Data processing device, data processing method and data processing program

Abstract

According to the present invention, a collation data generation unit (103) generates collation data used for estimating the condition of a machine in correspondence with a construction requirement, which is a requirement when constructing the machine. A sensor data acquisition unit (104) acquires sensor data of a diagnosis target machine (200), which is a machine for which construction has finished. A condition estimation unit (105) collates the sensor data and the collation data and estimates the condition of the diagnosis target machine (200).

Description

[Technical Field]

[0001] The present disclosure relates to a technique for estimating a state of a device. [Background technology]

[0002] In the past, in order to diagnose the state of a device, sensor data was acquired from the device to be diagnosed. Then, the acquired sensor data was compared with pre-prepared sensor data at the time of an abnormality (hereinafter referred to as abnormality data) to estimate the location and cause of the abnormality in the device to be diagnosed. However, depending on the type of equipment, abnormalities may occur infrequently, making it impossible to acquire and store abnormality data. Furthermore, it is not easy to prepare a plurality of types of abnormality data corresponding to a plurality of types of abnormality causes.

[0003] Therefore, in the technology of Patent Document 1, data simulating an abnormality is generated by applying hypothetical abnormal conditions to a physical model or a mathematical model (hereinafter, both are collectively referred to as a model) generated using sensor data under normal conditions (hereinafter, referred to as normal data).The technology of Patent Document 1 then obtains a simulated diagnostic pattern using the data simulating an abnormality. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6856443 Summary of the Invention [Problem to be solved by the invention]

[0005] When diagnosing a plurality of devices of the same model that are used in different installation environments using the technology of Patent Document 1, the following two problems can be considered. (1) Problems of reduced diagnostic accuracy due to differences in construction conditions The technology of Patent Document 1 generates a model from normal data of a certain device (referred to as device M). The normal data of another device (referred to as device N) of the same model as device M may differ from the normal data of device M due to differences in the installation conditions between device M and device N. For this reason, if device N is diagnosed using a model generated using the normal data of device M, the diagnostic accuracy may decrease. (2) Occurrence of duplicate processing One possible solution to the problem (1) above is to generate a model for each device. However, the technology of Patent Document 1 requires acquiring normal data for each device and generating a model for each device using the normal data for each device. As such, the technology of Patent Document 1 results in redundant processing in acquiring normal data and generating a model.

[0006] The primary objective of the present disclosure is to solve these problems. More specifically, the present disclosure aims to enable accurate estimation of the state of equipment, regardless of the installation conditions, without causing duplicate processing. [Means for solving the problem]

[0007] The data processing device according to the present disclosure includes: a matching data generating unit that generates matching data used to estimate the state of the equipment in accordance with construction conditions that are conditions when the equipment is constructed; a sensor data acquisition unit that acquires sensor data of a diagnosis target device that has been installed; The diagnostic device further includes a state estimation unit that compares the sensor data with the comparison data and estimates the state of the device to be diagnosed. [Effects of the Invention]

[0008] According to the present disclosure, the state of equipment can be estimated with high accuracy without causing duplicate processing, regardless of the construction conditions under which the equipment is constructed. [Brief explanation of the drawings]

[0009] [Figure 1]FIG. 1 is a diagram showing an example of the functional configuration of a fault diagnosis device according to a first embodiment. [Figure 2] FIG. 3 is a diagram showing an example of tree information according to the first embodiment. [Figure 3] FIG. 3 is a diagram showing an example of matching data according to the first embodiment. [Figure 4] FIG. 4 is a diagram showing an example of a matching process according to the first embodiment. [Figure 5] FIG. 3 is a diagram showing an example of tree information according to the first embodiment. [Figure 6] 3 is a flowchart showing an example of the operation of the fault diagnosis device according to the first embodiment. [Figure 7] FIG. 3 is a diagram showing an example of tree information according to the first embodiment. [Figure 8] FIG. 10 is a diagram showing an example of the functional configuration of a fault diagnosis device according to a second embodiment. [Figure 9] 10 is a flowchart showing an example of the operation of the fault diagnosis device according to the second embodiment. [Figure 10] FIG. 10 is a diagram showing an example of the functional configuration of a fault diagnosis device according to a third embodiment. [Figure 11] 10 is a flowchart showing an example of the operation of the fault diagnosis device according to the third embodiment. [Figure 12] FIG. 1 is a diagram showing an example of the hardware configuration of a fault diagnosis device according to a first embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, embodiments will be described with reference to the drawings. In the following description of the embodiments and the drawings, the same reference numerals denote the same or corresponding parts.

[0011] Embodiment 1 ***Configuration Description*** Fig. 1 shows an example of the functional configuration of fault diagnosis equipment 100 according to this embodiment. Fig. 12 shows an example of the hardware configuration of fault diagnosis equipment 100. The fault diagnosis device 100 is a computer. Furthermore, fault diagnosis equipment 100 corresponds to a data processing device, the operating procedure of fault diagnosis equipment 100 corresponds to a data processing method, and the program that realizes the operation of fault diagnosis equipment 100 corresponds to a data processing program.

[0012] Before describing in detail an example of the functional configuration and hardware configuration of the fault diagnosis device 100, an outline of the operation of the fault diagnosis device 100 will first be described.

[0013] Fault diagnosis device 100 acquires design information and construction conditions of the device, and generates collation data by associating the design information with the construction conditions.

[0014] The verification data is data used to estimate the state of the device. The collation data is data that simulates sensor data from the diagnosis target device 200, which will be described later. The collation data includes normal simulation data and abnormal simulation data. The normal simulation data is data that simulates the sensor data of the diagnosis target device 200 in a normal state. The abnormal simulation data is data that simulates the sensor data of the diagnosis target device 200 in an abnormal state. When there is no need to distinguish between the normal simulation data and the abnormal simulation data, both the normal simulation data and the abnormal simulation data are referred to as simulation data.

