JP2007109139A - Monitoring diagnostic apparatus and monitoring diagnostic method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology capable of diagnosing deficient situations of equipment even when no abnormal phenomenon to be referred to has yet happened. <P>SOLUTION: On the basis of an abnormality decision database 121 and abnormality diagnostic database 122, a status of the equipment which is decided to be abnormal is retrieved in a cases in the past database 123 and a knowledge database 124 and its influence degree and abnormality level are diagnosed. Thus, even when no deficiencies have yet happened, the deficient situations of the equipment can be diagnosed. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a monitoring / diagnosis apparatus and a monitoring / diagnosis method for equipment used in a power plant or the like.

  Conventionally, there is a technique described in Patent Document 1 as a technique for monitoring equipment used in a plant or the like and diagnosing the presence or absence of abnormality.

  The technique described in Patent Document 1 determines the presence or absence of abnormality based on the detection result of a sensor installed in equipment used in a plant, information relating to a signal item determined to be a failure, and a failure corresponding to this item. The failure status of the device is diagnosed by referring to the failure event from the case database stored in association with the information related to the event.

JP 2003-295937 A

  However, in the technique described in Patent Document 1, when a detection result of a sensor that is not stored in the case database occurs, it is not possible to search for a failure event to be referred to and to perform a diagnosis. It was.

  Therefore, an object of the present invention is to provide a technique capable of diagnosing a malfunction state of a device even when a malfunction event that should be referred to in the past has not occurred.

  In order to solve the above problems, the present invention acquires information from a knowledge database related to a failure event when there is no corresponding information in the case database storing past cases, and the degree of influence or abnormality is acquired from the acquired information. Provide a technology for determining the level and diagnosing equipment malfunctions.

  For example, the present invention is an apparatus monitoring and diagnosing apparatus, an abnormality diagnosis database storing a device name of a monitoring target apparatus that is a monitoring target, the process state quantity, and the state information, and the monitoring target A past case database storing device names of devices, information related to failures that occurred in the past in the monitored devices, and information related to handling the failures, and devices of non-monitored devices other than the monitored devices A storage unit storing a name, information relating to a failure of the non-monitoring target device, information relating to handling the failure, and a process state quantity of the monitoring target device, the monitoring target When it is determined that a failure has occurred in the device, the abnormality diagnosis database is based on the device name of the monitored device and the process state quantity. The state information is read out from, and the past case database is searched based on at least one of the device name of the monitoring target device and the state information, and the knowledge database is searched when there is no corresponding past case A processing unit.

  As described above, according to the present invention, even when there is no corresponding information in the case database storing past cases, the information is acquired from the knowledge database regarding the failure event, and the influence level and the abnormality level are obtained from the acquired information. Since the determination can be made, the failure status of the device can be diagnosed even when there is no failure event that should be referred to in the past.

  FIG. 1 is a schematic diagram of a monitoring / diagnosis apparatus 100 according to an embodiment of the present invention.

  As shown in FIG. 1, the monitoring diagnosis apparatus 100 includes an arithmetic processing unit 110, a storage unit 120, a communication unit 130, an I / F (Interface) unit 140, an input unit 150, and a display unit 160. And.

  The arithmetic processing unit 110 includes an abnormality determination processing unit 111, a past case search processing unit 112, a knowledge search processing unit 113, an influence degree evaluation processing unit 114, an abnormality level determination processing unit 115, and a registration unit 116. Have.

  The abnormality determination processing unit 111 acquires a detection signal from a sensor attached to the monitored device via the I / F unit 140, calculates a process state quantity from the detection signal, and determines an abnormality. The data stored in the database 121 is read out, and the presence / absence of an abnormality is determined based on the determination method stored in the abnormality diagnosis database 122.

  Specifically, first, the logic No. input in the method (logic) column 121c set for each device in the abnormality determination database 121 shown in FIG. The determination method to be used for this logic No is specified from the abnormal event determination file 122A being used (in the abnormal event determination file 122A, the determination method marked with a circle in the column of the method (logic) name 122b is the determination method to be used). ) Next, the specified determination method is searched from the determination method name column 122b of the determination method file 122B shown in FIG. 5B, and the sensor type and determination formula are specified in the line of the searched determination method. The presence / absence of an abnormality is determined by determining the process state quantity measured by the specified sensor type using the specified determination formula.

