Power plant maintenance management apparatus, maintenance management method and maintenance management system

Abstract

To provide a power plant maintenance management technique which improves usability for users.SOLUTION: A power plant maintenance management apparatus comprises: an action plan creation unit 1107 for creating at least one action plan corresponding to a failure of a power plant; an action plan evaluation unit 1108 for evaluating and recommending the created action plan; and an action plan arrangement unit 1109 for arranging the recommended action plan.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a maintenance management device, a maintenance management method, and a maintenance management system for a power generation plant.

Background Art

[0002] For example, in power generation plants such as wind power plants, solar power plants, hydroelectric power plants, geothermal power plants, thermal power plants, and nuclear power plants, since they are large-scale systems in which a large number of devices act complexly, there are multiple stakeholders such as power generation companies, power transmission and distribution companies, power retail companies, equipment companies, and maintenance management service companies. A technique for detecting signs of failures that may occur in a large-scale system and proposing countermeasures is known as FMEA (Failure Mode and Effects Analysis).

[0003] Although it is not a technique related to a maintenance management device for a power generation plant, a technique for selecting a cost-effective measure from measures for solving regional problems has been proposed (Patent Document 1). Similarly, although it is not a technique related to a maintenance management device for a power generation plant, a technique for maintaining the safety of a communication network during a disaster or the like has also been proposed (Patent Document 2).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] In the prior art, it is possible to detect a failure of a power generation plant and output a corresponding countermeasure (action plan), but it is not possible to appropriately evaluate the action plan, resulting in poor usability for the user.

[0006] The present invention has been made in view of the above problems, and provides a maintenance management device, a maintenance management method, and a maintenance management system for a power generation plant that can improve usability for the user.

Means for Solving the Problems

[0007] To solve the above problems, a maintenance management device for a power generation plant according to one aspect of the present invention includes an action plan generation unit that generates at least one action plan for coping with a failure of the power generation plant, an action plan evaluation unit that evaluates and recommends the generated action plan, and an action plan sorting unit that sorts the recommended action plan.

Effects of the Invention

[0008] According to the present invention, it is possible to evaluate, recommend, and sort an action plan for coping with a failure of a power generation plant.

Brief Description of the Drawings

[0009]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28

Figure 29

Figure 30

Figure 31

Figure 32

Figure 33

Figure 34

Figure 35

Figure 36

Figure 37

Figure 38

Figure 39

Figure 40

Mode for Carrying Out the Invention

[0010] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the present disclosure, when a failure is detected in a power generation plant including a plurality of assets, an action plan set in advance for each stakeholder is generated. The action plans are ranked for each stakeholder, and similar action plans are integrated and organized into one. The organized action plans are provided to the user through a GUI (Graphical User Interface). The user's selection is utilized to update the failure model and the action plan recommendation engine. This description shows one example included in the present disclosure and should not be used to narrow the scope of the present invention.

[0011] Here, an action plan is information including actions, costs, and results. An action plan and an action are strictly distinguishable. However, hereinafter, there may be cases where an action plan and an action are not strictly distinguished.

Example

[0012] The first embodiment will be described with reference to FIGS. 1 to 40. FIG. 1 shows an overall overview of the maintenance management system 1. The maintenance management system 1 includes, for example, an action plan generation unit 1107, an action plan ranking unit 1108, an action plan arrangement unit 1109, an asset model server 1110, and a user verification unit 111. These system configurations will be described later with reference to FIG. 2.

[0013] The action plan generation unit 1107 has a function of generating an action plan corresponding to a failure of a power generation plant for each stakeholder. The action plan generation unit 1107 includes, for example, a function F1a that outputs actions, costs, and results for each failure mode, a human model 012, a function F1b that generates an action plan, and a list 002 of the generated action plans.

[0014] The action plan ranking unit 1108 is an example of an "action plan evaluation unit" and has a function of evaluating and recommending the generated action plans. The action plan ranking unit 1108 includes, for example, an action plan recommendation engine F02, a function 005 that inputs model serving to the action plan recommendation engine F02, a weight 008, and a ranked action plan 006.

[0015] The action plan arrangement unit 1109 has a function of arranging action plans. The action plan arrangement unit 1109 includes, for example, an exclusion filter F03, a constraint list 003, an action plan 019, a function F04 that simplifies the action plan, and a simplified action plan 013.

[0016] The user verification unit 1111 has a function for verifying the organized action plan with the first user (energy vendor). The user verification unit 1111 includes, for example, functions F5a and F5b for providing a GUI, function 014 for saving the output of the GUI, and a function for accepting selections by the first user.

[0017] The asset model server 1110 shown in FIG. 1 does not have to be part of the maintenance management system 1. The action plan generation unit 1107 only needs to be able to obtain the necessary information from the asset model server 1110.