[0015] Design information is information obtained from the design drawings of the equipment. Specifically, the design information describes the specifications of the equipment that are determined when the equipment is designed. The specifications of the equipment include, for example, details of the equipment configuration, details of the parts included in the equipment, details of the equipment functions, etc.

[0016] Construction conditions are the conditions for constructing the equipment. Construction involves installing the equipment in a space (building, etc.) according to the construction drawings. Construction may involve construction work. Construction conditions are conditions that are obtained from construction drawings and must be determined when constructing equipment. Construction conditions vary depending on the construction method of the equipment and the type of space (building, etc.) in which the equipment is installed. Construction conditions include values that affect the operating status of the equipment.

[0017] When an air conditioner is used as an example of equipment, the design information indicates the model of the outdoor unit and the model of the indoor unit as details of the equipment configuration. The design information also indicates details of parts included in the equipment, such as the outer diameter and wall thickness of the pipes. The design information also indicates details of the equipment's functions, such as the fan air volume of the heat exchanger, the heat transfer correction coefficient, and the range of motion of the solenoid valve.

[0018] Furthermore, when an air conditioner is used as an example of equipment, the installation conditions include the height difference between the indoor heat exchanger and the outdoor heat exchanger, the amount of refrigerant charged, the number of indoor units to be installed, the length of piping, and the like.

[0019] In this embodiment, for the sake of simplicity, it is assumed that one piece of design information exists for one model. An example in which multiple pieces of design information exist for one model will be described later. Regarding the construction conditions, it is assumed that there are multiple construction conditions for one model. Fault diagnosis equipment 100 generates collation data for each construction condition. That is, fault diagnosis equipment 100 generates normal simulation data and abnormal simulation data as collation data for each of a plurality of construction conditions.

[0020] The fault diagnosis device 100 acquires sensor data from the diagnosis target device 200. The fault diagnosis device 100 is connected to the diagnosis target device 200 via the Internet or the like. The diagnosis target device 200 is a device that has already been installed in the space, and is the same model as the device whose design information and installation conditions have been acquired. The sensor data is data collected by a sensor installed in the diagnosis target device 200 .

[0021] The fault diagnosis device 100 compares the sensor data of the diagnosis target device 200 with multiple comparison data (normal simulation data and abnormal simulation data) for multiple installation conditions. The fault diagnosis device 100 estimates the state (normal / abnormal) of the diagnosis target device 200 by determining the simulation data that is most similar to the sensor data. As described above, the fault diagnosis device 100 generates the matching data for each construction condition. Therefore, no matter which of the construction conditions the diagnosis target device 200 is constructed under, the fault diagnosis device 100 can accurately estimate the state of the diagnosis target device 200 by matching the sensor data with the matching data.

[0022] Then, the fault diagnosis device 100 presents the estimated state of the diagnosis target device 200 as a diagnosis result to the user 300. The user 300 is, for example, a maintenance technician for the diagnosis target device 200.

[0023] Next, an example of the hardware configuration of the fault diagnosis equipment 100 will be described with reference to FIG.

[0024] As shown in FIG. 12, fault diagnosis equipment 100 includes, as hardware, a processor 901, a main memory device 902, an auxiliary memory device 903, a communication device 904, and an input / output device 905. The functions of the design information acquisition unit 101, construction condition acquisition unit 102, matching data generation unit 103, sensor data acquisition unit 104, state estimation unit 105 and diagnosis result output unit 106 shown in FIG. 1 are realized by, for example, a program. The auxiliary storage device 903 stores programs that realize the functions of the design information acquisition unit 101, the construction condition acquisition unit 102, the matching data generation unit 103, the sensor data acquisition unit 104, the state estimation unit 105, and the diagnosis result output unit 106. These programs are loaded from the auxiliary storage device 903 to the main storage device 902. Then, the processor 901 executes these programs to perform the operations of a design information acquisition unit 101, a construction condition acquisition unit 102, a matching data generation unit 103, a sensor data acquisition unit 104, a state estimation unit 105, and a diagnosis result output unit 106, which will be described later. FIG. 12 shows a schematic diagram of a state in which the processor 901 is executing a program that realizes the functions of the design information acquisition unit 101, the construction condition acquisition unit 102, the matching data generation unit 103, the sensor data acquisition unit 104, the state estimation unit 105, and the diagnosis result output unit 106. The communication device 904 communicates with the diagnosis target device 200 via the Internet or the like. The input / output device 905 is, for example, a keyboard, a mouse, a display, etc. The input / output device 905 receives instructions from the user 300. The input / output device 905 also presents various types of information to the user 300.

[0025] Next, an example of the functional configuration of the fault diagnosis equipment 100 will be described with reference to FIG.

[0026] The design information acquisition unit 101 acquires design information. For example, the design information acquisition unit 101 acquires the design information from a designer of the equipment. Alternatively, the design information acquisition unit 101 may acquire the design information from the equipment. The design information acquisition unit 101 stores the acquired design information in the construction condition holding unit 107 .

[0027] The construction condition acquisition unit 102 acquires a plurality of construction conditions. The construction condition acquisition unit 102 acquires the construction conditions from, for example, a designer of the equipment. The construction condition acquisition unit 102 stores the acquired construction conditions in the construction condition storage unit 107 .

[0028] The collation data generating unit 103 generates collation data for each construction condition. More specifically, the matching data generating unit 103 generates a physical model or a mathematical model (hereinafter simply referred to as a model) from the design information and construction conditions stored in the construction condition storing unit 107. Then, the matching data generating unit 103 generates matching data from the model for each construction condition. As described above, the matching data includes normal simulation data and abnormal simulation data. The matching data generating unit 103 generates matching data before the sensor data acquiring unit 104 acquires sensor data from the diagnosis target device 200 . The matching data generating unit 103 stores the generated matching data in the matching data holding unit 108 . The process performed by the matching data generating unit 103 corresponds to a matching data generating process.

