Fault prediction system

Abstract

A failure prediction system (1) has: an operation data acquisition section (201) that acquires operation data of a hydraulic excavator (10); an inspection data acquisition section (202) that acquires inspection data of the hydraulic excavator (10); a parts replacement repair data acquisition section (203) that acquires parts replacement repair data of the hydraulic excavator (10); and a failure prediction section (205) that predicts a failure probability of each portion of the hydraulic excavator (10) based on the operation data acquired by the operation data acquisition section (201), the inspection data acquired by the inspection data acquisition section (202), and the parts replacement data acquired by the parts replacement repair data acquisition section (203).

Description

Technical Field

[0001] This invention relates to a fault prediction system for predicting faults in hydraulic excavators and other working machinery. Background Technology

[0002] In fault prediction for operating machinery, the probability of a fault varies depending on the operating environment and frequency of use due to the diverse range of parts involved. Therefore, various technologies have been proposed. For example, Patent Document 1 discloses a fault prediction system that improves prediction accuracy by learning while comparing predicted fault values ​​with actual performance (whether or not an event occurred) over a specified period.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2016-157280 Summary of the Invention

[0006] However, in the fault prediction system described in Patent Document 1, the comparison is based on whether the predicted event occurs within a specified period, rather than on the predicted fault value. Therefore, actual performance outside the specified period, including emergency response, is not taken into account. Consequently, it is difficult to improve the prediction accuracy.

[0007] This invention was made to solve this technical problem, and its purpose is to provide a fault prediction system that can improve the accuracy of fault prediction.

[0008] The fault prediction system of the present invention predicts faults in operating machinery, characterized by comprising: an operation information acquisition unit for acquiring operation information of the operating machinery; an inspection information acquisition unit for acquiring inspection information of the operating machinery; a component replacement and repair information acquisition unit for acquiring component replacement and repair information of the operating machinery; and a fault prediction unit that predicts the fault probability of each part of the operating machinery based on the operation information acquired by the operation information acquisition unit, the inspection information acquired by the inspection information acquisition unit, the component replacement and repair information acquired by the component replacement and repair information acquisition unit, and the deviation information between the fault probability of each part of the operating machinery and the actual inspection results of the operating machinery stored in the storage unit.

[0009] In the fault prediction system of the present invention, the fault prediction unit predicts the fault probability of each part of the working machine based on the operation information obtained by the operation information acquisition unit, the inspection information obtained by the inspection information acquisition unit, the component replacement and repair information obtained by the component replacement and repair information acquisition unit, and the deviation information between the fault probability of each part of the working machine and the actual inspection results of the working machine stored in the storage unit. By further incorporating inspection information, component replacement and repair information, and the deviation information between the fault probability and the actual inspection results into the operation information of the working machine to predict the fault probability, the accuracy of fault prediction can be improved.

[0010] Invention Effects

[0011] According to the present invention, the accuracy of fault prediction can be improved. Attached Figure Description

[0012] Figure 1 This is a schematic diagram illustrating the configuration of a fault prediction system according to an implementation method.

[0013] Figure 2 This is a block diagram illustrating the fault prediction system implemented in this way.

[0014] Figure 3 This is a graph representing an example of failure probability data.

[0015] Figure 4 This is a diagram representing an example of inspection item data.

[0016] Figure 5 This is a diagram representing an example of key component data.

[0017] Figure 6 This is a flowchart representing the server's control and processing.

[0018] Figure 7 This is a flowchart representing the control and processing of the portable terminal's inspection item screen.

[0019] Figure 8 This is a flowchart illustrating the control and processing of the main components' screens on a portable terminal.

[0020] Figure 9 This is a flowchart representing the control processing of the fault probability screen on a portable terminal.

[0021] Figure 10 This is an example of a screen showing the inspection items on a portable terminal.

[0022] Figure 11 This is an example of a screen showing the main components of a portable terminal.

[0023] Figure 12This is an example of a screen displaying the probability of failure of a portable terminal.

