Information processing device, information processing method, and information processing program
The information processing device generates new confirmation rules and builds an estimation model to accurately identify equipment abnormalities by replacing existing elements, addressing the issue of omitted rules in existing techniques.
Patent Information
- Application Number
- JP2024559048
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2023-05-08
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2043-05-08
AI Technical Summary
Existing techniques for estimating equipment abnormality fail to correctly identify the cause due to the omission of confirmation rules, making it difficult to generate comprehensive maintenance knowledge networks.
An information processing device that generates new confirmation rules by replacing existing confirmation target elements with new ones, calculates their contribution, and selects rules based on a prescribed condition to build an estimation model for accurate abnormality detection.
Prevents the failure of generating confirmation rules and enables correct estimation of equipment abnormalities, reducing response time to such events.
Smart Images

Figure 0007595823000001 
Figure 0007595823000002 
Figure 0007595823000003
Abstract
Description
[Technical field]
[0001] The present disclosure relates to a technique for estimating the cause of an abnormality in a device. [Background technology]
[0002] Patent Document 1 discloses a technique for estimating the cause of an abnormality in a device. In the technology of Patent Document 1, the cause of an abnormality corresponding to the model is estimated based on a network model common to all models (common maintenance knowledge network). More specifically, a confirmation rule related to the abnormality that has occurred is extracted from the confirmation rules set in the common maintenance knowledge network. The confirmation rule is a rule used when checking the state of the equipment to infer the cause of the abnormality. Then, a maintenance knowledge network corresponding to the abnormality is generated using the extracted confirmation rules. Next, confirmation rules corresponding to failures that do not occur in the model in which the abnormality has occurred are removed. As a result, a maintenance knowledge network corresponding to that model is generated. By using this maintenance knowledge network as a Bayesian network, the cause of the abnormality can be inferred. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] Patent Publication No. 2022-096546 Summary of the Invention [Problem to be solved by the invention]
[0004] In the technique of Patent Document 1, a designer needs to associate a confirmation rule with every anomaly in a common maintenance knowledge network. In other words, the technique of Patent Document 1 requires the designer to analyze the causal relationship between the anomaly and the anomaly cause for every anomaly, and to generate confirmation rules that reflect the analysis results. However, it is difficult for a designer to generate confirmation rules that cover all anomalies, and there is a possibility that some confirmation rules may be overlooked. If a confirmation rule is overlooked, the cause of the anomaly cannot be correctly inferred. As described above, the technique of Patent Document 1 has a problem in that a situation may arise in which the cause of an abnormality cannot be correctly estimated due to omission of generation of a confirmation rule.
[0005] A primary objective of the present disclosure is to solve such problems. Specifically, a primary objective of the present disclosure is to prevent the generation of confirmation rules from being overlooked and to enable the cause of an abnormality to be correctly estimated. [Means for solving the problem]
[0006] The information processing device according to the present disclosure includes: a rule generating unit that generates a plurality of new confirmation rules different from the existing confirmation rules by replacing existing confirmation target elements, which are at least any one of the confirmation target elements of the device and the confirmation target elements of the surrounding environment of the device, described in existing confirmation rules, which are existing confirmation rules used when confirming the state of the device to estimate the cause of an abnormality in the device, with new confirmation target elements, which are elements not described in the existing confirmation rules; a contribution degree calculation unit that calculates a contribution degree of each of the new confirmation rules to the estimation of the cause of the abnormality of the device; and a selection unit that selects, from the plurality of new confirmation rules, a new confirmation rule whose calculated contribution degree matches a specified selection condition. Effect of the Invention
[0007] According to the present disclosure, it is possible to prevent omission of generation of confirmation rules and correctly infer the cause of an abnormality. [Brief description of the drawings]
[0008] [Figure 1] FIG. 2 is a diagram showing an example of a functional configuration of the information processing device according to the first embodiment. [Diagram 2]FIG. 2 is a diagram showing an example of a hardware configuration of the information processing device according to the first embodiment. [Diagram 3] 5 is a flowchart showing an example of an operation in a preparation stage of the information processing device according to the first embodiment. [Figure 4] FIG. 4 is a diagram showing an example of existing confirmation rule information according to the first embodiment. [Diagram 5] FIG. 4 is a diagram showing an example of new confirmation rule information according to the first embodiment. [Figure 6] FIG. 4 is a diagram showing an example of a calculation result of a contribution degree according to the first embodiment. [Figure 7] FIG. 4 is a diagram showing an example of abnormal event information according to the first embodiment. [Figure 8] FIG. 4 is a diagram showing an example of analysis result information according to the first embodiment. [Figure 9] FIG. 4 is a diagram showing an example of extraction of analysis result information according to the first embodiment. [Figure 10] 4 is a flowchart showing an example of an operation at a utilization stage of the information processing device according to the first embodiment. [Figure 11] FIG. 4 is a diagram showing an example of measurement data according to the first embodiment. [Figure 12] FIG. 4 is a diagram showing an example of state check information according to the first embodiment. [Figure 13] FIG. 2 is a diagram showing an example of display information according to the first embodiment. [Figure 14] FIG. 11 is a diagram showing an example of display information according to the second embodiment. [Figure 15] FIG. 11 is a diagram showing an example of visualization method information according to the second embodiment. [Figure 16] FIG. 13 is a diagram showing an example of measurement data according to the third embodiment. [Figure 17] FIG. 13 is a diagram showing an example of abnormal event information according to the third embodiment. [Figure 18] FIG. 13 is a diagram showing an example of abnormal event information according to the third embodiment. [Figure 19] FIG. 13 is a diagram showing an example of analysis result information based on an error code in the third embodiment. [Figure 20] FIG. 13 is a diagram showing an example of abnormal event information according to the fourth embodiment. [Figure 21] FIG. 13 is a diagram showing the counting results of the number of occurrences according to the fourth embodiment. [Figure 22] FIG. 13 is a diagram showing an example of display information according to the fourth embodiment. [Figure 23] 13 is a flowchart showing an example of an operation in a preparation stage of the information processing device according to the fifth embodiment. [Figure 24] 20 is a flowchart showing an example of the operation in a preparation stage of the information processing device according to the sixth embodiment. [Diagram 25] FIG. 4 is a diagram showing an example of analysis result information according to the first embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0009] Hereinafter, the embodiments will be described with reference to the drawings. In the following description of the embodiments and in the drawings, the same reference numerals denote the same or corresponding parts.
[0010] Embodiment 1 ***Configuration Description*** FIG. 1 shows an example of a functional configuration of an information processing device 100 according to the present embodiment. FIG. 2 shows an example of a hardware configuration of the information processing device 100 according to the present embodiment.