  Further, when the abnormality determination processing unit 111 determines that the calculated process state quantity is abnormal based on the determination formula, the device name of the device determined to be abnormal is acquired from the device information column 121a of the abnormality determination database 121. In addition, the state information related to the state represented by the process state amount is acquired from the state information column 122f of the determination method file 122B of the abnormality diagnosis database 122, and the information is sent to the past case search processing unit 112 described later. .

  In the past case search processing unit 112, when the abnormality determination unit 111 determines abnormality, a similar case occurs in the past based on the device name and status information sent from the abnormality determination processing unit 111. If there is a hit case, information on the hit case is acquired from the past case database, converted into a predetermined format, and displayed on the display unit 160. On the other hand, if there is no hit case, the device name and status information sent from the abnormality determination processing unit 111 is converted into a predetermined format and displayed on the display unit 160.

  In the present embodiment, the keyword search is performed from the device name column 123f and the occurrence event column 123h of the past case database 123 based on the device name and status information. However, the present invention is not limited to such a mode.

  When the knowledge search processing unit 113 receives an instruction via the input unit 150 to close the display screen of the device name and status information displayed when there is no case hit in the past case search processing unit 112, FIG. The displayed knowledge search screen 170 is displayed on the display unit 160, the knowledge database 124 is searched under the input conditions, information about the hit knowledge is acquired, converted into a predetermined format, and displayed on the display unit 160. Process to display.

  The knowledge search screen 170 includes a search target selection field 171, a search expression input field 172, a keyword input field 173, a weight value input field 174, a search button 175, and a reset button 176.

  The search target selection column 171 is a column for selecting a target (database) in which information to be searched is stored. In the present embodiment, an in-house check column 171a, an Internet check column 171b, and other check columns 171c are provided. When the input unit 150 selects a check box, information is searched in the selected target.

  Here, when “inside” is selected, a search is performed on the knowledge database 124, and when “Internet” is selected, a search is performed on the Internet via a predetermined search engine. When “others” is selected, when a search database other than the past case database 123 and the knowledge database 124 is stored in the storage unit 120, the search database is searched (this embodiment). In the embodiment, since such a search database is not provided, the search is not performed even if “others” is selected).

  In the search expression input field 172, a predetermined search expression for searching the knowledge database 124 is input.

  In the keyword input field 173, a keyword for performing a search is input in each field.

  In the present embodiment, the search expression input field 172 and the keyword input field 173 are searchable when input is made to either one, and so-called AND search is performed when input is made to both. To be done.

  In the weight value input field 174, the weight value of the keyword input in the keyword input field 173 adjacent to the weight value input field 174 is input. In the present embodiment, numerical values “1”, “2”, and “3” are input, and the degree of weight (importance) is set to be “1”> “2”> “3”. ing.

  When the search button 175 is pressed via the input unit 150, the keyword selected in the search expression input field 172 or the keyword input in the keyword input field 173 for the target selected in the search target selection field 171. The search is executed.

  In the present embodiment, the search expression input in the search expression input field 172 or the keyword input in the keyword input field 173 is used as the device item 124a, the component item 124b, and the like in the knowledge database 124 shown in FIG. Search is performed using the deterioration / damage mode item 124c and the sensor signal / operation information item 124e.

  In addition, when a keyword is input in the plurality of keyword input fields 173, a so-called “and search” is performed.

  When the reset button 176 is pressed via the input unit 150, the items selected or input are reset in the search target selection field 171, the search expression input field 172, the keyword input field 173, and the weight value input field 174, and become blank. It comes to return.

  Here, in the search in the knowledge database 124, the search target is filtered based on the input weight value and the number of matches of the searched keyword, and the weight is high by these weight values, and the number of matches is The more knowledge there is, the more preferential information is displayed on the display unit 160. In addition, when the priority according to the weight value and the number of matches matches, the higher weight value is given priority. Further, when there are a plurality of items that match the keyword, the priority may be determined based on the number of keyword matches in the character information input in the character information item 124h.