[0029] The sensor data acquiring unit 104 acquires sensor data from the diagnosis target device 200. More specifically, the sensor data acquiring unit 104 acquires the sensor data via the communication device 904. The sensor data acquisition unit 104 outputs the acquired sensor data to the state estimation unit 105. The processing performed by the sensor data acquisition unit 104 corresponds to a sensor data acquisition process.

[0030] When the state estimation unit 105 acquires the sensor data from the sensor data acquisition unit 104, it reads out the verification data (normal simulation data and abnormal simulation data) for all the construction conditions from the verification data storage unit . Then, the state estimation unit 105 checks (compares) the sensor data with each piece of simulation data. As a result of the check, the state estimation unit 105 determines the simulation data that is most similar to the sensor data. Then, the state estimation unit 105 estimates the state of the diagnosis target device 200 based on the simulation data that is most similar to the sensor data. That is, if the simulation data that is most similar to the sensor data is normal simulation data, the state estimation unit 105 estimates that the diagnosis target device 200 is normal. On the other hand, if the simulation data that is most similar to the sensor data is abnormal simulation data, the state estimation unit 105 estimates that the diagnosis target device 200 is abnormal.

[0031] Furthermore, if the collation data generating unit 103 generates abnormality simulation data for each abnormality cause, the state estimating unit 105 can estimate the abnormality cause. For example, assume that the collation data generating unit 103 has generated abnormality simulation data corresponding to abnormality cause α and abnormality simulation data corresponding to abnormality cause β. In this case, if the simulation data most similar to the sensor data is the abnormality simulation data corresponding to abnormality cause α, the state estimation unit 105 can estimate that the diagnosis target device 200 is abnormal, and further estimate that the cause of the abnormality is abnormality cause α.

[0032] The processing performed by the state estimation unit 105 corresponds to a state estimation processing.

[0033] The diagnosis result output unit 106 presents the estimated state (normal / abnormal), which is the diagnosis result of the state estimation unit 105, to the user 300. In addition, if the state estimation unit 105 estimates an abnormality as the state of the diagnosis target device 200 and also estimates the cause of the abnormality, the diagnosis result output unit 106 also presents the estimated cause of the abnormality to the user 300. Specifically, the diagnostic result output unit 106 displays the estimated state (and the estimated cause of the abnormality) on the input / output device 905 (display). The diagnostic result output unit 106 may further display the waveform of the sensor data, the matching data with the highest similarity, and the similarity between the sensor data and the matching data.

[0034] The construction condition holding unit 107 holds the design information acquired by the design information acquisition unit 101 and the construction conditions acquired by the construction condition acquisition unit 102 in association with each other.

[0035] The collation data storage unit 108 stores the collation data for each construction condition generated by the collation data generation unit 103.

[0036] FIG. 2 shows an example of tree information according to this embodiment. The tree information in Fig. 2 represents the relationship between the construction conditions, the verification data, and the state. The tree information in Fig. 2 is stored in the verification data storage unit 108. The state estimation unit 105 estimates the state of the diagnosis target device 200 by referring to the tree information in Fig. 2.

[0037] In the tree information in Figure 2, the top node is the "design information" node. The "construction conditions" node is placed below the "design information" node. The "verification data" node is placed below the "construction conditions" node. The "status" node is placed below the "verification data" node. The "status" node indicates the state that the equipment can be in. Specifically, the "status" node has a "normal" node and an "abnormal" node. "Abnormality α", "Abnormality β", "Abnormality γ", and "Abnormality δ" represent types of abnormalities that the device can have.

[0038] FIG. 2 shows tree information including nodes for "Construction Condition (I)" and "Construction Condition (II)." If other construction conditions (such as "Construction Condition (III)") exist, the nodes for the other construction conditions (such as "Construction Condition (III)") are also included in the tree information. Then, nodes for "Matching Data" and "Status" are placed below the nodes for the other construction conditions.

[0039] "Matching data (I)-1" is connected to "Status: Normal." Therefore, "Matching data (I)-1" simulates sensor data when the device is normal, and is normal simulation data. On the other hand, "Matching data (I)-2" is connected to "Status: Abnormal α." Therefore, "Matching data (I)-2" simulates sensor data when the device is abnormal, and is abnormal simulation data. "Matching data (I)-3" is connected to "Status: Abnormal β." Therefore, "Matching data (I)-3" simulates sensor data when the device is abnormal, and is abnormal simulation data. Similarly, "matching data (II)-1" is connected to "state: normal." Therefore, "matching data (II)-1" simulates sensor data when the device is normal, and is normal simulation data. On the other hand, "matching data (II)-2" is connected to "state: abnormal γ." Therefore, "matching data (II)-2" simulates sensor data when the device is abnormal, and is abnormal simulation data. "matching data (II)-3" is connected to "state: abnormal δ." Therefore, "matching data (II)-3" simulates sensor data when the device is abnormal, and is abnormal simulation data.

[0040] 2 is stored in the collation data holding unit 108. The state estimation unit 105 estimates the state of the diagnosis target device 200 by referring to the tree information shown in FIG. For example, if the matching data having the highest similarity to the sensor data is "matching data (I)-2", the state estimation unit 105 estimates the state of the diagnosis target device 200 to be "abnormal α".

[0041] FIG. 3 shows an example of the matching data.

[0042] In the example of FIG. 3, each piece of collation data simulates sensor data (time-series data) from sensors X, Y, and Z arranged in the device. That is, the "verification data (I)-1" simulates the sensor data of the sensors X, Y, and Z when the device is "normal." "Verification data (I)-2" simulates the sensor data of sensors X, Y, and Z when the device is in "abnormal condition α." "Verification data (I)-3" simulates the sensor data of sensors X, Y, and Z when the device is in "abnormal β" state. The collation data in Fig. 3 simulates sensor data for each sensor. Alternatively, the collation data may simulate composite data obtained by combining multiple pieces of sensor data from multiple sensors.

[0043] FIG. 4 shows an example of a process of matching sensor data acquired from the diagnosis target device 200 with matching data.