[0011] The operation of information processing device 100 according to this embodiment is roughly divided into a preparation stage and an application stage. In the preparation stage, the information processing device 100 generates a confirmation rule. The confirmation rule is a rule used when confirming the state of the device 200 in order to estimate the cause of an abnormality in the device 200. The device 200 is, for example, an elevator, an air conditioner, or the like. Furthermore, in a preparation stage, the information processing device 100 generates an estimation model by using the confirmation rules. The estimation model is a trained model for estimating the cause of an abnormality in the device 200. In the utilization stage, when an abnormality occurs in the device 200, the information processing device 100 checks the state of the device using the check rule. The information processing device 100 also estimates the cause of the abnormality in the device 200 using the check result and the estimation model. Then, the information processing device 100 presents the estimation result to the user of the information processing device 100. The user of the information processing device 100 identifies the cause of the abnormality in the device 200 by referring to the estimation result. The user of the information processing device 100 is, for example, a designer of the confirmation rules, a maintenance worker of the device 200, or the like.
[0012] The information processing device 100 according to this embodiment is a computer. The operation procedure of the information processing device 100 corresponds to an information processing method, and the program that realizes the operation of the information processing device 100 corresponds to an information processing program.
[0013] As shown in FIG. 2, the information processing device 100 includes, as hardware, a processor 901, a main storage device 902, an auxiliary storage device 903, a communication device 904, and an input / output device 905. 1, the information processing device 100 includes, as its functional configuration, a rule generating unit 111, a rule information storage unit 112, a contribution degree calculating unit 113, etc. The functions of the rule generating unit 111, the contribution degree calculating unit 113, the selecting unit 114, the estimation model generating unit 116, the data receiving unit 121, the state checking unit 122, the cause estimating unit 123, and the display unit 124 shown in FIG. 1 are realized by, for example, a program. The auxiliary storage device 903 stores programs for implementing the functions of the rule generating unit 111, the contribution degree calculating unit 113, the selecting unit 114, and the like. These programs are loaded from the auxiliary storage device 903 to the main storage device 902. Then, the processor 901 executes these programs. FIG. 2 shows a schematic diagram of a state in which a processor 901 is executing a program that realizes the functions of the rule generating unit 111, the contribution degree calculating unit 113, the selecting unit 114, and the like. The rule information storage unit 112, the abnormal event information storage unit 115, and the estimation model storage unit 117 are realized by a main storage unit 902 and / or an auxiliary storage unit 903. The communication device 904 communicates with the device 200 . The input / output device 905 receives instructions from a user of the information processing device 100. In addition, the input / output device 905 displays various information to the user of the information processing device 100. The input / output device 905 is a mouse, a keyboard, a display, or the like.
[0014] 1, rule generation unit 111 uses existing confirmation rules to generate multiple confirmation rules that are different from the existing confirmation rules. In the following, the existing confirmation rules are referred to as existing confirmation rules. In addition, in the following, each of the multiple confirmation rules generated by rule generation unit 111 using the existing confirmation rules is referred to as a new confirmation rule. The process performed by rule generating unit 111 corresponds to a rule generation process.
[0015] The rule information storage unit 112 stores the confirmation rule information. The confirmation rule information includes a confirmation rule, the details of which will be described later. The rule information storage unit 112 stores existing check rule information and new check rule information as the check rule information. The existing check rule information includes an existing check rule. The new check rule information includes a new check rule.
[0016] The contribution degree calculation unit 113 calculates the contribution degree of each new confirmation rule to the estimation of the cause of the abnormality in the device 200. The contribution degree calculation unit 113 may also calculate the contribution degree of an existing confirmation rule. That is, the contribution degree calculation unit 113 may calculate the contribution degree of each of a plurality of new confirmation rules and existing confirmation rules. The contribution degree calculation unit 113 calculates, for example, the importance of the feature amount of Random Forest, the weight of regression analysis, and the like as the contribution degree. The process performed by the contribution degree calculation section 113 corresponds to a contribution degree calculation process.
[0017] The abnormal event information storage unit 115 stores abnormal event information. The abnormal event information indicates abnormal events that have occurred in the device 200 in the past. The abnormal event information will be described in detail later.
[0018] The selection unit 114 selects, from among the multiple new confirmation rules generated by the rule generation unit 111, a new confirmation rule whose contribution degree calculated by the contribution degree calculation unit 113 matches a specified selection condition. If the contribution degree calculation unit 113 also calculates the contribution degree of the existing confirmation rule, the selection unit 114 selects, from the multiple new confirmation rules and the existing confirmation rules, a confirmation rule whose contribution degree matches the selection condition. The confirmation rule selected by the selection unit 114 is called a selected confirmation rule. The process performed by the selection unit 114 corresponds to a selection process.
[0019] The estimation model generation unit 116 generates an estimation model using the selection confirmation rule and the abnormal event information.
[0020] The estimation model storage unit 117 stores the estimation model generated by the estimation model generation unit 116.
[0021] The data receiving unit 121 receives measurement data when an abnormality occurs in the device 200. The measurement data indicates a value measured by the device 200 or a value measured in the surrounding environment of the device 200. The measurement data will be described in detail later.
[0022] The status confirmation unit 122 confirms the status of the device 200 using the measurement data and the selection confirmation rule.
[0023] The cause estimation unit 123 estimates the cause of the abnormality in the device 200 by using the state confirmation result of the state confirmation unit 122 and the estimation model.
[0024] The display unit 124 displays the result of the status confirmation by the status confirmation unit 122 and the result of the estimation of the cause of the abnormality by the cause estimation unit 123.
[0025] ***Explanation of Operation*** FIG. 3 is a flowchart showing an example of the operation of the information processing device 100 in the preparation stage according to the present embodiment. An example of the operation of the information processing device 100 in the preparation stage will be described with reference to FIG.
[0026] In step S111, the rule generation unit 111 generates a new confirmation rule using an existing confirmation rule. The procedure for generating a new confirmation rule by rule generating unit 111 will be described below.
[0027] FIG. 4 shows an example of existing confirmation rule information 1121 including existing confirmation rules. A rule number, a check item, and a check rule are included in the existing check rule information 1121. The check rules included in the existing check rule information 1121 are existing check rules.