  FIG. 3 shows a knowledge search result screen 180 that is an example of a search result screen of the knowledge database 124. As shown in FIG. 3, the knowledge search result screen 180 has a display field 181 for each knowledge. The display field 181 is displayed on the screen of the display unit 160 by selecting a tab 182. The display of is switched. In the display column 181, the knowledge displayed in the display column 160 located on the left side in FIG. 3 has a higher weight in the search keyword and a larger number of matches with the search keyword.

  The display column 181 includes an image display column 181a such as a drawing or a photograph of a part or the like that is the subject of knowledge, a character information display column 181b that describes the content of the knowledge, and related knowledge, reference materials, and knowledge targets. A supplementary information display field 181c for displaying supplementary information such as the specifications of the parts and the like is provided.

  Further, a menu display field 183 for displaying menu selection buttons such as printing and changing the display method is provided above the display field 181 in FIG.

  The knowledge search processing unit 113 performs a search based on the keyword input in the keyword input field 173 using a predetermined search search engine when there is no knowledge hit in the knowledge database 124. When no keyword is input in the keyword input field 173, a search is performed based on the device information and the state information specified by the abnormality determination processing unit 111. The search result thus performed is displayed on the display unit 160 in a predetermined display format.

  When the specific knowledge is hit in the knowledge search processing unit 113, the influence degree evaluation processing unit 114 identifies the hit knowledge having the highest weight in the search keyword and the largest number of matches with the search keyword, Information regarding the degree of influence of the knowledge is acquired from the knowledge database, converted into a predetermined format, and displayed on the display unit 160.

  Here, the information on the degree of influence is information such as cost and time (number of days) required to solve the malfunction when the malfunction occurs in the component, etc., and in this embodiment, the recovery is performed. Information such as the time required (number of days), the total number of people required for handling, the number of processes to deal with, the cost required for restoration, the amount of loss caused by malfunctions, etc. are displayed in a predetermined format, but not limited to these It is not done.

  Note that when the specific knowledge is not hit by the knowledge search processing unit 113, the influence evaluation processing unit 114 does not perform these display processes.

  The abnormality level determination processing unit 115 evaluates the state of the component of the device determined to be abnormal from a predetermined physical model or empirical formula, calculates the failure probability of the device as a whole from the failure probability of this component, It is determined by dividing it into different levels.

  Specifically, for a device determined to have an abnormality, the total equivalent operation time is calculated from the operation start date and operation time for each component of the device from the abnormality level determination database 125, and this total equivalent The operation time and a threshold value calculated from the management time (in this embodiment, a value between 50 and 80% of the management standard is selected in advance for each part) are compared, and the parts that exceed this threshold are weibull. The failure probability is calculated from the distribution, and the failure probability of the entire device is calculated according to the importance of the parts exceeding the threshold. For example, the damage probability of the entire device is determined in advance so that the importance of each component constituting the device is 100% as a whole, and the damage of the entire device is added by adding the importance of the component exceeding the threshold Probability can be calculated. The total equivalent operation time is calculated by multiplying the total operation time calculated from the operation start date and operation time by a predetermined coefficient.

  Here, in this embodiment, this level division is divided into level 1, level 2 and level 3. Level 1 has a failure probability of the entire device of 0% to less than 10%, and level 2 has a failure probability of the entire device. The level division is performed so that the probability of damage of the entire device is 50% or more and 100% or less in 10% or more and less than 50%, but it is not limited to such a mode.

  The registration unit 116 performs processing for inputting predetermined information to the abnormality determination database 121, the past case database 123, and the knowledge database 124. Specifically, the information input by the input unit 150 is registered in each database.

  The arithmetic processing unit 110 described above can be realized by executing a predetermined program stored in the storage unit 120 by a CPU (Central Processing Unit).

  The storage unit 120 includes an abnormality determination database 121, an abnormality diagnosis database 122, a past case database 123, and a knowledge database 124.

  As shown in FIG. 4, the abnormality determination database 121 includes a device information column 121a, a sensor type column 121b, a determination method column 121c, an abnormal event code column 121d, and a remarks column 121e. Predetermined information is input in advance using the registration unit 116 and the input unit 150.