[0018] FIG. 2 is a configuration diagram of the maintenance management system 1. The maintenance management system 1 will be described from the perspective of its system configuration. The maintenance management system 1 includes, for example, a maintenance management device 1105, a power generation plant 1104, a digital twin system 1106, and computer terminals 1101(1) and 1101(2), and these devices are communicably connected via a communication network 1108. Specifically, the maintenance management system 1 only needs to be able to obtain the necessary information from the sensor group 11041 of the power generation plant 1104 and the digital twin system 1106, and it does not have to include the power generation plant 1104 and the digital twin system 1106 within the maintenance management system 1.

[0019] The power generation plant 1104 is, for example, a wind power plant, a solar thermal power plant, a hydroelectric power plant, a thermal power plant, etc. The power generation plant 1104 includes, for example, a sensor group 11041 and a physical asset 11042. The physical asset 11042 is various facilities, devices, parts, etc. that make up the power generation plant 1104. The sensor group 11041 detects the state of the physical asset 11042 and outputs a signal. The sensor group 11041 includes, for example, a photoelectric switch, a mechanical switch, an image sensor, a laser sensor, a temperature sensor, a pressure sensor, an anemometer, an illuminance sensor, an ultrasonic sensor, etc.

[0020] The digital twin system 1106 includes, for example, an asset model server 1110 and a failure model 015. The digital twin system 1106 generates an asset model corresponding to the physical asset 11042 of the power plant 1104 and reflects the data detected by the sensor group 11041 in the asset model. The digital twin system 1106 can predict the occurrence of failures and the like from the state of the asset model.

[0021] At least one computer terminal 1101 can access the maintenance management device 1105. The first computer terminal 1101(1) is used by an energy vendor who can be called the first user. The second computer terminal 1101(2) is used by a second user who can be called a technology provider. When not distinguishing between both computer terminals 1101(1) and 1101(2), they may be referred to as the computer terminal 1101.

[0022] Here, the energy vendor is an organization that operates the power plant 1104. The technology provider is an organization that provides technologies such as devices, equipment, and software to the power plant 1104. In this embodiment, as will be described later, an action plan is generated according to the KPI for each technology provider, but it is the energy vendor who determines the action plan to be executed from among the generated action plans. The model of the power plant in the virtual space does not necessarily faithfully reproduce all of the actual power plant 1104, and furthermore, there are also non-physical circumstances that cannot be detected by the sensor group. Therefore, the energy vendor, who is the ultimate responsible person for the power plant 1104, finally determines the action plan to be executed. Note that the energy vendor may seek the opinions of each technology provider when finally determining the action plan to be adopted.

[0023] The computer terminal 1101 includes a GUI_1102 such as a web browser, for example. The user can input and output information to and from the maintenance management device 1105 via the GUI_1105. An example of the GUI screen will be described later.

[0024] The maintenance management device 1105 is a computer that manages the maintenance of the physical assets 11042 of the power plant 1104. The maintenance management device 1105 includes, for example, functional modules such as an action plan generation module 1107, an action plan ranking module 1108, an action plan sorting module 1109, and a user verification module 1111, and a data set described later. The functional modules 1107, 1108, 1109, 1111 correspond to the functional units 1107, 1108, 1109, 1111 shown in FIG. 1.

[0025] The data set possessed by the maintenance management device 1105 will be described. The maintenance management device 1105 includes, for example, a human model 012, an action plan list 002, a model serving input 005, a weight 008, a ranked action plan 006, an input to the filter 020, an output from the filter 019, a simplified action plan 013, and an output from the GUI 014.

[0026] Details of each dataset will be described later. Briefly, the human model 012 is a model that presents the relationship between assets, failures, actions, costs, and results to the user. The action plan list 002 is a list of action plans generated for each stakeholder. The model serving input 005 is data input for deploying and operating the action recommendation engine F02, which is a machine learning model, in a production environment. The weight 008 is data used when updating the failure model 015 and the action recommendation engine F02. The ranked action plan 006 is an action plan obtained by ranking the action plans generated for each stakeholder. The additional input to the filter 020 is an additional condition input to the filter F03 for excluding action plans that violate the constraints among the ranked action plans. The filter output 019 is data (action plans) output from the exclusion filter F03. The simplified action plan 013 is an action plan in which similar action plans are integrated into one and organized. The GUI 014 is data provided to the user via the GUI_1102.

[0027] Figure 3 is a hardware configuration diagram of an information processing apparatus that can be used in the maintenance management system 1. The computer 1200 shown in FIG. 3 can be used, for example, as the maintenance management device 1105. The computer 1200 shown in FIG. 3 can also be used as the digital twin system 1106 and the computer terminal 1101.