[0028] The rule number is an identification number of the confirmation rule. A check item is a check item among multiple check items that is the target of a check rule. A check item is an item that should be checked by the state checking unit 122, which will be described later, when an abnormality occurs in the device 200. In FIG. 4, the check items are abstracted as check item 1, check item 2, etc., but in actual operation, specific check items of the device 200 are listed. For example, a check item may be a check item as to whether "the signal value of the device 200 when an abnormality of the device 200 is detected is an appropriate value." Another example of a check item may be a check item as to whether "the order of changes in signal values of two or more signals of the device 200 is appropriate." The confirmation rule indicates details for realizing the confirmation targeted by the confirmation item. Check item 1 in FIG. 4 is a check item for "whether the signal value of device 200 is appropriate when an abnormality is detected in device 200." To realize the check targeted by check item 1, the confirmation rule corresponding to check item 1 describes a rule that reads "Is the value of signal A ON when an abnormality is detected?" "Signal A" is a signal used by device 200. Furthermore, the confirmation rules (existing confirmation rules) of the existing confirmation rule information 1121 indicate elements (confirmation target elements) that are targets of confirmation by the state confirmation unit 122. In the example of Fig. 4, "signal A" in the first row, "signal B" and "signal C" in the second row, and "signal D" and "signal E" in the third row are the confirmation target elements. 4, the signals used by device 200 are described in the existing confirmation rules as confirmation target elements. In addition, values (e.g., temperature, humidity, etc.) measured in the surrounding environment of device 200 may be described in the existing confirmation rules as confirmation target elements. Also, only values measured in the surrounding environment of device 200 may be described in the existing confirmation rules as confirmation target elements. The elements to be confirmed that are described in the existing confirmation rules are called existing confirmation elements. In the following, for the sake of simplicity, it is assumed that all existing confirmation target elements are signals used in the device 200.
[0029] The rule generation unit 111 replaces an existing confirmation target element with a new confirmation target element that is an element not described in the existing confirmation rule to generate a new confirmation rule. The rule generation unit 111 replaces an existing confirmation target element with a new confirmation target element in a brute force manner to generate a new confirmation rule. The rule information storage unit 112 stores confirmation target element information in which elements that can be used as new confirmation target elements are listed. In this embodiment, the confirmation target element information lists signal names of signals used in the device 200. The rule generation unit 111 generates new confirmation rules by using the signal names listed in the confirmation target element information as new confirmation target elements.
[0030] FIG. 5 shows an example of new confirmation rule information 1122 including a new confirmation rule. The new confirmation rule information 1122 also includes a rule number, a confirmation item, and a confirmation rule. The confirmation rule included in the new confirmation rule information 1122 is a new confirmation rule. The rule number indicates the relationship with the rule number of the existing confirmation rule from which the new confirmation rule was generated. Specifically, a new confirmation rule generated using an existing rule with rule number 1 is assigned a rule number with a suffix added to rule number 1, such as rule number 1-a. The items to be checked are the same as those shown in FIG. The confirmation rule in the new confirmation rule information 1122 (new confirmation rule) is a confirmation rule obtained by replacing an existing confirmation target element of the original existing confirmation rule with a new confirmation target element. For example, in the new confirmation rule of rule number 1-a, the existing confirmation target element "signal A" of the existing confirmation rule of rule number 1 is replaced with a new confirmation target element "signal B." Similarly, in the new confirmation rule of rule number 2-a, the existing confirmation target elements "signal B" and "signal C" of the existing confirmation rule of rule number 2 are replaced with new confirmation target elements "signal A" and "signal B". Similarly, in the new confirmation rule of rule number 3-a, the existing confirmation target elements "signal D" and "signal E" of the existing confirmation rule of rule number 3 are replaced with new confirmation target elements "signal A" and "signal B". 5, in order to clarify the relationship between the new confirmation rule and the existing confirmation rule, the rule number of the new confirmation rule includes the rule number of the existing confirmation rule from which it was generated. Alternatively, the rule number of the new confirmation rule may be a rule number that is unrelated to the rule number of the existing confirmation rule from which it was generated.
[0031] In this way, the rule generation unit 111 generates a new confirmation rule by replacing an existing confirmation target element with a new confirmation target element that is an element not described in the existing confirmation rule.
[0032] The rule generating unit 111 outputs existing confirmation rule information 1121 and new confirmation rule information 1122 to the contribution degree calculating unit 113 . In this embodiment, an example will be described in which contribution degree calculation unit 113 calculates the contribution degrees of an existing confirmation rule and a new confirmation rule. For this reason, rule generation unit 111 outputs existing confirmation rule information 1121 and new confirmation rule information 1122 to contribution degree calculation unit 113. When contribution degree calculation unit 113 calculates only the contribution degree of a new confirmation rule, rule generation unit 111 outputs only new confirmation rule information 1122 to contribution degree calculation unit 113.
[0033] 3, in step S112, the contribution degree calculation unit 113 calculates the contribution degree of the existing confirmation rule and the new confirmation rule. More specifically, the contribution degree calculation unit 113 calculates the contribution degree for each confirmation rule in the rule information storage unit 112 using Random Forest, Permutation Importance, or the like for the abnormal event information in the abnormal event information storage unit 115. The contribution degree calculation unit 113 uses the importance of the feature amount to calculate the contribution degree. In this embodiment, the feature amount is a confirmation rule. When the contribution degree calculation unit 113 uses Random Forest, the contribution degree calculation unit 113 can calculate the importance of the feature amount by referring to the statistics of the degree of reduction of the Gini impurity. Furthermore, the contribution calculation section 113 may calculate the importance of the feature amount using permutation importance or the like. Furthermore, the contribution degree calculation section 113 may calculate the contribution degree using another method. The contribution degree calculation unit 113 outputs existing confirmation rule information 1121, new confirmation rule information 1122, and the contribution degree for each confirmation rule to the selection unit 114. As described above, the abnormal event information indicates a past abnormal event that occurred in the device 200. The contribution degree calculation unit 113 acquires the abnormal event information from the abnormal event information storage unit 115.
[0034] FIG. 7 shows an example of abnormal event information. The abnormal event information is composed of an event number, a plurality of signal values, and an abnormality cause. The event number is the number of the abnormal event. The signal values are time-series transitions of the values of each signal measured after an abnormality occurs in the device 200. Each signal is a 1-bit signal. Furthermore, the signal value may include values other than a 1-bit signal, such as a voltage value. The cause of the abnormality is a cause of the abnormality specified by the user of the information processing device 100.
[0035] The contribution degree calculation unit 113 analyzes whether or not each selection confirmation rule applies for each event number of the abnormal event information, based on the signal value of the abnormal event information. Here, it is assumed that the contribution degree calculation unit 113 analyzes whether the existing confirmation rule and the new confirmation rule apply. Fig. 25 shows analysis result information describing the analysis results of the contribution degree calculation unit 113. "1", "1-a", "1-b", "2", "2-a" and "4-z" in Fig. 25 are rule numbers of the existing confirmation rule and the new confirmation rule. In the analysis result information in Fig. 25, a value of 1 means that the existing confirmation rule or the selected confirmation rule was applicable to the multiple signal values of each abnormal event. A value of 0 means that the existing confirmation rule or the selected confirmation rule was not applicable. For example, assume that the confirmation rule for "1" is "Is the value of signal A ON when an abnormality is detected?" If the contribution calculation unit 113 analyzes the signal value of signal A of event number 1 and finds that the value of signal A is ON, then a value of 1 is set for the combination of "event number 1" and "1" as shown in FIG.