  In the device information column 121a, the name of the device to which the sensor is attached is input.

  The sensor type column 121b inputs the type of sensor attached to the device input in the device information column 121a.

  In the technique (logic) input column 121c, a logic No. for determining an abnormality from the process state quantity calculated from the detection signal of the sensor attached to the device input in the device information column 121a is input.

  The abnormal event code column 121d includes an abnormal event code No. assigned in advance for each abnormal event in which the sensor specified by the sensor type 121b is determined to be abnormal by the determination method specified by the logic No in the method (logic) column 121c. Enter.

  In the remarks column 121e, an additional item for each device input in the device information column 121a can be freely input.

  As shown in FIG. 5, the abnormality diagnosis database 122 includes an abnormal event determination file 122A [see FIG. 5A] and a determination method file 122B [see FIG. 5B].

  In the abnormal event determination file 122A, a determination method name column 122b is provided in the leftmost column of FIG. 5A, and each determination method name is input to the vertical column in the determination method name column 122b. Further, a technique (logic) column 122b is provided in the uppermost line of FIG. 5A, and a logic No is input to the horizontal line in this technique (logic) column 122b. Then, whether or not to use the determination method for the logic No is input to the region specified by the intersection of the row of the logic No and the column of the determination method. Here, in this embodiment, when the determination method is used with the logic No, “◯” is input, while when it is not used, “×” is input.

  In the lowermost line of FIG. 5A, it is possible to specify a criterion for determining an abnormality when one or more determination methods are used with a specific logic No. Here, in this embodiment, when it is determined to be abnormal when it is determined as abnormal by all the determination methods marked with “◯” in the column, “○” is displayed, and “○” is displayed in the column. When it is determined to be abnormal when any of the attached determination methods is determined, “Δ” is input.

  As shown in FIG. 5B, the determination method file 122B includes a determination method name column 122c, a sensor type column 122d, a determination formula column 122e, and a status information column 122f.

  The determination method name column 122c inputs a determination method name corresponding to the determination method name column 122b of the abnormal event determination file 122A shown in FIG.

  In the sensor type column 122d, a sensor type name for calculating the process state quantity used in the determination method specified in the determination method name column 122c is input.

  In the determination formula column 122e, a determination formula for determining an abnormality based on the process state quantity calculated by the sensor specified in the sensor type column 122d is input. Here, in FIG. 5B, a, b, and d are process state quantities calculated from the detection signals of the sensor types A, B, and D, respectively, and Δc / Δt is calculated from the detection signal of the sensor type C. The process state quantity c is increased per predetermined time, and S, T, U, V, and W are predetermined threshold values.

  In the state information column 122f, information representing a state determined to be abnormal by determining the process state amount calculated by the sensor specified in the sensor type column 122d based on the determination formula column 122e is input. For example, when the sensor type A is a temperature sensor and the determination formula “a> S” is “a> 100 ° C.”, information such as “high temperature” is input in the state information column 122f.

  As shown in FIG. 6, the past case database 123 includes a classification number column 123a, an abnormal event code column 123b, a customer column 123c, a site column 123d, a number series / axis column 123e, and a device column 123f. And a date / time column 123g, an occurrence event column 123h, a cause column 123i, a countermeasure column 123j, a recovery date column 123k, and a comment column 123l.

  A predetermined classification number for classifying abnormal cases that occurred in the past is entered in the classification number column 123a.

  In the abnormal event code column 123b, a predetermined abnormal event code is input in correspondence with the abnormal event code column 121d of the abnormality determination database 121 shown in FIG.

  In the customer column 123c, the name of the customer who is using the device that detects the abnormality is input.

  In the site column 123d, the customer specified in the customer column 123c inputs information about a factory, a building, or the like that uses a device that detects an abnormality.

  In the number series / axis column 123e, if there is a number series No. or axis No. such as a factory or a building where an apparatus for detecting an abnormality is used, these pieces of information are input.

  In the device column 123f, the name of the device determined to be abnormal is input.

  In the date / time column 123g, the date and time when the abnormality occurred are entered.

