Plant fault diagnosis system

• US 5,070,468 A
• Filed: 10/12/1990
• Issued: 12/03/1991
• Est. Priority Date: 07/20/1988
• Status: Expired due to Fees

• Current Assignee: Idemitsu Kosan Company Limited, Mitsubishi Jukogyo Kabushiki Kaisha (Mitsubishi Heavy Industries, Ltd.)
• Original Assignee: Idemitsu Kosan Company Limited, Mitsubishi Jukogyo Kabushiki Kaisha (Mitsubishi Heavy Industries, Ltd.)
• Inventors: Takahashi, Koichi, Niinomi, Toshihiko, Kaseda, Shigeru, Funakoshi, Ryohei
• Primary Examiner(s): Black, Thomas G.
• Application Number:
  US07/598,770
• Time in Patent Office: 417 Days
• Field of Search: 364/550, 364/551.01, 364/551.02, 364/513, 364/500
• US Class Current: 702/185
• CPC Class Codes: B01J 19/002 Avoiding undesirable reactions or side-effects, e.g. avoiding explosions, or improving the yield by suppressing side-reactions
• CPC Class Codes: B01J 2219/00268 Detecting faulty operations
• CPC Class Codes: G01R 19/16528 using digital techniques or performing arithmetic operations using digital techniques to measure a voltage or a current, see G01R19/25
• CPC Class Codes: G05B 23/0235 based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold
• CPC Class Codes: G05B 23/027 Alarm generation, e.g. communication protocol; Forms of alarm
• CPC Class Codes: G05B 23/0272 Presentation of monitored results, e.g. selection of status reports to be displayed; Filtering information to the user
• CPC Class Codes: G08B 23/00 Alarms responsive to unspecified undesired or abnormal conditions
• CPC Class Codes: Y10S 706/914 Process plant

Abstract

The present invention relates to a plant fault diagnosing system for diagnosing the operating conditions of the plant, for example, change in the physical quantity or whether any fault is present in each of the instrumentations or not. This system makes reference and makes directly comparison the process data of the plant or a diagnosing pattern processed by treating the process data in a predetermined process with a predetermined information for detecting showing the symptom of an abnormality or a predetermined reference value. When any deviation from a predetermined allowable range or coincidence of the patterns is detected, it determines that there is any abnormality in the plant, and issues an alarm to, for example, an operator.

Claims

1. A plant fault diagnosing system, comprising:

instrumentation means for providing process data representing physical operational characteristics of a plant process, diagnosing means for diagnosing faults in the plant in accordance with a relationship between said process data and at least one of predetermined information regarding fault diagnosis and a changeable reference value based on said process data, said diagnosing means including means for generating a diagnosing pattern which symbolically represents said process data;

means for displaying the diagnosis given by said diagnosing means;

a memory for storing said predetermined information and also for outputting said predetermined information to said diagnosing means, said predetermined information including importance factors which define the relative importance of the diagnosed faults;

said diagnosing means including means for estimating a first amount of time which will elapse between the diagnosis and the actual occurrence of each fault based upon the physical properties of the plant process and said process data, means for estimating a second amount of time required to avoid the occurrence of the fault, means for subtracting said second amount of time from said first amount of time to thereby calculate an estimated floating time available before a fault avoidance operation should begin, and means for calculating an urgency factor based on said estimated floating time and said importance factor; and

said displaying means including means for ranking and displaying the faults in sequential order based upon respective magnitudes of said urgency factor and said floating time.

2. The plant fault diagnosing system according to claim 1, wherein said diagnosing means includes instrument failure diagnosing means for diagnosing a failure of said instrumentation means, and general fault diagnosing means for diagnosing abnormalities other than the failures of said instrumentation means by ignoring process data associated with any failure of said instrumentation means as detected by said instrument failure diagnosing means.

3. The plant fault diagnosing system according to claim 1, including an average value calculator connected to said diagnosing means and having means for calculating an average value of said process data over a predetermined time interval, said average value of said process data being said changeable reference value, wherein said diagnosing means includes means for calculating the difference between said average value and said process data, said diagnosing means having means for issuing an alarm when said difference deviates from a predetermined allowable range, and said average value calculator having holding circuit means for preventing calculation of said average value during any time period when said alarm is being issued.

4. The plant fault diagnosing system according to claim 3, including an allowable range calculator connected to said diagnosing means and having means for calculating a statistical region where said difference between said average value and said process data is distributed and for determining said allowable range based on said statistical region.

5. The plant fault diagnosing system according to claim 4, including means for applying reset signals to said average value calculator and said allowable range calculator, said calculators having respective means responsive to said reset signals for causing said average value and said allowable range to be calculated again by using process data from a time point at which said reset signals are supplied.

6. The plant fault diagnosing system according to claim 3, including means for applying a reset signal to said average value calculator, said average value calculator having means responsive to said reset signal for causing said average value to be calculated again by using process data from a time point at which said reset signal is supplied.

7. The plant fault diagnosing system according to claim 3, wherein said diagnosing means includes means for processing said diagnosing pattern by using the alarm.

8. The plant fault diagnosing system according to claim 1, including an average value calculator which is connected to said diagnosing means and includes means for calculating an average value of said process data over a predetermined time interval, said average value of said process data being said changeable reference value, wherein said diagnosing means includes means for calculating the difference between said average value and said process data, and means for issuing an alarm when the difference between said average value and said process data deviates from a predetermined allowable range for a predetermined continued time period.

9. The plant fault diagnosing system according to claim 1, including an average value calculator connected to said diagnosing means and having means for calculating an average value of said process data over a predetermined time interval, said average value of said process data being said changeable reference value, wherein said diagnosing means includes means for calculating the difference between the average value and the process data, and means for issuing an alarm when said difference deviates from a predetermined allowable range, and said diagnosing means including means for processing said diagnosing pattern by using said alarm.

10. The plant fault diagnosing system according to claim 1, including knowledge memory means for supplying said predetermined information regarding fault diagnosis to said diagnosing means, said predetermined information representing possible faults and symptoms associated with said possible faults, said possible faults being classified in said knowledge memory means according to one of the likelihood and intensity of occurrence of said symptoms in conjunction with each said possible fault.

11. The plant fault diagnosing system according to claim 10, wherein said diagnosing means includes means for relating said process data to said predetermined symptom information and for thereby classifying an actual fault in the plant as one of said possible faults represented by said predetermined information.

12. The plant fault diagnosing system according to claim 1, wherein said displaying means includes means for displaying each of said importance factor, said urgency factor, said first and second amounts of time, and said floating time.

13. A fault diagnosing system for diagnosing faults in an operative target system, comprising:

data collection means for collecting data representing a physical operational state of the target system;

diagnosing means for interpreting said data and determining on the basis thereof whether said data indicates that a fault will occur in the target system, said diagnosing means including fault time estimating means for estimating a first amount of time which will elapse between said indication of said fault and an actual occurrence of said fault in the target system, avoidance time estimating means for estimating a second amount of time which will be required for taking corrective action to avoid the occurrence of said fault, and floating time estimating means for subtracting said second amount of time from said first amount of time to thereby calculate an estimated floating time available before said corrective action should begin; and

display means for displaying faults indicated by said data, including ranking means for ranking said faults for display based on the magnitudes of their respective floating times.

14. The fault diagnosing system according to claim 13, wherein said display means includes further means for displaying each of said first amount of time, said second amount of time, and said floating time associated with each indicated fault.