[0044] The state estimation unit 105 calculates the similarity between each piece of matching data and the sensor data by using, for example, Euclidean distance, Manhattan distance, or DTW (Dynamic Time Wrapping). Then, the state estimation unit 105 selects the matched data with the highest similarity. Furthermore, the state estimation unit 105 identifies the "state" node connected to the matched data with the highest similarity, with reference to the tree information in FIG. In the example of Figure 4, the matching data with the highest similarity to the sensor data is "matching data (I)-2." In the tree information of Figure 2, the "state" node connected to "matching data (I)-2" is "abnormality α." Therefore, the state estimation unit 105 estimates that the state of the diagnosis target device 200 is "abnormal α."

[0045] Instead of the tree information shown in FIG. 2, the tree information shown in FIG. 5 may be used. FIG. 5 shows an example of tree information in which the "Cause of Abnormality" node is located below the "Abnormality" node. In the example of Figure 5, "abnormality α" has "abnormality cause α-1" and "abnormality cause α-2" below it. "abnormality β" has "abnormality cause β-1" and "abnormality cause β-2" below it. "abnormality γ" has "abnormality cause γ-1" and "abnormality cause γ-2" below it. "abnormality δ" has "abnormality cause δ-1" and "abnormality cause δ-2" below it.

[0046] 5 is used, the tree information shown in Fig. 5 is stored in the collation data storage unit 108. The state estimation unit 105 refers to the tree information shown in Fig. 5 to estimate the state of the diagnosis target device 200 and the cause of the abnormality. For example, as shown in FIG. 4, when the matching data having the highest similarity to the sensor data is matching data (I)-2, the state estimation unit 105 estimates the state of the diagnosis target device 200 as “abnormality α”, and further estimates “abnormality cause α-1” and “abnormality cause α-2” as the causes of the abnormality.

[0047] ***Explanation of Operation*** Next, an example of the operation of the fault diagnosis equipment 100 according to this embodiment will be described. FIG. 6 is a flowchart showing an example of the operation of the fault diagnosis equipment 100.

[0048] First, the design information acquisition unit 101 acquires design information of the equipment, and the construction condition acquisition unit 102 acquires construction conditions of the equipment (step 1-1). As described above, the design information acquisition unit 101 acquires one piece of design information, and the construction condition acquisition unit 102 acquires multiple construction conditions. The design information acquisition unit 101 stores the acquired design information in the construction condition holding unit 107. The construction condition acquisition unit 102 also stores the acquired construction conditions in the construction condition holding unit 107.

[0049] Next, the verification data generating unit 103 generates verification data for each construction condition (step 1-2). Then, the matching data generating unit 103 stores the generated matching data in the matching data holding unit 108.

[0050] Next, the sensor data acquisition unit 104 acquires sensor data of the diagnosis target device 200 (step 1-3). Then, the sensor data acquisition unit 104 outputs the sensor data to the state estimation unit 105 (step 1-4).

[0051] Next, the state estimation unit 105 compares the sensor data with the collation data (step 1-5) and calculates the similarity (step 1-6).

[0052] Next, the state estimation unit 105 estimates the "state" associated with the collated data with the highest similarity in the tree information of FIG. 2 as the state of the diagnosis target device 200 (step 1-7). When the tree information in FIG. 5 is used, if the matching data with the highest similarity is associated with "abnormality" in the tree information in FIG. 5, the state estimation unit 105 estimates the "cause of abnormality" associated with the "abnormality" as the cause of the abnormality.

[0053] Next, the state estimation unit 105 notifies the diagnosis result output unit 106 of the state of the diagnosis target device 200 estimated in step 1-7 as the diagnosis result (step 1-9). Furthermore, if the state estimation unit 105 also estimates the cause of the abnormality in step 1-7, it also notifies the diagnosis result output unit 106 of the estimated cause of the abnormality as the diagnosis result.

[0054] Finally, the diagnosis result output unit 106 presents the diagnosis result of the state estimation unit 105 to the user 300 (step 1-9).

[0055] ***Explanation of the effect of the embodiment*** In this manner, in this embodiment, collation data is generated for each of a plurality of construction conditions. Therefore, according to this embodiment, the state of equipment can be accurately estimated regardless of the construction conditions under which the equipment was constructed, without the need for redundant processing such as generating a model for each construction condition.

[0056] As mentioned above, this embodiment is based on the premise that there is only one piece of design information. If there are two or more pieces of design information for one model, the tree information shown in FIG. 7 is used. In the tree information in Fig. 7, the "design information" nodes include a "design information 1" node and a "design information 2" node. Also, the "construction conditions" nodes include a "construction conditions (I)" node and a "construction conditions (II)" node. In the example of FIG. 7, "Construction Conditions (I)" and "Construction Conditions (II)" are placed under the node "Design Information 1." Under "Construction Conditions (I)," "Verification Data 1-(I)-1," "Verification Data 1-(I)-2," and "Verification Data 1-(I)-3" are placed. Under "Verification Data 1-(I)-1," "Normal" is placed. Under "Verification Data 1-(I)-2," "Abnormal α" is placed. Under "Verification Data 1-(I)-3," "Abnormal β" is placed. Although not shown in FIG. 7, under "Construction Conditions (II)," "Verification Data 1-(II)-1," "Verification Data 1-(II)-2," and "Verification Data 1-(II)-3" are placed. Under "Verification Data 1-(II)-1," "Normal" is placed. Under "Verification Data 1-(II)-2," "Abnormal α" is placed. "Abnormal β" is placed below "Matching Data 1-(II)-3." In addition, nodes for "construction conditions (I)" and "construction conditions (II)" are placed below the node for "design information 2." Under "construction conditions (I)," "verification data 2-(I)-1," "verification data 2-(I)-2," and "verification data 2-(I)-3" are placed. Under "verification data 2-(I)-1," "normal" is placed. Under "verification data 2-(I)-2," "abnormal α" is placed. Under "verification data 2-(I)-3," "abnormal β" is placed. Although not shown in FIG. 7, under "construction conditions (II)," "verification data 2-(II)-1," "verification data 2-(II)-2," and "verification data 2-(II)-3" are placed. Under "verification data 2-(II)-1," "normal" is placed. Under "verification data 2-(II)-2," "abnormal α" is placed. "Abnormal β" is placed below "Matching Data 2-(II)-3."