[0028] The computer 1200 includes, for example, a processor 1201, a memory 1202, a storage 1203, a communication unit 1204, and a user interface unit 1205. In the figure, the user interface is abbreviated as "UI".

[0029] A predetermined computer program stored in the storage 1203 is transferred to the memory 1202. The processor 1201 realizes each function described in FIG. 1 by executing the computer program stored in the memory 1202. The processor 1201 is not limited to a CPU (Central Processing Unit), and may include, for example, a GPU (Graphics Processing Unit) or an ASIC (Application Specific Integrated Circuit).

[0030] The communication unit 1204 is a device for communicating with other devices via the communication network 1103. The user interface unit 1205 is a device for exchanging information between the user using the computer 1200 and the computer 1200. The user interface unit 1205 includes an information input device and an information output device. Examples of the information input device include a keyboard, a pointing device, a touch panel, and a voice instruction device. Examples of the information output device include a monitor display, a printer, a speaker, and a voice synthesizer. The information input device and the information output device may be integrated.

[0031] The computer 1200 is connectable to the storage medium MM. The storage medium MM is configured, for example, like a memory device, a hard disk device, an optical disk device, a magneto-optical disk device, a magnetic tape device, etc., and stores computer programs and data non-temporarily. The storage medium MM can transfer and store computer programs and data to the computer 1200. It is also possible to transfer and store computer programs and data from the computer 1200 to the storage medium MM. A computer program that realizes each function of the maintenance management device 1105 is stored in the storage medium MM, the storage medium MM is connected to another computer, and the computer program stored in the storage medium MM can be installed on the other computer. Thereby, another computer can be made to function as the maintenance management device 1105.

[0032] The communication network 1103 may be a public communication network such as the Internet or a dedicated communication network. The communication network 1103 may be either a wireless communication network or a wired communication network, or a combination of a wireless communication network and a wired communication network.

[0033] Figure 4 schematically shows an example of integrating similar action plans. The graph and processing in Figure 4 are not presented to the user. It is an internal process of the maintenance management device 1105. In step S1301, an action plan that satisfies the constraint conditions and a KPI (Key Performance Indicator) for selecting the action plan are input to the action plan sorting module 1109. The action plan sorting module 1109 integrates similar action plans as described below by a process (function) F04 (S1302 - S1308).

[0034] The action plan sorting module (also referred to as the action plan sorting unit) 1109 assigns the action plan as a black dot mark to a two - dimensional map with two axes of cost and result (S1302). The horizontal axis of the two - dimensional graph is the cost for executing the action. The vertical axis of the two - dimensional graph is the result when the action is executed. The action plan sorting module 1109 calculates the distance between each action plan in the two - dimensional graph (S1303).

[0035] The action plan sorting module 1109 obtains the constraint conditions regarding cost and result from the constraint list 003 (S1305). The constraint conditions are displayed as dotted lines on the two - dimensional graph as a cost threshold and a result threshold.

[0036] The action plan sorting module 1109 groups actions with a distance less than or equal to a predetermined value (S1306). This grouping is called clustering. In FIG. 4, three clusters are shown. Among these clusters, the cluster indicated by the dotted line contains two actions that cross the result threshold. One action included in the dotted-line cluster exceeds the result threshold, and the other action included in the dotted-line cluster is below the result threshold. Therefore, those two actions included in the dotted-line cluster do not exist on the same side with respect to the result threshold and are contradictory.

[0037] The action plan sorting module 1109 integrates clusters without contradictions into one action (S1307). The clusters without contradictions are shown by the solid-line ellipses on the left side of the two-dimensional graph. The integrated cluster becomes one black mark as a whole. In this way, the action plan sorting module 1109 reduces the number of action plans by clustering the ranked action plans and integrating them into non-contradictory action plans (S1308).

[0038] FIG. 5 schematically shows how the model is updated according to the user's selection. On the left side of FIG. 5, the current weights for each action plan are shown (S1401). In the graph shown in FIG. 5, the vertical axis represents the ranked weights, and the horizontal axis represents the action plans. The simplified action plans are output from GUI output 014 to GUI_1102 and displayed (S1402).

[0039] A simplified action plan is an action plan that is simplified and displayed so that it is easy for the user to select. One action plan includes a plurality of sub-action plans. For example, the action plan of "replace the turbine" includes a series of steps such as "stop the generator", "remove the cable", "remove the old turbine", "install the new turbine", "attach the cable", and "trial operation". And action plans are usually generated in multiple copies for each failure of one physical asset. Further, in this embodiment, an action plan is generated for each stakeholder. Therefore, the user needs to select the action plan that the user determines to be the most appropriate from among a large number of action plans. Therefore, the details of the action plan are omitted, and a simplified action plan is presented to the user.

