System and method for diagnosing abnormalities in plant control system

• US 7,212,952 B2
• Filed: 03/25/2005
• Issued: 05/01/2007
• Est. Priority Date: 09/27/2002
• Status: Active Grant

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Yamashita, Tatsuo, Hosaka, Hideo, Okazaki, Mitsuyoshi, Asatsu, Seiichi, Watanabe, Shunzo
• Primary Examiner(s): Barlow, Jr.; John E.
• Assistant Examiner(s): Washburn; Douglas N.
• Application Number:
  US11/088,844
• Publication Number:
  US 20050203696A1
• Time in Patent Office: 767 Days
• Field of Search: 702/182, 702/183
• US Class Current: 702/183
• CPC Class Codes: G05B 23/0281 Quantitative, e.g. mathematical distance; Clustering; Neural networks; Statistical analysis

Abstract

Cause of abnormalities of an abnormal state of a control system of a plant of a turbine are estimated and diagnosed from characteristics of abnormal phenomena. By using a control system abnormality cause-and-effect matrix in which characteristics of a plurality of abnormal phenomena constructed on a knowledge base are correlated with a plurality of causes by correlating (weighting) scores, plant devices causing abnormality are inferred and diagnosed.

Claims

1. A control system abnormality diagnosis system of a plant comprising:

a control system abnormality cause-and-effect matrix in which abnormality causes generated in a plurality of devices constituting the control system of the plant are correlated with characteristics of plural kinds of abnormal phenomena generated in the control system because of abnormalities of the devices for running states of the plant by weighting scores in accordance with correlativity thereof;

a control system abnormality cause-and-effect coefficient matrix in which for the characteristics of the plural kinds of abnormal phenomena generated in the control system of the plant, weighting factors are set with respect to the abnormality causes generated in the plurality of devices constituting the control system differently for states of associated devices thereof,

an abnormal phenomenon input element that inputs the characteristics of the abnormal phenomena generated in the control system together with the running state of the plant and the states of the associated devices;

a cause analysis element that analyzes the abnormality causes generated in the control system of the plant based on the weighting scores obtained by the control system abnormality cause-and-effect matrix and the weighting factors obtained by the control system abnormality cause-and-effect coefficient matrix from the characteristics of the abnormal phenomena input from the abnormal phenomenon input element, the running state of the plant, and the states of the associated devices; and

a diagnosing result output element that outputs the abnormality causes of the plant analyzed by the cause analysis element as diagnosing results.

2. The control system abnormality diagnosis system of the plant according to claim 1,

wherein the abnormal phenomenon input element displays the plural kinds of abnormal phenomena generated in the control system, the characteristics of the abnormal phenomena and the running state described in the control system abnormality cause-and-effect matrix, and the states of the associated devices described in the control system abnormality cause-and-effect coefficient matrix item by item, and inputs the characteristics of the abnormal phenomena generated in the control system, the running state of the plant, and the states of the associated devices by selecting the displayed items.

3. The control system abnormality diagnosis system of the plant according to claim 1 or 2, further comprising a communication element connected to an external communication terminal through a network,

wherein the abnormal phenomenon input element inputs the characteristics of the abnormal phenomena generated in the control system in accordance with communication with the external communication terminal connected by the communication element together with the running state of the plant and the states of the associated devices.

4. A control system abnormality diagnosis system of a plant connected to an external communication terminal through a network, comprising:

a control system abnormality cause-and-effect matrix in which abnormality causes generated in a plurality of devices constituting the control system of the plant are correlated with characteristics of plural kinds of abnormal phenomena generated in the control system because of abnormalities of the devices for running states of the plant by weighting scores in accordance with correlativity thereof;

a control system abnormality cause-and-effect coefficient matrix in which for the characteristics of the plural kinds of abnormal phenomena generated in the control system of the plant, weighting factors are set with respect to the abnormality causes generated in the plurality of devices constituting the control system differently for states of associated devices thereof,