[0057] 7 is used, the tree information shown in Fig. 7 is stored in the collation data storage unit 108. The state estimation unit 105 estimates the state of the diagnosis target device 200 by referring to the tree information shown in Fig. 7.

[0058] Furthermore, the tree information in Fig. 5 may be combined with the tree information in Fig. 7. That is, a "Cause of Abnormality" node may exist below the "Abnormality" node in the tree information in Fig. 7, as in Fig. 5.

[0059] Embodiment 2 In the fault diagnosis device 100 according to the first embodiment, if the operation conditions cannot be acquired, the verification data cannot be generated, and therefore the state of the diagnosis target device 200 cannot be diagnosed. Furthermore, in the fault diagnosis device 100 according to the first embodiment, if only some of the operation conditions can be acquired, the collation data will be insufficient, which may result in insufficient diagnostic accuracy.

[0060] Fault diagnosis device 100 according to this embodiment generates assumed installation conditions that are assumed when installing equipment as assumed installation conditions. Then, fault diagnosis device 100 generates collation data in correspondence with a plurality of installation conditions including the assumed installation conditions, thereby solving the above-mentioned problem.

[0061] In this embodiment, differences from the first embodiment will be mainly described. The matters not explained below are the same as those in the first embodiment.

[0062] ***Configuration Description*** FIG. 8 shows an example of the functional configuration of fault diagnosis equipment 100 according to this embodiment. Compared to FIG. 1, in FIG. 8, an assumed construction condition generating unit 109 is added.

[0063] The assumed construction condition generating unit 109 generates assumed conditions when constructing the equipment as assumed construction conditions. When the construction condition acquisition unit 102 cannot acquire the construction conditions or when the construction condition acquisition unit 102 can only acquire some of the construction conditions, the assumed construction condition generation unit 109 generates assumed construction conditions to compensate for the missing construction conditions. The assumed construction condition generating unit 109 generates construction conditions assumed when constructing equipment based on the design information. Then, the assumed construction condition generating unit 109 stores the generated assumed construction conditions in the construction condition holding unit 107. If the construction condition acquisition unit 102 has acquired even a small number of construction conditions, the assumed construction condition generation unit 109 may use the construction conditions acquired by the construction condition acquisition unit 102 to generate assumed construction conditions. The functions of the assumed construction condition generating unit 109 are also realized by a program, similar to the design information acquiring unit 101 etc. The processor 901 executes the program that realizes the functions of the assumed construction condition generating unit 109.

[0064] In this embodiment, the construction condition holding unit 107 holds a plurality of construction conditions including the assumed construction conditions generated by the assumed construction condition generating unit 109 . In other words, if the construction condition acquisition unit 102 is able to acquire even a small number of construction conditions, the construction condition storage unit 107 stores the construction conditions that the construction condition acquisition unit 102 has acquired and the assumed construction conditions generated by the assumed construction condition generation unit 109.

[0065] In this embodiment, the matching data generating unit 103 generates matching data for a plurality of construction conditions including the assumed construction conditions generated by the assumed construction condition generating unit 109 . When generating the matching data, the matching data generating unit 103 does not need to distinguish between the construction conditions acquired by the construction condition acquiring unit 102 and the assumed construction conditions generated by the assumed construction condition generating unit 109.

[0066] The tree information used in this embodiment is the tree information shown in any one of Fig. 2, Fig. 5, and Fig. 7. The tree information used in this embodiment may also be tree information obtained by combining the tree information in Fig. 5 and the tree information in Fig. 7. However, in this embodiment, at least one of the construction conditions included in the tree information is an assumed construction condition generated by the assumed construction condition generating unit 109. That is, taking the tree information in FIG. 2 as an example, at least one of "construction condition (I)" and "construction condition (II)" is an assumed construction condition.

[0067] ***Explanation of Operation*** Next, an example of the operation of the fault diagnosis equipment 100 according to this embodiment will be described. FIG. 9 is a flowchart showing an example of the operation of fault diagnosis equipment 100 according to this embodiment.

[0068] First, the design information acquisition unit 101 acquires the design information of the equipment, and the construction condition acquisition unit 102 acquires the construction conditions of the equipment (step 2-1). In this embodiment, the construction condition acquisition unit 102 may not be able to acquire some or all of the construction conditions. The construction condition acquisition unit 102 notifies the assumed construction condition generation unit 109 that some or all of the construction conditions could not be acquired.

[0069] Next, the assumed construction condition generating unit 109 generates assumed construction conditions (step 2-2).

[0070] Next, the matching data generating unit 103 generates matching data (step 2-3). As described above, the matching data generating unit 103 generates matching data for each of a plurality of construction conditions including the assumed construction conditions generated by the assumed construction condition generating unit 109.

[0071] The processing from step 2-4 onwards is the same as the processing from step 1-3 onwards in Fig. 6. Therefore, the description of these processing steps will be omitted.

[0072] ***Explanation of the effect of the embodiment*** In this embodiment, assumed construction conditions are generated, and verification data is generated using the assumed construction conditions. Therefore, according to this embodiment, even if some or all of the construction conditions cannot be obtained, the verification data generated using the assumed construction conditions can accurately estimate the state of the equipment, regardless of the construction conditions under which the equipment was constructed.

[0073] Embodiment 3 In the first and second embodiments, the collation data corresponding to all the construction conditions is prepared in advance. Therefore, in the configurations of the first and second embodiments, the storage resources are strained by the large amount of collation data. Furthermore, in the first and second embodiments, when sensor data is acquired, the sensor data is compared with all of the matching data. Therefore, in the configurations of the first and second embodiments, the sensor data is compared with a large amount of matching data, which puts a strain on the computational resources.