[0040] A user who is a member of the energy vendor selects an action plan that matches the actual situation of the power generation plant 1104 from among the action plans simplified and displayed on the GUI (S1403). The result of the user's selection is reflected in the weight 008. The weight of the action plan selected by the user is increased (S1404). In the graph of FIG. 5, the leftmost action plan is selected, and its weight is increased as shown by the dotted line (S1405). Among the weights stored in the weight 008, the weight corresponding to the action plan selected by the user is increased and updated (S1406). Based on the updated weight, the graph of the action plan presented to the user is also updated (S1407).

[0041] Figure 6 is an example of the first GUI. GUI_1 is the main screen among the GUIs displayed on GUI_1102. GUI_1 includes, for example, user name G101, asset search G102, failure model G103, structural model G104, learning model G105, action plan for stakeholder A G106, action plan for stakeholder B G107, comparison G108, stakeholder name G109 and its importance G110, cost name G111 and new generation G112 and threshold G113, result name G114 and new generation G115 and threshold G116, generation of action plan map G117, save G118, acceptance of recommendation G119.

[0042] User name G101 is information that identifies the user who has logged in to the maintenance management service provided by the maintenance management device 1105. Asset search G102 is a search window for searching for a target asset from a list of physical assets 11042 of the power generation plant 1104. Failure model G103 is a selection part for selecting a model used to detect the failure of the physical asset 11042. Structural model G104 is a selection part for selecting a structural model. Learning model G105 is a selection part for selecting a learning model.

[0043] Action plan for stakeholder A G106 shows the content of the action plan recommended by the action plan ranking module 1108 as an appropriate action plan for the first stakeholder among the stakeholders of the power generation plant 1104. Action plan for stakeholder B G107 shows the content of the action plan recommended by the action plan ranking module 1108 as an appropriate action plan for the second stakeholder.

[0044] The first stakeholder A and the second stakeholder B are stakeholders arbitrarily selected from among the stakeholders of the power generation plant 1104. Here, for the sake of easy explanation, two stakeholders A and B are taken as examples.

[0045] The comparison G108 is an instruction part for comparing action plans for each stakeholder to be compared.

[0046] The stakeholder name G109 and its importance G110 are a designation part for designating the name of the stakeholder and its importance.

[0047] The cost name G111, new generation G112, and threshold G113 are a designation part for the user to set the cost name, new generation, and threshold of the action plan. The result name G114, new generation G115, and threshold G116 are a designation part for the user to set the result, new generation, and threshold of the action plan.

[0048] The generation of the action plan map G117 is an instruction part for the user to instruct the generation of a map that assigns the action plan to a two-dimensional map. The save G118 is an instruction part for saving the content generated in FIG. 6 to the GUI output 014. The recommendation approval G119 is an instruction part for notifying the maintenance management device 1105 that the user has approved the action plan recommended by the maintenance management device 1105.

[0049] FIG. 7 is an example of the second GUI. GUI_2 is a screen that compares the action plans of each stakeholder and displays the comparison results.

[0050] GUI_2 includes, for example, the name G201 of the stakeholder A to be compared, the name G202 of the stakeholder B, the comparison button G203, the comparison result G204, and the return button G205.

[0051] In the comparison result G204, the result of comparing the action plan of stakeholder A and the action plan of stakeholder B is displayed in tabular form. The table includes rank, recommended action plan, cost, and result. Among the action plans recommended for stakeholder A and the action plans recommended for stakeholder B, the different parts between the two are displayed in different colors. For example, based on the action plan of stakeholder A, the parts where the action plan of stakeholder B differs from the action plan of stakeholder A may be displayed in red. Note that since there are usually multiple, rather than just one, action plans recommended for each stakeholder, the lists or tables of the action plans of the stakeholders are compared with each other.

[0052] Figure 8 is an example of the third GUI. GUI_3 is a screen for setting KPIs related to action plans for each stakeholder. GUI_3 includes, for example, an action plan map G301, an action plan being edited G302, the current cost KPI_G303, the current result KPI_G304, and save G305.

[0053] The action plan map G301 assigns one or more possible action plans for a certain physical asset 11042 to a two-dimensional map. The horizontal axis of the action plan map indicates the current cost KPI (here, "amount of money"). The vertical axis of the action plan map indicates the current result KPI (here, "energy output"). In the example of Figure 8, seven action plans from action plan A to G are shown.

[0054] In the current cost KPI_G303, the evaluation results of each action plan in the current cost KPI are displayed in a bar graph. In the current cost KPI_G304, the evaluation results of each action plan in the current result KPI are displayed in a bar graph. Any graph that makes it easy to visually identify the superiority or inferiority among action plans is acceptable, not limited to bar graphs. By pressing Save G305, the stakeholder causes the cost KPI and result KPI set in GUI_3 to be saved to GUI output 014.