[0036] In step S113, the selection unit 114 selects, from among the existing confirmation rules and the new confirmation rules, a confirmation rule whose contribution degree matches a specified selection condition. For example, the selection unit 114 selects a confirmation rule using a selection condition of "select a confirmation rule with a contribution degree equal to or greater than a threshold value." The selection unit 114 may select a confirmation rule using a selection condition such as "select a confirmation rule having a contribution degree within the top n ranks." The selection unit 114 may select a confirmation rule using a selection condition that "selects the confirmation rule with the highest degree of contribution for each rule number." FIG. 6 shows an example in which the selection unit 114 selects, as the selected confirmation rule, a confirmation rule having a contribution degree equal to or greater than a threshold value "0.60." The selection unit 114 outputs existing confirmation rule information 1121 and / or new confirmation rule information 1122 including the selected confirmation rule to the estimation model generation unit 116. Furthermore, the selection unit 114 stores in the rule information storage unit 112 existing confirmation rule information 1121 and / or new confirmation rule information 1122 including the selected confirmation rule.
[0037] In step S115, the estimation model generating unit 116 generates an estimation model. The estimation model generation unit 116 generates an estimation model using the existing confirmation rule information 1121 and / or new confirmation rule information 1122 acquired from the selection unit 114 and the abnormal event information.
[0038] The estimation model generation unit 116 analyzes whether or not each selection confirmation rule applies for each event number of the abnormal event information, based on the signal value of the abnormal event information. Here, it is assumed that the estimation model generating unit 116 analyzes whether or not each of the selection confirmation rules selected in FIG. 6 applies. Fig. 8 shows analysis result information describing the analysis results of the estimation model generation unit 116. "2-m", "3-f", "1", "4-g", "3-p" and "2-r" in Fig. 8 are the rule numbers of the selection confirmation rules selected in Fig. 6. In the analysis result information in Figure 8, a value of 1 means that the selective confirmation rule was applied to the multiple signal values of each abnormal event, and a value of 0 means that the selective confirmation rule was not applied. For example, assume that the selection confirmation rule for "2-m" is a rule that asks, "Did the value of signal G become OFF after the value of signal F became ON?" If the result of the estimation model generation unit 116 analyzing the signal values of signal F and signal G of event number 1 shows that the value of signal G becomes OFF after the value of signal F becomes ON, then a value of 1 is set for the combination of "event number 1" and "2-m" as shown in FIG.
[0039] The estimation model generation unit 116 learns the relationship between the selection confirmation rule and the cause of the abnormality, and generates an estimation model for inferring the cause of the abnormality. More specifically, the estimation model generation unit 116 uses statistical causal inference, which is a method of statistically inferring causal relationships between input data and output data, to infer selection confirmation rules that have causal relationships with abnormality causes and construct a graph. Then, the estimation model generation unit 116 applies a Bayesian network to this graph to generate an estimation model that infers failures. In this embodiment, one estimation model is generated, but the estimation model generating unit 116 may generate an estimation model for each cause of anomaly. More specifically, the estimation model generation unit 116 extracts records corresponding to each abnormality cause from the analysis result information, and then generates an estimation model by learning the extracted records. Fig. 9 shows an example in which the estimation model generation unit 116 extracts records of anomaly cause 1 from the analysis result information. For example, as shown in Fig. 9, if anomaly cause 1 is described in event number 1, event number 5, event number 11, event number 14, and event number 25 in the analysis result information, the estimation model generation unit 116 extracts these records as learning targets from the analysis result information. Then, the estimation model generation unit 116 learns the extracted records to generate an estimation model. The estimation model generation unit 116 generates estimation models for other anomaly causes in the same procedure.
[0040] The estimation model generation unit 116 stores the analysis result information and the generated estimation model in the estimation model storage unit 117.
[0041] Next, an example of the operation of the information processing device 100 in the utilization stage will be described with reference to FIG. FIG. 10 is a flowchart showing an example of the operation of the information processing device 100 in the utilization stage.
[0042] When an abnormality occurs in the device 200, the data receiving unit 121 periodically receives measurement data from the device 200 in step S121. The data receiving unit 121 receives measurement data measured by the device 200 for each of a plurality of confirmation items targeted by the selected confirmation rule. The data receiving unit 121 outputs the received measurement data to the state checking unit 122 . The data receiving unit 121 outputs, for example, as shown in FIG. 11, measurement data indicating the time transition of the value of each signal to the state checking unit 122.
[0043] Next, in step S122, the status confirmation unit 122 confirms the status of the device 200 using the measurement data and the selection confirmation rule. Specifically, the state checking unit 122 checks whether or not each selection check rule is satisfied with respect to the time transition of the value of each signal. 12 shows the status check information. The status check information indicates the result of the status check performed by the status check unit 122. In FIG. 12, "2-m", "3-f", "1", "4-g", "3-p" and "2-r" are rule numbers of the selection confirmation rules. In the state check result information in Fig. 12, a value of 1 means that the selective check rule was met for the time transition of the value of each signal, and a value of 0 means that the selective check rule was not met. The state confirmation unit 122 outputs the measurement data of FIG. 11 and the state confirmation result information of FIG.
[0044] Next, in step S123, the cause estimation unit 123 estimates the cause of the abnormality in the device 200. More specifically, the cause estimation unit 123 acquires the analysis result information (FIG. 8) from the estimation model storage unit 117. Then, the cause estimation unit 123 compares the state check result information acquired from the state check unit 122 with the analysis result information to select candidates for the cause of the abnormality. In the status check result information of FIG. 12, selection check rule 1, selection check rule 2-m, selection check rule 3-f, and selection check rule 4-g are applicable. Here, it is assumed that in a past abnormal event in which the cause of the abnormality was identified as abnormality cause 1 in Fig. 8, these selection confirmation rules were applicable with a high probability. Also, in a past abnormal event in which the cause of the abnormality was identified as abnormality cause 2 in Fig. 8, these selection confirmation rules were applicable with a medium probability. Furthermore, it is assumed that in a past abnormal event in which the cause of the abnormality was identified as abnormality cause 3, these selection confirmation rules were applicable with a low probability. In this case, the cause estimation unit 123 selects abnormality causes 1 to 3 as candidates for the abnormality cause. Then, the cause estimation unit 123 acquires the estimation model from the estimation model storage unit 117. Furthermore, the cause estimation unit 123 applies the measurement data of Fig. 11 to the estimation model to estimate the cause of the abnormality. The cause estimation unit 123 calculates, for example, the probability that each candidate is the cause of the anomaly as an estimation result. Here, for example, the cause estimation unit 123 calculates an estimation result such as "the probability that the cause of the anomaly is anomaly cause 1 is 80%, the probability that the cause is anomaly cause 2 is 16%, and the probability that the cause is anomaly cause 2 is 4%." Then, the cause estimation unit 123 outputs the estimation result of the cause of the abnormality to the display unit 124.