  In the occurrence event column 123h, information related to a state when it is determined to be abnormal based on the process state amount calculated from the signal measured by the sensor is input.

  In the cause column 123i, the cause when the state determined to be abnormal occurs is input.

  In the countermeasure column 123j, a countermeasure taken for a state determined to be abnormal is input.

  The date of recovery is entered in the recovery date column 123k.

  In the comment field 123l, any additional items can be freely entered.

  In the present embodiment, since keyword search is performed on the past case database 123, information is input in these fields using a predetermined character string.

  As shown in FIG. 7, the knowledge database 124 includes a device item 124a, a component item 124b, a deterioration / damage mode item 124c, a damage model item 124d, a sensor signal / operation information item 124e, and an influence degree. The information item 124f, the screen information item 124g, and the manual item 124h are formed to have a tree structure.

  The name of the device in which the sensor is installed is input in the device item 124a.

  In the component item 124b, the name of a component that is a component of the device input in the device item 124a and that is measured by the sensor is input. In FIG. 7, a line is drawn between related items.

  In the deterioration / damage mode item 124c, a matter that causes the component to deteriorate or be damaged is input.

  In the damage model item 124d, a model of how a part is damaged is classified in advance, and the classification number of the model is input.

  In the sensor signal / operation information item 124e, information on a process state quantity for measuring by a sensor that a component has deteriorated or damaged and information on a state specified by the process state quantity are input.

  In the influence degree information item 124f, information such as cost and time (number of days) necessary for solving the defect when the defect occurs is input. In this embodiment, information such as the time (number of days) required for recovery, the total number of persons required for response, the number of processes required for response, the cost required for recovery, and the amount of loss caused by a malfunction are displayed in a predetermined format. However, it is not limited to these.

  In the image information item 124g, image information such as a drawing or a photograph of a part or the like to be displayed in the display field 181 of the knowledge search result screen 180 when the knowledge search processing unit 113 performs a search is input.

  In the character information item 124h, character information such as a description of a part to be displayed in the display field 181 of the knowledge search result screen 180 is input.

  In the present embodiment, since the keyword search is performed on the knowledge database 124, information is input as a predetermined character string in these items.

  As shown in FIG. 8, the abnormal level determination database 125 includes a part name field 125a, a part number field 125b, a management criteria field 125c, a threshold value list 125d, an operation start date field 125e, A file 125A having a time column 125f and an importance column 125g is created for each device.

  The part name is input in the part name column 125a.

  In the part number column 125b, an identification number for each part determined by a predetermined method is input.

  In the management standard column 125c, the replacement standard (life) of each part is specified and input by the period (time).

  In the threshold value list 125d, a period (time) from the start of use at which the probability that the part is damaged becomes high is input as a percentage with respect to the period (time) input in the management reference field 125c. Specifically, in this embodiment, a numerical value between 50 and 80% is input. This is because 50 to 80% of the replacement standard (lifetime) of each part has passed, and it is considered that the probability that the part will be damaged becomes high, but it is limited to such a numerical range. That's not true.

  In the operation start date column 125e, the date of starting use of the part is input.

  In the operation time column 125f, how long a part is used per day is input.

  In the importance column 125g, the importance of the part with respect to the damage of the entire device is input so that it becomes 100% for all the parts.

  The storage unit 120 described above can be realized by an auxiliary storage device such as a hard disk.

  The communication unit 130 is an interface for performing communication via a LAN (Local Area Network). In the present embodiment, a NIC (Network Interface card) is used.

  The I / F unit 140 is an interface for sending a signal from the sensor to the abnormality determination unit 111.

  The input unit 150 is an input device such as a mouse or a keyboard.

  The display unit 160 is a display device such as a display.

  An example of use of the monitoring / diagnosis apparatus 100 configured as described above is shown in FIG.

  FIG. 9 shows an example in which the monitoring / diagnosis apparatus 100 is applied to the gas turbine power generation facility 200.

  The gas turbine power generation facility 200 includes, for example, a gas turbine 210 and a power generator 220A driven by the gas turbine 210. Further, in order to perform combined power generation, an exhaust heat recovery boiler 221 and a steam turbine 222, a condenser 223, and a generator 220B. The gas turbine 210 includes a gas turbine compressor 211, a combustor 212, and a turbine 213.