[0055] Figure 9 shows processes F1a and F1b for generating an action plan. In the first process F1a, the action plan generation module 1107 (also referred to as the action generation unit 1107) connects to the failure model 015 of the digital twin system 1106 via the communication network 1103 (S101). The action plan generation module 1107 obtains a list 010 of failure models from the failure model 015. The user specifies actions, costs, and results (S102). As a result, an action list 016, a cost list 018, and a result list 017 are generated, and these lists 016, 018, 017 are stored in the human model 012, which is a model created by the user (S107).

[0056] In the second process F1b, the action plan generation module 1107 obtains data from the human model 012 (S108). The action plan generation module 1107 performs a simulation to obtain results for all conditions (S103).

[0057] The action plan generation module 1107 determines whether the result obtained in step S103 satisfies the cost KPI (S104). If the obtained result satisfies the cost KPI (S104: YES), it proceeds to step S105; if not (S104: NO), it returns to step S103.

[0058] The action plan generation module 1107 determines whether the result obtained in step S103 satisfies the result KPI (S105). If the obtained result satisfies the result KPI (S105: YES), it proceeds to step S106; if not (S105: NO), it returns to step S103.

[0059] The action plan generation module 1107 outputs actions, costs, and results for each KPI and each condition (S106). Thereby, the action plan list 002 is generated.

[0060] Figure 10 shows the process F3 of excluding action plans that do not satisfy the constraint conditions. The action plan sorting module 1109 acquires the action plan list 002 and the constraint list 003 of the physical asset 11042, and analyzes the action plan list using these data (S201).

[0061] The action plan sorting module 1109 determines whether the nth row of the action plan list violates the constraints in the constraint list 003 (S202). If the action plan sorting module 1109 determines that there is a violation (S202: YES), it deletes the violating row (S203a).

[0062] If the target row does not violate the constraints (S202: NO), the action plan sorting module 1109 determines whether all rows of the action plan list 002 have been inspected (S203b). If there are uninspected rows (S203: NO), it moves the pointer to the next row (S204a) and returns to step S201.

[0063] If all rows have been inspected (S203: YES), the action plan sorting module 1109 saves it as a list of filtered action plans (S204b). Thereby, the filtered action plan 106 is generated.

[0064] Figure 11 shows the process F5b of defining the second stakeholder B. The stakeholder B, who is a technology provider, uses the second computer terminal 1101(2) to access the maintenance management device 1105 and display GUI_1 (S301). The maintenance management device 1105 acquires the cost and result of each action plan specified by the stakeholder B (S302).

[0065] The maintenance management device 1105 acquires the weight of each action plan specified by the stakeholder B (S303). The maintenance management device 1105 saves the information specified in steps S302 and S303 as the content of the stakeholder B (S304). As a result, a list 006 of ranked action plans for the stakeholder B is generated.

[0066] Figure 12 shows the process F2 of recommending an action plan. The action plan ranking module 1108 acquires the training data 004 and the model serving 005. The action plan ranking module 1108 sets the "condition" column of the filtered action plan to "measured value" and the "condition" column of the model serving to "predicted value" (S401).

[0067] The action plan ranking module 1108 calculates the weight by calculating the difference between the predicted value and the measured value (S402). Furthermore, the action plan ranking module 1108 divides the list 106 of filtered action plans into rows that match and do not match the "process / asset ID" of the model serving 005 (S402).

[0068] When it matches the "process / asset ID" (split 1), the action plan ranking module 1108 sets a weight of 100% for the best fit, a weight of 51% for the worst fit, and interpolates the weights for intermediate fits (S403).

[0069] If it does not match the "Process / Asset ID" (Split 2), the Action Plan Ranking Module 1108 sets a weight of 50% for the best fit and 0% for the worst fit, and interpolates the weights for the intermediate fits (S404).

[0070] The Action Plan Ranking Module 1108 integrates the split tables into one and sorts them in descending order of weight (S405). This generates a list 012 of the ranked action plans for the stakeholders. FIG. 13 shows a process F5c of comparing action plans for each stakeholder among the stakeholders. The flowchart of FIG. 13 is a subroutine of FIG. 15 described later. The Maintenance Management Device 1105 acquires the ranked action plans for each stakeholder (S504) and extracts the action plan of the first stakeholder A and the action plan of the second stakeholder B.

[0071] The Maintenance Management Device 1105 sets the "Condition" column of the filtered action plan to "Measured Value" and the "Condition" column of the model serving to "Predicted Value" (S502).