[0045] Finally, in step S124, the display unit 124 displays the result of the estimation of the cause of the abnormality on the display of the input / output device 905.
[0046] For example, the display unit 124 displays display information 1240 as exemplified in Fig. 13. Note that in Fig. 13, "cause of abnormality" is abbreviated to "cause". Display information 1240 shows an estimation result 1241 by the cause estimation unit 123. The estimation result 1241 shows candidates for the abnormality cause selected by the cause estimation unit 123 (referred to as cause candidates in FIG. 13) and their probabilities. Furthermore, the display information 1240 displays a correlation diagram 1242 between the candidates for the cause of anomaly (shown as Cause 1, Cause 2, and Cause 3 in FIG. 13) and the selection confirmation rule. In the correlation diagram 1242, the relationship between the candidate for the cause of anomaly with the highest probability (Cause 1 in FIG. 13) and the selection confirmation rule related to that candidate is highlighted. Furthermore, a list 1243 of selection confirmation rules associated with the candidate causes is displayed in the display information 1240. In the list 1243, the selection confirmation rules associated with the candidate cause of anomaly with the highest probability in the correlation diagram 1242 (cause 1 in FIG. 13) are listed.
[0047] Thereafter, the user of the information processing device 100 refers to the display information 1240 and investigates the cause of the abnormality in the device 200. When the user of the information processing device 100 is able to identify the cause of the abnormality in the device 200, the user adds the identified cause of the abnormality to the measurement data. Then, the user of the information processing device 100 uses the input / output device 905 to store the measurement data to which the cause of the abnormality has been added as abnormal event information in the abnormal event information storage unit 115.
[0048] In the above, the state confirmation unit 122 confirms the state of the device 200 by using the measurement data and the selection confirmation rule in step S122. Alternatively, the user of the information processing device 100 may manually check the state of the device 200 using the measurement data and the selection check rules. In this case, the user of the information processing device 100 inputs a confirmation result corresponding to the state confirmation information of FIG. The cause estimation unit 123 performs step S123 using the confirmation result input by the user of the information processing device 100.
[0049] ***Description of the Effects of the Embodiment*** In this embodiment, multiple new confirmation rules are generated by replacing existing confirmation target elements of existing confirmation rules with multiple new confirmation target elements in a brute force manner. Therefore, according to this embodiment, it is possible to prevent the generation of confirmation rules from being overlooked and to correctly infer the cause of an abnormality. Furthermore, by correctly estimating the cause of the abnormality, it is possible to reduce the time required for the user of the information processing device 100 to take action (to identify the cause of the abnormality) when an abnormality occurs.
[0050] Moreover, according to this embodiment, it is possible to generate a confirmation rule that is effective in estimating the cause of an abnormality that the user of the information processing device 100 is not aware of.
[0051] Furthermore, in this embodiment, a selection confirmation rule related to the estimation of the cause of the abnormality is displayed in the display information 1240. Therefore, according to this embodiment, the user of the information processing device 100 can refer to the displayed selection confirmation rule in the task of identifying the cause of the abnormality.
[0052] Embodiment 2 In this embodiment, the display unit 124 displays display information 1240 that is different from that in the first embodiment. In this embodiment, differences from the first embodiment will be mainly described. It should be noted that matters not explained below are the same as those in the first embodiment.
[0053] FIG. 14 shows an example of display information 1240 displayed by the display unit 124 according to the present embodiment. For drawing reasons, detailed illustration of the estimation result 1241 and the correlation diagram 1342 is omitted in Fig. 14. However, it is assumed that the estimation result 1241 and the correlation diagram 1342 similar to those in Fig. 13 are displayed in Fig. 14 as well.
[0054] In this embodiment, display unit 124 displays pass / fail result 1244 obtained by the status check by status check unit 122. Pass / fail result 1244 indicates whether or not each selection check rule was met in the status check by status check unit 122. "YES" indicates that the selection confirmation rule was met in the state confirmation by the cause estimation unit 123. On the other hand, "NO" indicates that the selection confirmation rule was not met. In the example of Fig. 14, it is assumed that the selection confirmation rule 3-f was determined not to be met in the state confirmation by the state confirmation unit 122. Furthermore, the display unit 124 may highlight the name of the signal that caused the result "No." In the example of Fig. 14, the reason why the selection confirmation rule 3-f was determined not to apply was "signal R," so the display unit 124 highlights "signal R." In this embodiment, cause estimation unit 123 outputs status confirmation information (FIG. 12) in addition to the estimation result to display unit 124. In this embodiment, it is assumed that the value 0 is set for selection confirmation rule 3-f in the status confirmation information (FIG. 12). Display unit 124 displays pass / fail result 1244 according to the value in the status confirmation information.
[0055] Moreover, in this embodiment, the display unit 124 visualizes time transitions 1245 of the signal values of the signals that are the confirmation target elements of the selected confirmation rule. Specifically, in selected confirmation rule 1, signal A is the confirmation target element, so the time transitions of the value of signal A in the measurement data are visualized. In selected confirmation rule 2-m, signals F and G are the confirmation target elements, so the time transitions of the values of signals F and G in the measurement data are visualized. In selected confirmation rule 3-f, signals E and R are the confirmation targets, so the time transitions of the values of signals E and R in the measurement data are visualized. The display unit 124 may change the visualization method for each signal. As shown in Fig. 15, the display unit 124 holds visualization method information in which a visualization method is specified for each signal, and visualizes the signal value in the method specified in the visualization method information. For drawing reasons, the row of selection confirmation rule 4-g shown in Fig. 13 is omitted in Fig. 14. However, also in Fig. 14, for selection confirmation rule 4-g, the signal value is visualized by the method specified in the visualization method information.
[0056] As described above, in this embodiment, the pass / fail result of each selection confirmation rule is displayed in the display information 1240. Also, in this embodiment, the display information 1240 visualizes the time transition of the value of the signal that is the confirmation target element of each selection confirmation rule. As described above, the user of the information processing device 100 can use the display information 1240 according to this embodiment to identify the cause of an abnormality. Furthermore, the user of the information processing device 100 can trust the estimation result of the cause estimation unit 123 by referring to the display information 1240 according to this embodiment.