an abnormal phenomenon input element that inputs the characteristics of the abnormal phenomena generated in the control system of the plant together with the running state of the plant and the states of the associated devices from the external communication terminal through the network;

a cause analysis element that analyzes the abnormality causes generated in the control system of the plant based on the weighting scores obtained by the control system abnormality cause-and-effect matrix and the weighting factors obtained by the control system abnormality cause-and-effect coefficient matrix from the characteristics of the abnormal phenomena input from the abnormal phenomenon input element, the running state of the plant, and the states of the associated devices; and

a diagnosing result output element that outputs the abnormality causes of the plant analyzed by the cause analysis element as diagnosing results to the external communication terminal through the network.

5. A communication terminal connected to a control system abnormality diagnosis system of a plant through a network,

the control system abnormality diagnosis system comprising:

a control system abnormality cause-and-effect matrix in which abnormality causes generated in a plurality of devices constituting the control system of the plant are correlated with characteristics of plural kinds of abnormal phenomena generated in the control system because of abnormalities of the devices for running states of the plant by weighting scores in accordance with correlativity thereof; and

a control system abnormality cause-and-effect coefficient matrix in which for the characteristics of the plural kinds of abnormal phenomena generated in the control system of the plant, weighting factors are set with respect to the abnormality causes generated in the plurality of devices constituting the control system differently for states of associated devices thereof, and

the communication terminal comprising;

an abnormal phenomenon input element which is connected to the control system abnormality diagnosis system through the network and which inputs the characteristics of the abnormal phenomena generated in the control system of the plant together with the running state of the plant and the states of the associated devices;

an abnormality cause reception element that receives, in response to the characteristics of the abnormal phenomena, the running state of the plant and the state of the associated devices input from the abnormal phenomenon input element, abnormality causes of the plant analyzed based on the control system abnormality cause-and-effect matrix and the control system abnormality coefficient matrix of the control system abnormality diagnosis system; and

a diagnosing result output element that outputs the abnormality causes of the plant received by the diagnosing result reception element as diagnosing results.

6. A control system abnormality diagnosis method of a plant comprising:

inputting characteristics of abnormal phenomena generated in the control system together with a running state of the plant and states of associated devices to a control system abnormality cause-and-effect matrix and a control system abnormality cause-and-effect coefficient matrix by using the control system abnormality cause-and-effect matrix in which abnormality causes generated in a plurality of devices constituting the control system of the plant are correlated with characteristics of plural kinds of abnormal phenomena generated in the control system because of abnormalities of the devices for running states of the plant by weighting scores in accordance with correlativity thereof and the control system abnormality cause-and-effect coefficient matrix in which for the characteristics of the plural kinds of abnormal phenomena generated in the control system of the plant, weighting factors are set with respect to the abnormality causes generated in the plurality of devices constituting the control system differently for the states of the associated devices thereof;

analyzing the abnormality causes generated in the control system of the plant based on the weighting scores obtained by the control system abnormality cause-and-effect matrix and the weighting factors obtained by the control system cause-and-effect coefficient matrix from the characteristics of the abnormal phenomena input together with the running state of the plant and the states of the associated devices; and

outputting the analyzed abnormality causes of the plant as diagnosing results.

7. The control system abnormality diagnosis method of the plant according to claim 6,

wherein the plural kinds of abnormal phenomena generated in the control system, the characteristics of the abnormal phenomena and the running state described in the control system abnormality cause-and-effect matrix, and the states of the associated devices described in the control system abnormality cause-and-effect coefficient matrix are displayed item by item, and the characteristics of the abnormal phenomena generated in the control system, the running state of the plant, and the states of the associated devices are input by selecting the displayed items.

8. The control system abnormality diagnosis method of the plant according to claim 6 or 7,

wherein the characteristics of the abnormal phenomena generated in the control system are input together with the running state of the plant and the states of the associated devices in accordance with communication with an external communication terminal connected by a communication element.