  Various sensors 230 are attached to the gas turbine compressor 211, the combustor 212, the turbine 213, the generators 220A and 220B, the exhaust heat recovery boiler 221, the steam turbine 222, and the condenser 223. These various sensors 230 send signals to the monitoring / diagnosis apparatus 100 in order to calculate the process state quantities of the attached devices, and the monitoring / diagnosis apparatus 100 receives these signals via the I / F unit 140. .

  Further, a router 240 is connected to the communication unit 130 of the monitoring / diagnosis apparatus 100, and the Internet 241 can be connected via the router 240.

  A monitoring diagnosis method in the monitoring diagnosis apparatus 100 used as described above will be described with reference to the flowchart shown in FIG.

  In the monitoring / diagnosis device 100, signals transmitted from various sensors 230 connected to each device are received via the I / F unit 140, and the abnormality determination processing unit 111 calculates a process state quantity from the received signal. Then, the presence or absence of abnormality is monitored (S250). Specifically, the abnormality determination processing unit 111 reads out data stored in the abnormality determination database 121, and obtains a logic No. of a method (logic) for determining abnormality from the type of sensor and the connected device. The determination method is specified from the logic No based on the abnormal event determination file 122A. Since the determination method is specified in the abnormal event determination file 122A, a determination expression for detecting an abnormality is specified based on the process state quantity obtained for each sensor type. Therefore, an abnormality in the process state quantity is monitored by this determination expression. To do.

  If it is not determined that there is an abnormality based on the comprehensive determination in the abnormal event determination file 122A (S251), the process returns to S250 and monitoring is continued.

  On the other hand, if it is determined that there is an abnormality based on the comprehensive determination in the abnormal event determination file 122A (S251), the past case search processing unit 112 searches for past cases (S252). In the past case search, a keyword search is performed based on the device name of the device determined to be abnormal and the state information regarding the state represented by the process state amount determined to be abnormal.

  If there is a keyword search that hits a past case (S252), the corresponding information in the past case database 123 is changed to a predetermined format and displayed on the display unit 160 (S253). Such display ends when the user issues a display end command via the input unit 150 (S254), and when the display of past cases ends, the process proceeds to S267.

  On the other hand, in the past case, when there is no hit case (S255), the display unit 160 displays the state information regarding the state represented by the device name of the device determined to be abnormal and the process state amount determined to be abnormal. (S255). Such display ends when the user issues a display end command via the input unit 150 (S256). When the display of past cases ends, the knowledge search processing unit 113 displays the knowledge search screen 170. (S257), and the user inputs predetermined items via the input unit 150 and instructs the search to start with the search button 175, whereby the search is started (S258).

  When a knowledge that matches the keyword or the search expression is found (S259), the knowledge search processing unit 113 extracts necessary information from the knowledge database 124 and displays the knowledge search result screen shown in FIG. 180 is generated, and the search target is filtered based on the input weight value and the number of matches of the searched keyword, and the one having a high weight and a large number of matches is preferentially displayed on the display unit 160. (S260). Such display ends when the user issues a display end command via the input unit 150 (S261).

  When the knowledge display is completed (S261), the impact evaluation processing unit 114 has the highest weight in the search keyword among the knowledge hit in the knowledge search processing unit 113, and the number of matches with the search keyword is large. Is acquired from the influence information item 124f of the knowledge database 124, converted into a predetermined format, and displayed on the display unit 160 (S262). Such display ends when the user issues a display end command via the input unit 150 (S263). And when such a display is complete | finished, it progresses to S267.

  On the other hand, when the knowledge that hits the keyword or the search expression is not found (S259), the knowledge search processing unit 113 uses the predetermined search search engine to enter the keyword or abnormality input in the keyword input field 173. A search is performed based on the device name and status information specified by the determination processing unit 111 (S264). The search results thus performed are displayed on the display unit 160 in a predetermined display format (S265). Such display ends when the user issues a display end command via the input unit 150 (S266), and when such display ends, the process returns to S250 to monitor process data.

