-site monitor that collects and stores data which is downloaded to a central site for diagnostic analysis. The central site can utilize any number of expert systems or rule-based systems to automatically detect malfunctions. Additionally, an operator or engineer can review the data to determine if the turbine is functioning properly. Other processes, such as, those processes occurring in airplanes, locomotives, steel mills, paper mills, and factories, can utilize equipment which is remotely monitored in a similar fashion to the turbine generator equipment described above.
To improve availability of equipment and performance of complex processes, the earliest possible detection of detrimental conditions is necessary. Early detection permits appropriate action to be taken to minimize the impact of such conditions and to prevent consequential damage. Early detection can be achieved through continuous detailed evaluation of the operating condition of the equipment (e.g., the turbine and generator equipment in a power plant). This requirement for early detection is especially critical when deterioration of equipment can happen in relatively short periods of time (in several minutes to a few hours), which, if left uncorrected, can have significant impact on the availability and performance of the equipment. Equally important to improved availability and performance of complex processes is the ability to accurately identify a developing condition. For example, in a power plant, the consequence of unnecessarily interrupting turbine operation is costly to an operator. Similarly, in other industrial processes, unnecessarily interrupting equipment is costly, especially in complex processes which utilize expensive equipment, such as, steel fabrication, paper fabrication, airplane operation and locomotive operation. Thus, the accurate diagnosis of detrimental conditions and developing conditions is critical to the efficient operation of complex processes.
The key factors to precise identification of detrimental conditions and developing conditions are accuracy and completeness of the data (e.g., the appropriate data logged in the most informative manner), timeliness of the data, and background fleet data to facilitate accurate analysis.
Generally, sensor signals are not continuously stored in memory because of the large amount of communication and storage overhead associated with the large number of sensors. Accordingly, some systems utilize a “deadband” technique to store sensor signals; the deadband technique only stores a signal when the signal exceeds a previously stored value by a predetermined amount (e.g., only stores those signals which exhibit a predetermined difference in value from a previous scan). With such systems, general trends and changes associated with the sensor signals over long periods of time can be stored efficiently in memory.
Alternatively, some systems utilize a time-coherent data storage technique where the sensors are all scanned simultaneously and periodically stored. The sensors are stored for a period of time or for a particular number of readings. However, the archival time interval and number of readings are limited due to memory requirements. In addition, this technique can miss recordation of rapid signal changes if the signal change does not occur during the time of the reading (e.g., in a period between readings).
Although change-detect data storage techniques can be effective for data storage where the evaluation or analysis depends solely upon changes of monitored parameters, change-detect data storage techniques do not provide accurate definitions of instantaneous relationships among parameters. Accurate definitions of instantaneous relationships among parameters are critical to predicting and diagnosing malfunctions. Although time-coherent data storage techniques are effective where instantaneous relationships among systems parameter can be solely used for evaluation and analysis, rapid changes in parameter values are often missed in such an archiving scheme.
Conventional systems generally rely on the central site to perform sophisticated diagnostic analysis to detect malfunctions and anomalies. The central site includes computer equipment which can perform expert system diagnostics on the data downloaded from the remote location. In addition, it includes experts who can validate the conclusions of the automated diagnostic results. However, with such a scheme, anomaly detection and malfunction diagnosis cannot occur until the data is made available to the central site. Accordingly, a delay can prevent the central site from uncovering serious and potentially hazardous conditions as quickly as possible without continuous transfer of data to the central location. On-line or real time expert system anomaly detection is not possible without continuous transfer of data to the central location. In addition, conventional systems which solely rely upon central site diagnostics require that communication overhead be utilized to transfer data when systems are working appropriately and anomalous conditions are not present (i.e., communication overhead is wasted transferring “healthy” data to the central site).
Accordingly, there is a need for a diagnostic system that is not susceptible to the disadvantages associated with a change-detect only or time-coherent data only storage technique. Further still, there is a need for a diagnostic system that performs local anomaly detection to reduce response time to anomalous equipment or process conditions and to reduce the transfer of healthy unit data to a central location. Even further still, there is a need for a diagnostic system that combines the advantages of time-coherent data storage techniques and change-detect data storage techniques. Yet further, there is a need for a diagnostic system which performs real time, on-line anomaly detection.
SUMMARY OF THE INVENTION
The present invention relates to a diagnostic system for analyzing operation of a process. The process is monitored by sensors, each of which senses a different parameter associated with the operation of the process. The diagnostic system includes a local diagnostic unit. The local diagnostic unit is located at a site associated with the process and receives sensor signals from the sensors. The local diagnostic unit periodically and simultaneously stores the sensor signals. The local diagnostic unit stores one sensor signal of the sensor signals whenever the one sensor signal deviates from a previously stored value by a first predetermined amount.
The present invention relates to a diagnostic system for analyzing operation of a process. The diagnostic system includes a local database, and a local diagnostic unit. The local diagnostic unit is located at a site associated with the process. The local diagnostic unit receives a plural
Almstead John Gamwell
Longtin Keith Andrew
Morjaria Maheshch Amritlal
Beausoleil Robert
Ziemer Rita A
LandOfFree
Remote diagnostic system and method collecting sensor data... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Remote diagnostic system and method collecting sensor data..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Remote diagnostic system and method collecting sensor data... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2944257