[0057] Embodiment 3 In this embodiment, the estimation model generation unit 116 generates an estimation model in a unit different from that in the first embodiment. In the first embodiment, the estimation model generation unit 116 generates an estimation model in a unit of anomaly cause. In this embodiment, the estimation model generation unit 116 generates an estimation model in a unit of an error code. In this embodiment, differences from the first embodiment will be mainly described. It should be noted that matters not explained below are the same as those in the first embodiment.
[0058] FIG. 16 shows an example of measurement data according to this embodiment. In FIG. 16, an error code has been added compared to the measurement data in FIG. The error code is added to the measurement data by the device 200 when an abnormality occurs in the device 200. The error code indicates the type of abnormality recognized by the device 200.
[0059] FIG. 17 shows an example of abnormal event information according to this embodiment. In this embodiment, the abnormal event information storage unit 115 stores the abnormal event information illustrated in FIG. In FIG. 17, an error code has been added compared to the abnormal event information in FIG. In this embodiment as well, when the display information 1240 is displayed on the display unit 124, the user of the information processing device 100 identifies the cause of the abnormality in the device 200 and adds the identified cause of the abnormality to the measurement data in Fig. 16. Then, the user of the information processing device 100 stores the measurement data to which the cause of the abnormality has been added as abnormal event information in the abnormal event information storage unit 115 using the input / output device 905. As a result, the abnormal event information storage unit 115 stores the abnormal event information illustrated in FIG. In the example of Fig. 17, error code α was set in the measurement data when abnormal events of event number 1 and event number 5 occurred. Furthermore, error code α was set in the measurement data when an abnormal event of event number 2 occurred. Furthermore, error code β was set in the measurement data when an abnormal event of event number 3 occurred. Furthermore, error code β was set in the measurement data when an abnormal event of event number 4 occurred. Then, through an analysis by the user of the information processing device 100, it was determined that the abnormal events of event number 1 and event number 5 were caused by abnormality cause 1. Furthermore, it was determined that the abnormal event of event number 2 was caused by abnormality cause 2. Similarly, it was determined that the abnormal event of event number 3 was caused by abnormality cause 3. Furthermore, it was determined that the abnormal event of event number 4 was caused by abnormality cause 4. 17, in the abnormal event information according to the present embodiment, an error code α is associated with the abnormal events of event number 1, event number 2, and event number 5. Also, an error code β is associated with the abnormal events of event number 3 and event number 4.
[0060] Note that different error codes may be set even for abnormal events with the same abnormality cause. Figure 18 shows an example in which error code α is set to the measurement data when an abnormal event with event number 1 occurs, and error code γ is set to the measurement data when an abnormal event with event number 5 occurs. Subsequent analysis reveals that both the abnormal events with event numbers 1 and 5 are caused by abnormality cause 1, but in the abnormal event information, the abnormal events with event numbers 1 and 5 are associated with different error codes.
[0061] In this embodiment, the contribution calculation unit 113 applies the confirmation rule to the abnormal event with the error code α in the abnormal event information (FIG. 17) to obtain analysis result information. For example, the analysis result information obtained by applying the confirmation rule to the abnormal event with the error code α is shown in the upper part of FIG. Next, the contribution degree calculation unit 113 uses the analysis result information to calculate the contribution degree of each confirmation rule when estimating the cause of the abnormality of the error code α. Next, the selection unit 114 selects a confirmation rule having a high degree of contribution for the error code α.
[0062] Similarly, the contribution calculation unit 113 applies the confirmation rule to the abnormal event with the error code β in the abnormal event information (FIG. 17) to obtain analysis result information. For example, the analysis result information obtained by applying the confirmation rule to the abnormal event with the error code β is shown in the lower part of FIG. Next, the contribution degree calculation unit 113 uses the analysis result information to calculate the contribution degree of each confirmation rule when estimating the cause of the abnormality of the error code β. Next, the selection unit 114 selects a confirmation rule with a high degree of contribution for the error code β. The selection unit 114 selects a confirmation rule for the error code β that is different from the confirmation rule for the error code α.
[0063] Thereafter, the estimation model generating unit 116 generates an estimation model for estimating the cause of the abnormality for each error code. 19, for the error code α, the estimation model generation unit 116 learns the confirmation rules with rule numbers "2-m", "3-f", "1", "4-g", "3-p", and "2-r" to generate an estimation model. Similarly, for the error code β, the estimation model generation unit 116 learns the confirmation rules with rule numbers "3-c", "2-r", "4", "1-b", "2", and "4-h" to generate an estimation model.
[0064] Furthermore, when an abnormality occurs in the device 200, the cause estimation unit 123 estimates the cause of the abnormality using an estimation model corresponding to an error code added to the measurement data. For example, when an error code α is added to the measurement data, the cause estimation unit 123 estimates the cause of the abnormality using an estimation model corresponding to the error code α.
[0065] As described above, in this embodiment, an estimation model is generated for each error code, so that the cause of an abnormality can be accurately estimated using an appropriate estimation model in accordance with the error code at the time of the occurrence of the abnormality.
[0066] Embodiment 4 In this embodiment, an example will be described in which the display unit 124 displays parts that are frequently replaced due to the cause of the abnormality estimated by the cause estimation unit 123. In this embodiment, differences from the first embodiment will be mainly described. It should be noted that matters not explained below are the same as those in the first embodiment.
[0067] FIG. 20 shows an example of abnormal event information according to this embodiment. In FIG. 20, a column for replacement parts is added to the abnormal event information in FIG. In this embodiment, when the display information 1240 is displayed on the display unit 124, the user of the information processing device 100 identifies the cause of the abnormality in the device 200 and determines whether or not a part of the device 200 needs to be replaced. If a part needs to be replaced, the user adds the identified cause of the abnormality and the name of the part that needs to be replaced to the measurement data. The user of the information processing device 100 then uses the input / output device 905 to store the measurement data to which the cause of the abnormality and the name of the part have been added as abnormal event information in the abnormal event information storage unit 115. As a result, the abnormal event information according to this embodiment indicates parts that have been replaced due to abnormalities that have occurred in the past, as shown in FIG. In this embodiment, the abnormal event information may also include an error code as in the third embodiment.
[0068] In the present embodiment, the cause estimating unit 123 estimates the cause of the abnormality in step S123 of FIG. 10, and then determines a part that is recommended for replacement. For example, assume that the cause estimation unit 123 calculates an estimation result that "the probability that the cause of the anomaly is anomaly cause 1 is 80%, the probability that the cause is anomaly cause 2 is 16%, and the probability that the cause is anomaly cause 2 is 4%," as described above. In this case, the cause estimation unit 123 counts the occurrence frequency of replacement parts for anomaly cause 1, which is the anomaly cause with the highest probability, as shown in FIG. Then, the display unit 124 displays the occurrence count of the replacement part for the abnormality cause 1 as the replacement count 1246 in association with the abnormality cause 1 in the estimation result 1241, as shown in FIG.