[0074] In this embodiment, fault diagnosis device 100 holds reference matching data that serves as a reference for generating matching data. Fault diagnosis device 100 also holds, for each construction condition, a generation rule for generating matching data for each construction condition from the reference matching data. When sensor data is acquired from the diagnosis target device 200, the fault diagnosis device 100 analyzes the sensor data and selects one of a plurality of construction conditions as a selected construction condition. Next, the fault diagnosis device 100 applies a generation rule corresponding to the selected construction condition to the reference matching data to generate a matching rule for the selected construction condition. The fault diagnosis device 100 then matches the generated matching rule with the sensor data to estimate the state of the diagnosis target device 200. Fault diagnosis equipment 100 according to this embodiment operates in this manner, and is therefore able to solve the above-mentioned problems.

[0075] In this embodiment, differences from the second embodiment will be mainly described. The matters not explained below are the same as those in the second embodiment.

[0076] ***Configuration Description*** FIG. 10 shows an example of the functional configuration of fault diagnosis equipment 100 according to this embodiment. 10, compared to Fig. 8, a generation rule generation unit 110, a reference matching data storage unit 111, and a generation rule storage unit 112 are added, and the matching data storage unit 108 is removed.

[0077] The generation rule generating unit 110 generates reference matching data and a generation rule for each construction condition.

[0078] As described above, the reference matching data is data that serves as a reference for generating matching data. The generation rule generation unit 110 generates the reference matching data in the following procedure.

[0079] First, the generation rule generating unit 110 designates one of a plurality of construction conditions as a designated construction condition. Furthermore, the generation rule generating unit 110 specifies one of the multiple states of the specified construction conditions as a specified state. The generation rule generation unit 110, for example, designates "construction condition (I)" as the designated construction condition out of "construction condition (I)" and "construction condition (II)" shown in Fig. 2. Furthermore, the generation rule generation unit 110 designates, for example, "normal" as the designated state out of "normal," "abnormal α," and "abnormal β" in "construction condition (I)" shown in Fig. 2. Then, the generation rule generating unit 110 generates the matching data of the specified state of the specified construction conditions as the reference matching data. When the generation rule generation unit 110 specifies "construction condition (I)" as the specified construction condition and "normal" as the specified state, the generation rule generation unit 110 generates simulated data of "matching data (I)-1" corresponding to "construction condition (I)" and "normal" as reference matching data. In other words, the generation rule generation unit 110 generates data that simulates the sensor data of each of sensors X, Y, and Z in the "matching data (I)-1" shown in FIG. 3 as reference matching data.

[0080] The generation rule for each construction condition is a rule for generating the matching data for each construction condition from the reference matching data. In this embodiment, the generation rule is made up of a construction condition relational expression and a state relational expression. The construction condition relational expression is a conversion formula for generating, from the reference matching data, matching data corresponding to the same state as the specified state of the reference matching data for the construction condition for which the generation rule is to be generated. The state relational expression is a conversion expression for generating, from the verification data generated using the construction condition relational expression, verification data corresponding to a state different from the designated state. The generation rule generating unit 110 generates a construction condition relational expression and a state relational expression for each construction condition other than the designated construction condition.

[0081] Assume that the generation rule generation unit 110 generates a generation rule for "construction condition (II)" shown in FIG. Here, as described above, it is assumed that the generation rule generation unit 110 has designated "construction condition (I)" as the designated construction condition and "normal" as the designated state. In other words, it is assumed that the generation rule generation unit 110 has generated, as reference matching data, data simulating the sensor data of each of sensors X, Y, and Z in "matching data (I)-1" shown in Fig. 3. The generation rule generating unit 110 generates, from the reference matching data, a conversion formula for generating matching data corresponding to "matching condition (II)" and "normal" as a matching condition relational formula. In addition, the generation rule generation unit 110 generates a conversion formula for generating matching data corresponding to "construction condition (II)" and "abnormality γ" from matching data corresponding to "construction condition (II)" and "normality" as a state relational formula for "abnormality γ." Furthermore, the generation rule generation unit 110 generates a conversion formula for generating matching data corresponding to "construction condition (II)" and "abnormality δ" from matching data corresponding to "construction condition (II)" and "normality" as a state relational formula for "abnormality δ."

[0082] In addition, for "construction condition (I)", the generation rule generation unit 110 generates a conversion formula for generating the matching data corresponding to "construction condition (I)" and "abnormality α" from the reference matching data as a state relational formula for "abnormality α". Furthermore, the generation rule generating unit 110 generates a transformation formula for generating the matching data corresponding to the "construction condition (I)" and the "abnormality β" from the reference matching data as a state relational formula for the "abnormality β."

[0083] The generation rule generation unit 110 stores the generated reference matching data in the reference matching data storage unit 111 . Furthermore, the generation rule generating unit 110 stores the generated generation rules (the construction condition relational expressions and the state relational expressions) for each construction condition in the generation rule holding unit 112.

[0084] The reference matching data storage unit 111 stores the reference matching rules.

[0085] The generation rule storage unit 112 stores a generation rule for each construction condition.

[0086] In this embodiment, when the state estimation unit 105 acquires sensor data from the sensor data acquisition unit 104 , the state estimation unit 105 outputs the sensor data to the matching data generation unit 103 .

[0087] The matching data generating unit 103 acquires the reference matching data from the reference matching data holding unit 111 .

[0088] Next, the matching data generation unit 103 compares the sensor data with the reference matching data and determines which construction condition generation rule (construction condition relational equation, state relational equation) should be applied to the reference matching data to obtain matching data that is most similar to the sensor data. Then, the matching data generating unit 103 selects the construction conditions that will yield matching data that is most similar to the sensor data. Here, the construction conditions selected by the matching data generating unit 103 are referred to as selected construction conditions.