[0072] FIG. 14 shows a process F4 of integrating and organizing similar action plans into one. As shown on the left side of FIG. 14, the Action Plan Organization Module 1109 represents all costs and results in a two-dimensional map based on the ranked action plan 006 (S601). The Action Plan Organization Module 1109 calculates a distance matrix for all action plans with thresholds (S602) and integrates the action plans with close distances into similar clusters (S603). This generates the clustered action plan 009.

[0073] The action plan sorting module 1109 deletes a predetermined action within a similar cluster based on the clustered action plan 09 and the constraint list 006 (S604). The predetermined action is (1) the action farthest from the centroid of the similar cluster and (2) the action on the same side of the constraint threshold. Actions that exceed the threshold of the constraint condition and actions that do not reach the threshold of the same constraint condition are not subject to deletion even if they belong to the same similar cluster. Through step S604, a list 007 of the simplified action plans is obtained. The list 007 of the simplified action plans is displayed on the GUI_1 and provided to the user.

[0074] Figure 15 shows the process F5a in which the user selects an action plan via the GUI. The user verification module 1111 determines whether the second stakeholder B is defined (S701). When the determination in step S701 is "YES", the user verification module 1111 compares the action plan of the first stakeholder A and the action plan of the second stakeholder B (S702a).

[0075] When the determination in step S701 is "NO", the user verification module 1111 determines whether to define the second stakeholder B (S702b). When defining the second stakeholder B (S702b: YES), the user verification module 1111 causes the simplified action plan for the second stakeholder B to be displayed on the GUI (S703a) and moves to step S702a.

[0076] When the determination in step S702b is "NO", the user verification module 1111 obtains the ranked action plan 006 of the first stakeholder A and causes the simplified action plan to be displayed on the GUI (S703b).

[0077] The user verification module 1111 updates the failure model 015 (processing F07 in FIG. 5 described later) or updates the action plan recommendation engine F02 by changing the weight of the action plan recommendation engine F02 based on the content selected by the user via the GUI (which action plan was selected) (S704).

[0078] FIG. 16 shows the process F6 of adjusting the weight according to the user's selection. When the user verification module 1111 selects an action plan via the GUI (S801), the user verification module 1111 increases the weight of the action plan selected by the user (S802) and updates the weight 008. The user verification module 1111 updates the ranking of the action plans based on the updated weight (S803).

[0079] FIG. 17 shows the process F7 of updating the failure model 015. The user verification module 1111 accesses the FMEA010 of the failure model 015 (FIG. 26 described later) (S901a) and detects the action plan selected by the user (S901b). The user verification module 1111 adds the action plan selected by the user to the FMEA010 of the failure model 015 based on the FMEA010 and the ranked action plans 006 of each stakeholder A and B (S903). As a result, the enhanced FMEA011 is generated.

[0080] Using FIGS. 18 to 40, a configuration example of the tables and lists used in the maintenance management system 1 will be described.

[0081] FIG. 18 is the list 002 of action plans. The action plan list 002 includes, for example, a process / asset ID for specifying a condition, an action, a process, or an asset, a cost KPI, a cost value, a result KPI, and a result value. The condition can be defined, for example, by the required time of the action.

[0082] Figure 19 is list 003 of asset constraints. List 003 is used by exclusion filter F03 to exclude action plans that violate the constraints. The constraint list 003 set for physical asset 11042 includes, for example, the target action, the asset / process ID that identifies the asset or process, the cost KPI, the conditions that the cost should meet, the result KPI, and the conditions that the result should meet.

[0083] Figure 20 is training data 004 input into the model. The data 004 for training the model includes, for example, conditions (time constraints), actions, process / asset IDs, cost KPIs, cost values, result KPIs, and result values.

[0084] Figure 21 is an example of the input for model serving 005. Model serving 005 includes, for example, conditions (time constraints) and process / asset IDs.

[0085] Figure 22 is list 006 of ranked action plans for each stakeholder. The ranked action plans 006 generated and ranked for each stakeholder include, for example, ranking weights, actions, process / asset IDs, cost KPIs, cost values, result KPIs, and result values.

[0086] Figure 23 is list 007 of sorted and ranked action plans. This list 007 includes, for example, ranking weights, actions, process / asset IDs, cost KPIs, cost values, result KPIs, and result values.

[0087] Figure 24 shows updated weights 008. The updated weights 008 associate, for example, the old ranking weights with the new ranking weights.

[0088] Figure 25 is clustered action plan 009. The clustered action plan 009 associates, for example, the cluster identification number with the identification information of the actions belonging to the cluster.

[0089] Figure 26 is a table of FMEA010. FMEA010 includes, for example, information for identifying failure models, cases, local impacts, and global impacts.