[0069] The user of the information processing device 100 refers to the display information 1240 and considers whether or not the current abnormality in the device 200 requires replacement of the part indicated in the replacement number 1246 .
[0070] In this embodiment, an example has been shown in which the number of replacements 1246 is displayed only for abnormality cause 1, which is the most probable abnormality cause. The number of replacements 1246 may also be displayed for abnormality cause 2 and abnormality cause 3, which are less probable abnormality causes.
[0071] As described above, in this embodiment, the number of parts replaced due to past abnormalities is displayed. Therefore, according to this embodiment, the user of the information processing device 100 can properly know the parts for which replacement should be considered.
[0072] Embodiment 5. In this embodiment, the calculation of the contribution degree by the contribution degree calculation section 113 and the selection of the selection confirmation rule by the selection section 114 are performed multiple times. In this embodiment, differences from the first embodiment will be mainly described. It should be noted that matters not explained below are the same as those in the first embodiment.
[0073] FIG. 23 is a flowchart showing an example of the operation of the information processing device 100 in the preparation stage according to the present embodiment.
[0074] In FIG. 23, step S111 is similar to that shown in FIG. In this embodiment, steps S112 and S113 are repeated until a prescribed number of repetitions is reached (YES in step S115). For example, if the number of repetitions is two, the contribution calculation unit 113 and the selection unit 114 operate as follows.
[0075] In the first step S112, the contribution degree calculation unit 113 calculates the contribution degree of each of the existing confirmation rule information 1121 and the new confirmation rule information 1122, for example, by using RandomForest. Then, in the first step S113, the selection unit 114 selects a plurality of selected confirmation rules based on the contribution degrees of the first step S112, and outputs the plurality of selected confirmation rules to the contribution degree calculation unit 113. In the second step S112, the contribution degree calculation unit 113 calculates the contribution degree of each of the multiple selection confirmation rules using, for example, permutation importance. Then, in the second step S113, the selection unit 114 selects a new selection confirmation rule from the multiple selection confirmation rules based on the contribution degree in the second step S112. The selection unit 114 outputs the new selection confirmation rule to the estimation model generation unit 116. The method of selecting the selection confirmation rule by the selection unit 114 is the same as that described in the first embodiment. Also, in FIG. 23, step S114 is the same as that shown in FIG.
[0076] In this manner, in this embodiment, the calculation of the contribution degree and the selection of the selected confirmation rule are repeated multiple times, so that the embodiment can select a more effective confirmation rule for estimating the cause of the abnormality.
[0077] Embodiment 6 In this embodiment, the contribution degree calculation unit 113 classifies the existing confirmation rules and multiple new confirmation rules into multiple groups and calculates the contribution degree for each group. Then, the selection unit 114 selects, for each group, a confirmation rule whose contribution degree matches a selection condition as a selected confirmation rule. In this embodiment, differences from the first embodiment will be mainly described. It should be noted that matters not explained below are the same as those in the first embodiment.
[0078] FIG. 24 is a flowchart showing an example of the operation of the information processing device 100 in the preparation stage according to the present embodiment.
[0079] In FIG. 24, step S111 is similar to that shown in FIG. In this embodiment, in step S116, the contribution degree calculation unit 113 classifies the multiple new confirmation rules and existing confirmation rules generated in step S111 into multiple groups. The contribution degree calculation unit 113 classifies the multiple new confirmation rules and existing confirmation rules into groups randomly, for example. The contribution degree calculation unit 113 may classify the multiple new confirmation rules and existing confirmation rules into groups by any method. Next, in step S1121, the contribution degree calculation section 113 calculates, for each group, the contribution degree of each of the confirmation rules included in the group. The method of calculating the contribution degree by contribution degree calculation section 113 is similar to that described in the first embodiment. Next, in step S1131, the selection unit 114 selects, for each group, a selection confirmation rule from among the confirmation rules included in the group, based on the contribution degree calculated in step S1121. The method of selecting the selection confirmation rule by the selection unit 114 is the same as that described in the first embodiment. Finally, in step S114, the estimation model generation unit 116 generates an estimation model using the selection confirmation rule for each group. The method of generating the estimation model by the estimation model generation unit 116 is similar to that described in the first embodiment.
[0080] In this embodiment, the confirmation rules are classified into a plurality of groups, the contribution degree is calculated for each group, and a selected confirmation rule is selected for each group. Therefore, even if a large number of new confirmation rules are generated, the selected confirmation rule can be efficiently selected.
[0081] Although the first to sixth embodiments have been described above, two or more of these embodiments may be combined for implementation. Alternatively, one of these embodiments may be partially implemented. Alternatively, two or more of these embodiments may be partially combined and implemented. Furthermore, the configurations and procedures described in these embodiments may be modified as necessary.
[0082] ***Additional Hardware Configuration Information*** Finally, a supplementary explanation of the hardware configuration of the information processing device 100 will be given. The processor 901 shown in FIG. 2 is an integrated circuit (IC) that performs processing. The processor 901 is a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or the like. The main storage device 902 shown in FIG. 2 is a RAM (Random Access Memory). The auxiliary storage device 903 shown in FIG. 2 is a ROM (Read Only Memory), a flash memory, a HDD (Hard Disk Drive), or the like. The communication device 904 shown in FIG. 2 is an electronic circuit that performs data communication processing. The communication device 904 is, for example, a communication chip or a NIC (Network Interface Card).
[0083] The auxiliary storage device 903 also stores an OS (Operating System). At least a part of the OS is executed by the processor 901 . The processor 901 executes at least a part of the OS, and executes programs that realize the functions of the rule generating unit 111, the contribution degree calculating unit 113, the selecting unit 114, and the like. The processor 901 executes the OS, which performs task management, memory management, file management, communication control, and the like. In addition, at least one of information, data, signal values, and variable values indicating the results of processing by the rule generation unit 111, contribution calculation unit 113, selection unit 114, etc. are stored in at least one of the main memory device 902, the auxiliary memory device 903, a register within the processor 901, and a cache memory. Furthermore, the programs for realizing the functions of the rule generating unit 111, the contribution degree calculating unit 113, the selecting unit 114, etc. may be stored in a portable recording medium such as a magnetic disk, a flexible disk, an optical disk, a compact disk, a Blu-ray (registered trademark) disk, a DVD, etc. Then, the portable recording medium storing the programs for realizing the functions of the rule generating unit 111, the contribution degree calculating unit 113, the selecting unit 114, etc. may be distributed.