[0024] Figure 13 This is a schematic diagram representing a variation of the fault prediction system.

[0025] Figure 14 This is a schematic diagram representing a variation of the fault prediction system. Detailed Implementation

[0026] Hereinafter, embodiments of the fault prediction system of the present invention will be described with reference to the accompanying drawings. In the description of the drawings, the same reference numerals are used for the same elements, and repeated descriptions are omitted. Furthermore, the following description uses a hydraulic excavator as an example of the work machinery to be subject to fault prediction; however, the work machinery is not limited to hydraulic excavators and may also include wheel loaders and other work machinery.

[0027] Additionally, the following description uses terms such as "part" and "component." The scope of "part" is generally broader than that of "component." For example, the "front part," which is a part, consists of components such as the boom, stick, and bucket. However, in the case of "engine," it can be either a part or a component. Therefore, depending on the context, sometimes "part" and "component" refer to the same object.

[0028] Figure 1 This is a schematic diagram illustrating the configuration of the fault prediction system according to the implementation method. Figure 2 This is a block diagram illustrating a fault prediction system according to an embodiment. The fault prediction system 1 of this embodiment includes: a plurality of hydraulic excavators 10; a server 20 configured to communicate with these hydraulic excavators 10 via a network 40; and a portable terminal 30 configured to communicate with the server 20 and the hydraulic excavators 10 via the network 40.

[0029] [About Hydraulic Excavators]

[0030] The hydraulic excavator 10 is manufactured at the factory and used by designated users (e.g., the owner or user of the hydraulic excavator) at work sites where civil engineering, construction, dismantling, dredging, or similar operations are carried out. The hydraulic excavator 10 is equipped with an operation data collection unit 101. This operation data collection unit 101, for example, is part of a control unit that controls the hydraulic excavator 10 as a whole, and sends the operation data of the hydraulic excavator 10 along with its own identification number to a server 20 and a portable terminal 30. Furthermore, it sends the data to the server 20 and the portable terminal 30 periodically (e.g., once a day).

[0031] [About portable devices]

[0032] The portable terminal 30, such as a smartphone, tablet, mobile phone, or PDA (Personal Data Assistant), is carried by the maintenance personnel or user of the hydraulic excavator 10. The maintenance personnel or user can request information such as inspection items for a specific hydraulic excavator 10 via the portable terminal 30. Specifically, when the maintenance personnel or user requests information such as inspection items for a specific hydraulic excavator 10, they can directly input the identification number of that specific hydraulic excavator 10 into the input section 304 of the portable terminal 30, or select the identification number of that specific hydraulic excavator 10 displayed in the input section 304, thereby requesting the server 20 via the portable terminal 30.

[0033] like Figure 2 As shown, the portable terminal 30 includes an inspection item screen 301, a main component screen 302, a fault probability screen (display unit) 303, and an input unit 304. The inspection item screen 301, the main component screen 302, and the fault probability screen 303 are switched according to the selection of maintenance personnel and users. These screens will be described in detail below.

[0034] The input unit 304 is used to receive input from maintenance personnel or users, such as a touch panel displayed on the portable terminal 30 and a keyboard. As described above, when maintenance personnel or users want information such as specific inspection items of the hydraulic excavator 10, they can enter an identification number into the input unit 304, thereby making a request to the server 20.

[0035] In addition, for example, when maintenance personnel perform any inspection work on the hydraulic excavator 10, they can input the inspection information (e.g., inspection results) and identification number of the hydraulic excavator 10 into the input unit 304, thereby sending the inspection information to the server 20.

[0036] Furthermore, for example, if maintenance personnel replace or repair any component of the hydraulic excavator 10, they can obtain a work association number via the input unit 304 as a management number accompanying the replacement or repair of that component. At this time, data related to the work association number is sent to the server 20 as component replacement and repair information.