  Further, when the display of the past case is completed (S254) or the display of the influence level is completed (S263), the abnormality level determination processing unit 115 sets the state of the component determined to be abnormal to a predetermined physical state. Evaluation is made from a model or an empirical formula, and the failure probability of the entire device including the part is divided into predetermined levels to determine an abnormal level (S267). Specifically, for a device determined to have an abnormality, the device 125A is read from the abnormality level determination database 125, and the total equivalent operation time is calculated from the operation start date and operation time for each part. Compare the total equivalent operation time with the threshold value calculated from the management time, calculate the failure probability for each part from the Weibull distribution for the parts exceeding this threshold value, and The damage probability of the entire device is calculated according to the degree.

  If the abnormality level is larger than 1 as a result of the abnormality level determination (S268), the abnormality level determination processing unit 115 changes the damage probability for each part and the damage probability of the entire device to a predetermined format and displays them. (S269). Such display ends when the user issues a display end command via the input unit 150 (S270). When the display ends, the process returns to S250 to monitor process data.

  If the abnormal level is 1 (S268), the process returns to S250 to monitor process data.

  In the embodiment described above, as shown in FIG. 9, all the processing is performed by one monitoring and diagnosing device 100. For example, as shown in FIG. The first monitoring and diagnosing device 300A is disposed at a position adjacent to the facility 200, and the second monitoring and diagnosing device 300B connected to the first monitoring and diagnosing device 300A via the router 240 and the Internet 241 is replaced with a gas turbine power generation facility. The first monitoring and diagnosing apparatus 300 </ b> A and the second monitoring and diagnosing apparatus 300 </ b> B may be arranged at a position away from 200 and perform the distributed processing.

  For example, as shown in FIG. 11, the first monitoring / diagnosis apparatus 300A has only the abnormality determination database 321A and the abnormality diagnosis database 322A, and the device name and state determined to be abnormal based on these databases. The information may be transmitted to the second monitoring / diagnosis apparatus 300B, and other processes may be performed by the second monitoring / diagnosis apparatus 300B. In this case, what is provided with at least the abnormality determination unit 111 may be used as the arithmetic processing unit 110 of the first diagnostic apparatus 300A.

  Further, as shown in FIG. 12, a first monitoring / diagnosis device 400A is arranged at a position adjacent to the gas turbine power generation facility 200, and the first monitoring / diagnosis device 400A is connected to the first monitoring / diagnosis device 400A via a router 240 and the Internet 241. The connected second monitoring and diagnosing device 400B is arranged at a position away from the gas turbine power generation facility 200, and the same processing is performed in parallel by the first monitoring and diagnosing device 400A and the second monitoring and diagnosing device 400B. Also good.

  In this case, the first monitoring and diagnosing device 400A and the second monitoring and diagnosing device 400B can adopt the same configuration as the monitoring and diagnosing device 100 shown in FIG. The sensor signal obtained from the gas turbine power generation facility 200 can be transmitted to the second monitoring and diagnosing device 400B.

1 is a schematic diagram of a monitoring and diagnosis apparatus 100. FIG. Schematic of the knowledge search screen 170. FIG. Schematic of the knowledge search result screen 180. FIG. Schematic of the abnormality determination database 121. FIG. 1 is a schematic diagram of an abnormality diagnosis database 122. FIG. Schematic of the past case database 123. FIG. Schematic of the knowledge database 124. Schematic diagram of an abnormal level determination database 125. FIG. Schematic which shows the usage example of the monitoring diagnostic apparatus. 5 is a flowchart showing a processing procedure in the monitoring diagnostic apparatus 100. Schematic which shows the modification of a monitoring diagnostic apparatus. Schematic which shows the modification of a monitoring diagnostic apparatus.