[0084] Furthermore, at least any one of the "parts" such as the rule generating part 111, the contribution calculating part 113, and the selecting part 114 may be read as "circuit" or "process" or "step" or "procedure" or "processing" or "circuitry". The information processing device 100 may be realized by a processing circuit. The processing circuit is, for example, a logic integrated circuit (IC), a gate array (GA), an application specific integrated circuit (ASIC), or a field programmable gate array (FPGA). In this case, the rule generating unit 111, the contribution calculating unit 113, the selecting unit 114, etc. are each realized as part of a processing circuit. In this specification, the higher-level concept of a processor and a processing circuit is called "processing circuitry." That is, a processor and a processing circuit are each specific examples of "processing circuitry." [Explanation of symbols]
[0085] 100 information processing device, 111 rule generation unit, 112 rule information storage unit, 113 contribution degree calculation unit, 114 selection unit, 115 abnormal event information storage unit, 116 estimation model generation unit, 117 estimation model storage unit, 121 data receiving unit, 122 status confirmation unit, 123 cause estimation unit, 124 display unit, 200 device, 901 processor, 902 main memory device, 903 auxiliary memory device, 904 communication device, 1121 existing confirmation rule information, 1122 new confirmation rule information, 1240 display information.
Claims
1. a rule generating unit that generates a plurality of new confirmation rules different from the existing confirmation rules by replacing existing confirmation target elements, which are at least any one of the confirmation target elements of the device and the confirmation target elements of the surrounding environment of the device, described in existing confirmation rules, which are existing confirmation rules used when confirming the state of the device to estimate the cause of an abnormality in the device, with new confirmation target elements, which are elements not described in the existing confirmation rules; a contribution degree calculation unit that calculates a contribution degree of each of the new confirmation rules to the estimation of the cause of the abnormality of the device; and a selection unit for selecting, from the plurality of new confirmation rules, a new confirmation rule whose calculated contribution degree meets a specified selection condition.
2. The contribution degree calculation unit Calculating the contribution of each of the plurality of new confirmation rules and the existing confirmation rules; The selection unit is The information processing apparatus according to claim 1 , wherein a confirmation rule whose calculated contribution rate matches the selection condition is selected from the plurality of new confirmation rules and the existing confirmation rules.
3. The contribution degree calculation unit Calculating a new contribution degree for each of the new confirmation rules selected by the selection unit; The selection unit is The information processing apparatus according to claim 1 , wherein a new confirmation rule whose new contribution degree matches the selection condition is selected from among the new confirmation rules whose new contribution degrees have been calculated by the contribution degree calculation section.
4. The contribution degree calculation unit Calculating a new contribution degree for each of the confirmation rules selected by the selection unit; The selection unit is The information processing apparatus according to claim 2 , further comprising: selecting a confirmation rule whose new contribution degree matches the selection condition from among the confirmation rules whose new contribution degrees have been calculated by the contribution degree calculation unit.
5. The contribution degree calculation unit classifying the plurality of new confirmation rules into a plurality of groups, and calculating the contribution degree of each new confirmation rule included in each group; The selection unit is The information processing apparatus according to claim 1 , wherein a new confirmation rule whose calculated contribution rate matches the selection condition is selected for each group.
6. The rule generating unit 2. The information processing apparatus according to claim 1, wherein a plurality of new confirmation rules are generated by replacing a plurality of existing confirmation target elements of a plurality of existing confirmation rules with a plurality of new confirmation target elements in a brute force manner.
7. The information processing device further comprises: The information processing apparatus according to claim 1 , further comprising an estimation model generating unit that generates an estimation model for estimating a cause of an abnormality in the device by using a selected confirmation rule that is the confirmation rule selected by the selection unit.
8. The estimation model generation unit The information processing device according to claim 7 , wherein the estimation model is generated for each abnormality cause using the selection confirmation rule and abnormal event information in which a past abnormal event of the device is indicated in association with one of a plurality of abnormality causes.
9. The estimation model generation unit The information processing device according to claim 7 , wherein the estimation model is generated for each error code using the selection confirmation rule and abnormal event information in which a past abnormal event of the device is indicated in association with one of a plurality of error codes.
10. The information processing device further comprises: a status confirmation unit that confirms a status of the device using the selection confirmation rule when an abnormality occurs in the device; The information processing apparatus according to claim 7 , further comprising a cause estimation unit that estimates a cause of the abnormality in the device by using a result of the confirmation by the state confirmation unit and the estimation model.
11. The information processing device further comprises:
8. The information processing device according to claim 7, further comprising a cause estimation unit that, when an abnormality occurs in the equipment, estimates a cause of the abnormality in the equipment using a confirmation result obtained by manually confirming the status of the equipment using the selected confirmation rule and the estimation model.
12. The information processing device further comprises: The information processing apparatus according to claim 10 , further comprising a display unit that displays a result of the confirmation by the state confirmation unit and a result of the estimation by the cause estimation unit.
13. The display unit is The information processing device according to claim 12 , wherein the result of the confirmation by the state confirmation unit is displayed as whether the selected confirmation rule is correct or not.
14. The display unit is The information processing apparatus according to claim 12 , wherein the confirmation target elements described in the selection confirmation rule are visualized.
15. The display unit is The information processing apparatus according to claim 12 , wherein a part replaced due to a past abnormality caused by the abnormality cause estimated by the cause estimation unit is displayed.
16. a computer replaces existing confirmation target elements, which are at least any one of the confirmation target elements of the device and the confirmation target elements of the surrounding environment of the device, described in an existing confirmation rule, which is an existing confirmation rule used when confirming the state of the device to estimate the cause of an abnormality in the device, with new confirmation target elements, which are elements not described in the existing confirmation rule, thereby generating a plurality of confirmation rules different from the existing confirmation rules as a plurality of new confirmation rules; the computer calculates a contribution of each of the new confirmation rules to an estimation of a cause of an abnormality in the device, The information processing method in which the computer selects, from the plurality of new confirmation rules, a new confirmation rule whose calculated contribution rate meets a specified selection condition.
17. a rule generation process for generating multiple new confirmation rules that are different from the existing confirmation rules by replacing existing confirmation target elements, which are at least any one of the confirmation target elements of the device and the confirmation target elements of the device's surrounding environment, described in existing confirmation rules, which are existing confirmation rules used when confirming the state of the device to estimate the cause of the device's abnormality, with new confirmation target elements, which are elements not described in the existing confirmation rules; a contribution degree calculation process for calculating a contribution degree of each of the new confirmation rules to the estimation of the cause of the abnormality of the device; and a selection process for selecting, from the plurality of new confirmation rules, a new confirmation rule whose calculated contribution degree meets a specified selection condition.
Citation Information
Patent Citations
Diagnosis supporting system, diagnosis supporting method and diagnosis supporting program
JP2005050115A
Root cause analysis and automation using machine learning
US20200382361A1
Cause estimation system and cause estimation method
JP2022096546A