[0089] The matching data generating unit 103 acquires the generation rules for the selected construction conditions from the generation rule storage unit 112 . Then, the matching data generating unit 103 applies the generation rules for the selected construction conditions to the reference matching data to generate matching data for the selected construction conditions. Specifically, the matching data generating unit 103 applies the construction condition relational equation of the selected construction condition to the reference matching data, and generates matching data corresponding to the selected construction condition and the same state as the specified state. Furthermore, the collation data generating unit 103 applies the state relational equation of the selected construction conditions to the generated collation data, and generates collation data corresponding to the selected construction conditions and states different from the designated state.

[0090] For example, if the selected construction condition is "construction condition (II)" shown in Figure 2, the matching data generation unit 103 applies the construction condition relational equation for "construction condition (II)" to the reference matching data to generate matching data (matching data (II)-1) corresponding to "construction condition (II)" and "normal." Furthermore, the matching data generation unit 103 applies the state relational equation for the selected construction condition "abnormality γ" to the generated "matching data (II)-1" to generate matching data (matching data (II)-2) corresponding to the "construction condition (II)" and "abnormality γ". In addition, the matching data generation unit 103 applies the state relational equation for the selected construction condition "abnormality δ" to the generated "matching data (II)-1" to generate matching data (matching data (II)-3) corresponding to the "construction condition (II)" and the "abnormality δ".

[0091] Then, the verification data generating unit 103 outputs the verification data of the selected construction conditions to the state estimating unit 105.

[0092] In addition, when the matching data generation unit 103 selects the "construction condition (I)" shown in Figure 2 as the selected construction condition, the matching data generation unit 103 applies the state relation equation for "abnormality α" of the "construction condition (I)" to the reference matching data, and generates matching data (matching data (I)-2) corresponding to the "construction condition (I)" and the "abnormality α". Similarly, the matching data generation unit 103 applies the state relation equation for "abnormality β" of "construction condition (I)" to the reference matching data to generate matching data (matching data (I)-3) corresponding to "construction condition (I)" and "abnormality β."

[0093] The state estimation unit 105 acquires the matching data of the selected construction conditions from the matching data generation unit 103, matches (compares) the sensor data with the matching rules of the selected construction conditions, and estimates the state (and cause of abnormality) of the diagnosis target device 200.

[0094] ***Explanation of Operation**+ Next, an example of the operation of the fault diagnosis equipment 100 according to this embodiment will be described. FIG. 11 is a flowchart showing an example of the operation of fault diagnosis equipment 100 according to this embodiment.

[0095] Step 3-1 is the same as step 2-1 in Figure 9. Also, step 3-2 is the same as step 2-2 in Figure 9. Therefore, the explanation of steps 3-1 and 3-2 will be omitted.

[0096] Next, the generation rule generation unit 110 generates reference matching data (step 3-3). As described above, the generation rule generation unit 110 specifies the specified construction conditions and the specified state, and then generates the matching data corresponding to the specified state of the specified construction conditions as the reference matching data.

[0097] Next, the generation rule generating unit 110 generates a generation rule for each construction condition (step 3-4). As described above, the generation rule generating unit 110 generates a construction condition relational expression and a state relational expression as generation rules for each construction condition.

[0098] Step 3-5 is the same as step 2-4 in Figure 9. Also, step 3-6 is the same as step 2-5 in Figure 9. Therefore, the explanation of steps 3-5 and 3-6 will be omitted.

[0099] After step 3-6, the state estimation unit 105 outputs the sensor data to the matching data generation unit 103. Then, the matching data generation unit 103 compares the sensor data with the reference matching data and selects construction requirements (step 3-7). As described above, the matching data generation unit 103 selects construction conditions that will yield matching data that is most similar to the sensor data. The construction conditions selected by the matching data generation unit 103 are the selected construction conditions. The collation data generating unit 103 acquires the rules for generating the selected construction conditions.

[0100] Next, the matching data generating unit 103 applies the generation rules of the selected construction conditions to the reference matching data to generate matching data (step 3-8). Then, the matching data generating unit 103 outputs the generated matching data to the state estimating unit 105.

[0101] Next, the state estimation unit 105 compares the verification data acquired from the verification data generation unit 103 with the sensor data, and estimates the state of the diagnosis target device 200 (step 3-9). In other words, the state estimation unit 105 estimates the state of the diagnosis target device 200 as the state corresponding to the simulation data that is most similar to the sensor data among the normal simulation data and abnormal simulation data included in the matching data acquired from the matching data generation unit 103.

[0102] Step 3-10 is the same as step 2-9 in Figure 9. Also, step 3-11 is the same as step 2-10 in Figure 9. Therefore, the explanation of steps 3-10 and 3-11 will be omitted.

[0103] ***Explanation of the effect of the embodiment***

[0104] In this embodiment, fault diagnosis equipment 100 only holds reference matching data and generation rules, and does not need to hold a large amount of matching data as in the first and second embodiments. In this embodiment, when the fault diagnosis device 100 acquires sensor data, it selects a generation rule and applies the selected generation rule to reference matching data to generate matching data. Then, the fault diagnosis device 100 simply matches the sensor data with the generated matching data. Therefore, there is no need to match the sensor data with a large number of matching data sets, as in the first and second embodiments. Therefore, according to this embodiment, it is possible to avoid a situation in which storage resources are strained due to holding a large amount of collation data, as in the first and second embodiments. Furthermore, according to this embodiment, it is possible to avoid a situation in which computation resources are strained by matching sensor data with a large amount of matching data, as in the first and second embodiments.

[0105] Although the first to third embodiments have been described above, two or more of these embodiments may be combined and implemented. 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.