Explanation of symbols

100, 300 Monitoring / diagnosis apparatus 110 Operation processing unit 111 Abnormality determination unit 112 Past case search processing unit 113 Knowledge search processing unit 114 Influence degree evaluation processing unit 115 Abnormal level determination processing unit 116 Registration unit 120 Storage unit 121 Abnormality determination database 122 Abnormal diagnosis Database 123 Past case database 124 Knowledge database 125 Abnormal level judgment database

Claims (8)

A device monitoring and diagnosis device,
An abnormality diagnosis database storing state information representing a state when it is determined to be abnormal based on a process state quantity calculated from a detection signal of a sensor attached to a monitoring target device that is a monitoring target, the monitoring target A past case database storing device names of devices, information related to failures that occurred in the past in the monitored devices, and information related to handling the failures, and devices of non-monitored devices other than the monitored devices A storage unit that stores a name, a knowledge database that stores information related to the failure of the non-monitoring target device, and information related to the handling of the failure,
When it is determined that there is an abnormality based on the process state amount of the monitored device, the state information is read from the abnormality diagnosis database, and based on at least one of the device name of the monitored device and the state information, Searching the past case database, and when there is no corresponding past case, an arithmetic processing unit for searching the knowledge database,
A monitoring and diagnosis apparatus characterized by. The monitoring and diagnosis apparatus according to claim 1,
It has a communication unit that can connect to the Internet,
When there is no corresponding item in the knowledge database, the arithmetic processing unit uses a predetermined search search engine on the Internet based on at least one of the device name of the monitoring target device and the status information. Doing a search,
A monitoring and diagnosis apparatus characterized by. The monitoring and diagnosis apparatus according to claim 1,
In the knowledge database, information on the degree of influence when a part constituting the monitoring target device has a defect is stored corresponding to the item,
The arithmetic processing unit, when there is a corresponding item in the knowledge database, to perform control to output information on the degree of influence corresponding to the item in a predetermined format;
A monitoring and diagnosis apparatus characterized by. The monitoring and diagnosis apparatus according to claim 1,
In the storage unit, an abnormal level determination including a device name of the monitored device, information on a component of the monitored device, information on a life of the component, and information on a use start date of the component The database is remembered,
The arithmetic processing unit, from the abnormality level determination database, based on the device name of the monitored device, information on the component of the monitored device, information on the life of the component, and the start date of use of the component Information, and, based on the current date, identify a part that is likely to be damaged from the parts, and calculate the probability of damage of the device from the importance of the part,
A monitoring and diagnosis apparatus characterized by. An abnormality diagnosis database storing state information representing a state when an abnormality is determined based on a process state quantity calculated from a detection signal of a sensor attached to a monitoring target device to be monitored, the monitoring target device Device name, past case database storing information related to failures that occurred in the past in the monitored device, and information related to handling the failures, and device names of non-monitored devices other than the monitored device A monitoring diagnostic device that monitors devices in a monitoring diagnostic device comprising: a storage unit storing a knowledge database storing information related to failures in the non-monitoring target device and information relating to handling the failures. And
When it is determined that there is an abnormality based on the process state quantity of the monitored device, a reading process of reading the state information from the abnormality diagnosis database;
A past case search process for searching the past case database based on at least one of the device name of the monitoring target device and the state information;
In the past case search process, if there is no corresponding past case, a knowledge search process for searching the knowledge database based on at least one of the device name of the monitoring target device and the state information;
A monitoring diagnosis method comprising: The monitoring diagnosis method according to claim 5,
In the knowledge search process, if there is no corresponding item, an Internet that performs a search in the Internet using a predetermined search search engine based on at least one of the device name of the monitored device and the status information Having a search process,
A monitoring diagnosis method characterized by the above. The monitoring diagnosis method according to claim 5,
In the knowledge database, information on the degree of influence when a part constituting the monitoring target device has a defect is stored corresponding to the item,
In the knowledge search process, when there is a corresponding item, an impact output process for outputting information on the impact corresponding to the item in a predetermined format,
A monitoring diagnosis method characterized by the above. The monitoring diagnosis method according to claim 5,
The storage unit includes a device name of the monitored device, information on a component of the monitored device, information on a life of the component, and information on a start date of use of the component. The database is remembered,
Based on the device name of the monitored device, information on the component of the monitored device, information on the life of the component, and information on the start date of use of the component are read from the abnormality level determination database. Identifying a part having a high possibility of damage from the parts based on the current date, and providing an abnormality level determination process for calculating a failure probability of the equipment from the importance of the part;
A monitoring diagnosis method characterized by the above.

Images