[0090] Figure 27 is a table of the improved FMEA011. The improved FMEA011 based on the action plan selected by the user includes, for example, failure modes, cases, local impacts, global impacts, and action plans. The action plan includes proposed actions, costs, and results.

[0091] Figure 28 is the human model 012 represented as a graphical database. The human model 012 presents to the user, in a tree structure, for example, a list of assets that are parent nodes, the ID of each asset included in the list, failure modes, repair actions corresponding to the failures, the costs required to execute the actions, and the results of the actions.

[0092] Repair actions include, for example, cleaning, blade replacement, generator shutdown, etc. Cost KPIs include, for example, energy output, cost, labor cost (man-hours), etc. Result KPIs include, for example, energy output, cost, labor cost, etc.

[0093] Figure 29 is the output 013 of module F04 for simplifying the action plan. The output 013 of module F04 for simplifying the action plan includes, for example, the clustered action 009 and the simplified action plan 007.

[0094] Figure 30 is the result table 014 of the balanced action plan among stakeholders. This table 014 includes, for example, ranking weights, actions, process / asset IDs, cost KPIs, cost values, result KPIs, and result values.

[0095] Figure 31 shows the structure of the failure model 015. The list of physical assets includes a plurality of physical assets 11042, but here we will explain by taking one physical asset as an example.

[0096] On the left side of Figure 31, the original data of the failure model 015 is shown by a solid line. On the right side of Figure 31, the data added by the maintenance management device 1105 is shown by a dotted line. In the original data, under the identification information (ID) of FMEA, the failure mode, failure case, local impact, and global impact are associated. The new data added by the maintenance management device 1105 has the cost and result associated under the identification information (ID) of the action plan. Therefore, the more the maintenance management device 1105 operates, for each physical asset in the failure model 015, the action plan adopted by the user (energy vendor), that is, the proven action plan selected and executed by that user, will be gradually added and registered. As a result, the failure model 015 can be strengthened according to the actual situation of the power generation plant 1104, and the maintainability can be improved.

[0097] Figure 32 is a list 016 of action plans for each physical asset stored in the human model 012.

[0098] Figure 33 is a list 017 showing the relationship between the action plan for each asset and the cost KPI stored in the human model 012. This list 017 includes, for example, the identification information of the physical asset, the action plan, the cost KPI, and the value of the cost.

[0099] Figure 34 is a result list 018 of action plans for each physical asset. The result list 018 includes, for example, the identification information of the physical asset, the action plan, the result KPI, and the value of the result.

[0100] As shown in Figure 35, the filter output 019 stores the model training input 004.

[0101] Figure 36 is the additional input 020 to the filtering module. This additional input 020 contains, for example, a list 003 of constraints for each physical asset.

[0102] As shown in Figure 37, the GUI output 014 stores the data 021 output by the action plan simplification module F04. The data 021 contains the balanced result data 014 among the stakeholders.

[0103] Figure 38 is a data store 022 that stores the weights of the model. The data store 022 stores the weights 008 for updating the action plan recommendation engine F02 as a machine learning model.

[0104] Figure 39 is the output 023 of the action plan recommendation engine module F02. This output 023 contains the ranked action plans 006 for each stakeholder.

[0105] Figure 40 is the output 024 of the simulation module F1b that generates the action plan. This output 024 contains a list 002 of action plans.

[0106] According to the maintenance management system 1 of this embodiment configured as described above, the user can efficiently and appropriately maintain the power generation plant.

[0107] According to the maintenance management system 1 of this embodiment, when a failure of a power generation plant, which is a large-scale system involving a plurality of stakeholders (energy vendors and each technology provider), is detected, the user, who is an energy vendor, can select an appropriate action plan according to the actual situation of the power generation plant.

[0108] In the maintenance management system 1 of this embodiment, after consolidating and organizing a plurality of action plans generated for each stakeholder, they are presented to the user who is an energy vendor. The maintenance management system 1 can reduce unnecessary action plans from among a wide variety of action plans and combine similar action plans into one. As a result, the user can relatively easily select an appropriate action plan. Therefore, the time for the user to make a judgment when the power generation plant 1104 fails can be shortened, the detected failure can be promptly addressed, and the reliability of the power generation plant can be improved.

[0109] In the maintenance management system 1 of this embodiment, based on the result selected by the user from among the recommended action plans, in order to improve the failure model 015 and the action plan recommendation engine F02, a model that suits the actual situation of the power generation plant 1104 can be created, and maintenance can be performed efficiently and appropriately.

[0110] Note that the present invention is not limited to the above-described embodiments and includes various modifications. The above embodiments have been described in detail for easy understanding of the present invention and are not necessarily limited to those having all the configurations described. Also, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment. Further, the configuration of another embodiment can be added to the configuration of one embodiment. Also, for a part of the configuration of each embodiment, other configurations can be added, deleted, or replaced.