[0106] ***Additional hardware configuration information*** Finally, a supplementary explanation of the hardware configuration of the fault diagnosis equipment 100 will be given. A processor 901 shown in FIG. 12 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. 12 is a RAM (Random Access Memory). The auxiliary storage device 903 shown in FIG. 12 is a ROM (Read Only Memory), a flash memory, an HDD (Hard Disk Drive), or the like. The communication device 904 shown in FIG. 12 is an electronic circuit that executes data communication processing. The communication device 904 is, for example, a communication chip or a NIC (Network Interface Card).

[0107] 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 also executes a program that realizes the functions of the design information acquisition unit 101 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 design information acquisition unit 101, etc. is 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 program that realizes the functions of the design information acquisition unit 101, 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 program that realizes the functions of the design information acquisition unit 101, etc. may be distributed.

[0108] Furthermore, at least one of the "units" such as the design information acquisition unit 101 may be read as a "circuit" or a "process" or a "procedure" or a "process" or a "circuitry." Furthermore, the fault diagnosis device 100 may be realized by a processing circuit, such as 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 design information acquisition unit 101 and the like are realized as parts of the processing circuits. In this specification, the term "processing circuitry" refers to a generic concept that encompasses a processor and a processing circuit. That is, a processor and a processing circuit are each specific examples of "processing circuitry." [Explanation of symbols]

[0109] 100 Fault diagnosis device, 101 Design information acquisition unit, 102 Construction condition acquisition unit, 103 Matching data generation unit, 104 Sensor data acquisition unit, 105 State estimation unit, 106 Diagnosis result output unit, 107 Construction condition storage unit, 108 Matching data storage unit, 109 Expected construction condition generation unit, 110 Generation rule generation unit, 111 Reference matching data storage unit, 112 Generation rule storage unit, 200 Diagnosis target device, 300 User, 901 Processor, 902 Main memory device, 903 Auxiliary memory device, 904 Communication device, 905 Input / output device.

Claims

1. a matching data generating unit that generates normal simulation data, which is data that simulates sensor data of the equipment in a normal state, and abnormal simulation data, which is data that simulates sensor data of the equipment in an abnormal state, as matching data used to estimate the state of the equipment, in correspondence with construction conditions that are conditions when the equipment is constructed; a sensor data acquisition unit that acquires sensor data of a diagnosis target device that has been installed; a state estimation unit that compares the sensor data with the comparison data, determines which of the normal simulation data and the abnormal simulation data is most similar to the sensor data, and estimates the state of the device to be diagnosed based on the simulation data that is most similar to the sensor data.

2. The matching data generation unit The data processing device according to claim 1 , wherein the sensor data acquisition unit generates a plurality of pieces of collation data in association with a plurality of construction conditions before acquiring the sensor data.

3. The matching data generation unit generating a plurality of pieces of matching data, each of which is associated with a state of the device and a cause of the state; The state estimation unit The data processing device according to claim 1 , wherein the sensor data is compared with the plurality of comparison data, and the state of the device to be diagnosed and the cause of the state are estimated.

4. The data processing device further comprises: an assumed construction condition generating unit that generates assumed conditions when constructing the equipment as assumed construction conditions; The matching data generation unit The data processing device according to claim 1 , wherein a plurality of pieces of collation data are generated in association with a plurality of construction conditions including the assumed construction conditions.

5. The matching data generation unit:

2. The data processing device according to claim 1, wherein after the sensor data acquisition unit acquires the sensor data, the sensor data acquisition unit analyzes the sensor data, selects one of a plurality of construction conditions, and generates the matching data in correspondence with the selected construction condition, which is the selected construction condition.

6. The data processing device further comprises: a reference matching data holding unit that holds reference matching data that is used as a reference for generating the matching data by the matching data generating unit; a generation rule storage unit that stores a generation rule for generating the matching data from the reference matching data for each construction condition; The matching data generation unit The data processing device according to claim 5 , wherein the generation rule of the selected construction condition is applied to the reference matching data to generate the matching data.

7. The reference matching data storage unit The reference matching data is stored as the reference matching data, and the reference matching data corresponds to a specified construction condition, which is a construction condition specified from the plurality of construction conditions, and a specified state, which is a state specified from a plurality of states that the device can be in, The generation rule storage unit For each construction condition other than the specified construction condition among the plurality of construction conditions, a construction condition relational expression for generating, from the reference verification data, verification data corresponding to a state identical to the specified state, and a state relational expression for generating, from the verification data generated using the construction condition relational expression, verification data corresponding to a state different from the specified state are held as the generation rule; The matching data generation unit 7. A data processing device according to claim 6, wherein the construction condition relational expression of the selected construction conditions is applied to the reference matching data to generate matching data corresponding to the selected construction conditions and a state that is the same as the specified state, and the state relational expression of the selected construction conditions is applied to the generated matching data to generate matching data corresponding to the selected construction conditions and a state that is different from the specified state.

8. The computer generates normal simulation data, which is data simulating sensor data of the equipment in a normal state, and abnormal simulation data, which is data simulating sensor data of the equipment in an abnormal state, as a plurality of pieces of matching data used to estimate the state of the equipment, corresponding to construction conditions, which are conditions under which the equipment is constructed; The computer acquires sensor data of the diagnosis target device, which is an installed device, The computer compares the sensor data with the plurality of comparison data, determines which of the normal simulation data and the abnormal simulation data is most similar to the sensor data, and estimates the state of the device to be diagnosed based on the simulation data that is most similar to the sensor data.

9. a matching data generation process for generating normal simulation data, which is data simulating sensor data of the equipment in a normal state, and abnormal simulation data, which is data simulating sensor data of the equipment in an abnormal state, as a plurality of matching data used to estimate the state of the equipment, in correspondence with installation conditions, which are conditions when installing the equipment; A sensor data acquisition process for acquiring sensor data of the diagnosis target device, which is an installed device; a data processing program that causes a computer to execute a state estimation process of collating the sensor data with the plurality of comparison data, determining which of the normal simulation data and the abnormal simulation data is most similar to the sensor data, and estimating the state of the device to be diagnosed based on the simulation data that is most similar to the sensor data.

Images