[0111] Each component of the present invention can be arbitrarily selected, and an invention having the selected configuration is also included in the present invention. Furthermore, the configurations described in the claims can be combined in combinations other than those explicitly stated in the claims.

[0112] This embodiment clearly discloses the following program claims. These program claims can be non-temporarily stored in the storage medium MM.

[0113] (Program Claim) A computer program that causes a computer 1200 to function as a maintenance management device 1105, an action plan generation unit that generates at least one action plan corresponding to a failure of the power plant, an action plan evaluation unit that evaluates and recommends the generated action plan, an action plan arrangement unit that arranges the recommended action plan, and a computer program comprising the same.

Description of Symbols

[0114] 1: Maintenance management system, 1101: Computer terminal, 1103: Communication network, 1104: Power plant, 1105: Maintenance management device, 1106: Digital twin system, 1107: Action plan generation unit, 1108: Action plan ranking unit, 1109: Action plan arrangement unit, 1110: Asset model server, 1111: User verification unit.

Claims

1. A maintenance management device for managing the maintenance of a power generation plant, comprising: an action plan generation unit that generates at least one action plan corresponding to a failure of the power generation plant; an action plan evaluation unit that evaluates and recommends the generated action plan; an action plan arrangement unit that arranges the recommended action plan; and a maintenance management device for a power generation plant.

2. The maintenance management device for a power generation plant according to claim 1, further comprising: a verification unit that presents the action plan arranged by the action plan arrangement unit to a computer terminal and receives a confirmation result from the computer terminal. A maintenance management device for a power generation plant.

3. The maintenance management device for a power generation plant according to claim 2, wherein: the verification unit reflects the received confirmation result in the processing by the action plan generation unit and the processing by the action plan evaluation unit. A maintenance management device for a power generation plant.

4. The maintenance management device for a power generation plant according to claim 3, wherein: the action plan generation unit generates an action plan for each stakeholder of the power generation plant. A maintenance management device for a power generation plant.

5. The maintenance management device for a power generation plant according to claim 4, wherein: the action plan generation unit generates the action plan for each stakeholder based on the cost required for generating the action plan and the execution result of the action plan. A maintenance management device for a power generation plant.

6. The maintenance management device for a power generation plant according to claim 5, wherein: the action plan evaluation unit evaluates and recommends the action plan for each stakeholder based on the cost, the execution result, and the weight input from the computer terminal. A maintenance management device for a power generation plant.

7. The maintenance management device for a power generation plant according to claim 6, wherein: the action plan arrangement unit excludes an action plan that meets a predetermined exclusion condition among the recommended action plans. A maintenance management device for a power generation plant.

8. The maintenance management device for a power generation plant according to claim 7, wherein: The action plan consolidation unit integrates similar action plans among the recommended action plans into one action plan. A maintenance management device for a power generation plant.

9. A maintenance management device for a power generation plant according to claim 6, The computer terminal includes a first computer terminal used by an energy vendor that operates the power generation plant among the stakeholders, and a second computer terminal used by a technology provider that provides the technology used in the power generation plant among the stakeholders. The action plan generation unit generates at least one action plan for coping with a failure of the power generation plant based on evaluation criteria set from the second computer terminal. The action plan evaluation unit evaluates and recommends the action plan for each stakeholder based on the cost, the execution result, and the weight input from the first computer terminal. The verification unit presents the consolidated action plan to the first computer terminal and receives a confirmation result from the first computer terminal. A maintenance management device for a power generation plant.

10. A maintenance management method for managing the maintenance of a power generation plant by a maintenance management device, The maintenance management device, generates at least one action plan for coping with a failure of the power generation plant, evaluates and recommends the generated action plan, and consolidates the recommended action plan. A maintenance management method for a power generation plant.

11. A maintenance management system for a power generation plant, a digital twin management system that receives signals from a sensor group provided in the power generation plant via a communication network, reflects the signals on a power generation plant in a virtual space, and detects a failure of the power generation plant; a maintenance management device that is connected to the digital twin management system via the communication network and manages the maintenance of the power generation plant; a computer terminal that is connected to the maintenance management device via the communication network and is used by a stakeholder of the power generation plant; comprising The maintenance management device includes a processor, a memory that stores a predetermined program code executed by the processor, and a communication interface unit connected to the communication network. By executing the predetermined program code, the processor generates at least one action plan corresponding to a failure of the power plant, evaluates and recommends the generated action plan, organizes the recommended action plan, sends the organized action plan to the computer terminal to request confirmation by the stakeholder, reflects the confirmation result from the computer terminal in the generation of the action plan and the recommendation of the action plan A maintenance management system for a power plant.

Images