[0037] Furthermore, for example, when an operator performs an inspection based on the failure probability predicted by the failure prediction unit 205 (described later), the actual inspection results along with the identification number of the hydraulic excavator 10 are input into the input unit 304 and sent to the server 20. The server 20 compares the actual inspection results obtained from the input results with the predicted values ​​and learns from the results.

[0038] [About the server]

[0039] Server 20 is the main computer constituting the fault prediction system 1. It is located at the head office, branch office, factory, or management center of the manufacturer of the hydraulic excavator 10. It periodically collects operating data from multiple hydraulic excavators 10 and centrally manages these hydraulic excavators 10. Server 20 is, for example, composed of a microcomputer that combines a CPU (Central Processing Unit) for performing calculations, ROM (Read-Only Memory) as a secondary storage device for storing programs used for calculations, and RAM (Random Access Memory) as a temporary storage device for storing calculation processes and temporary control variables. It performs calculations and judgments by executing the stored programs.

[0040] In this embodiment, the server 20 predicts the failure probability of various parts of the hydraulic excavator 10 based on the operation data of the managed object, namely the hydraulic excavator 10, and sends the predicted results to the portable terminal 30. To achieve this, the server 20 in this embodiment has an operation data acquisition unit (operation information acquisition unit) 201, an inspection data acquisition unit (inspection information acquisition unit) 202, a component replacement and repair data acquisition unit (component replacement and repair information acquisition unit) 203, a learning unit 204, a fault prediction unit 205, an inspection item acquisition unit 206, a major component acquisition unit 207, and a data storage unit 210.

[0041] The operation data acquisition unit 201 acquires the operation data sent from the hydraulic excavator 10, and stores and accumulates the acquired operation data together with the identification number of the hydraulic excavator 10 in the data storage unit 210. Figure 2 The operation data history 211 is the operation data of the hydraulic excavator 10 accumulated over a specified period (e.g., from the time it left the factory until now), and is stored in the data storage unit 210. This operation data is data about operation information, such as the start time of operation of the hydraulic excavator 10, the end time of operation, position information, and information detected by each sensor.

[0042] The inspection data acquisition unit 202 acquires the inspection information of the hydraulic excavator 10 and stores the acquired results as inspection data 212 along with the identification number of the hydraulic excavator 10 in the data storage unit 210. The inspection information of the hydraulic excavator 10 is information sent to the input unit 304 by maintenance personnel as described above. The inspection data 212 includes the inspection date and time, inspection items, inspection work content, and inspection results.

[0043] The component replacement and repair data acquisition unit 203 acquires component repair and replacement information related to the aforementioned "operation association number," and stores the acquired component repair and / or replacement results as component replacement and repair data 213 along with the identification number of the hydraulic excavator 10 in the data storage unit 210. The component replacement and repair data 213 includes component repair details, component replacement details, repair date and time, and replacement date and time.

[0044] Learning Department 204 performs machine learning using known methods such as decision trees and gradient boosting algorithms, which are its evolutionary forms. Machine learning involves using operational data history of other hydraulic excavators collected or accumulated at the desired prediction point as explanatory variables and business data including either inspection / repair history or component replacement history implemented relative to these hydraulic excavators as objective variables, and then creating a fault model that derives the probability of failure. Furthermore, inspection / repair history and component replacement history are objective variables, representing the fact of failure. Therefore, if the operational data history includes failure history, inspection / repair history and component replacement history are not needed. On the other hand, if the operational data history does not include failure history, inspection / repair history and component replacement history are necessary. In other words, when generating the fault model described later, data related to inspection / repair history and component replacement history are referenced depending on whether failure history is included in the operational data history.

[0045] The fault prediction unit 205 predicts the failure probability of each part of the hydraulic excavator 10 based on the operational data history 211 collected from the hydraulic excavator 10 under normal circumstances, the inspection data 212 fed back from the portable terminal when performing inspections and repairs on any hydraulic excavator 10, and the component replacement and repair data 213. The prediction results are stored in the data storage unit 210 as fault probability data 215 along with the identification number of the hydraulic excavator 10. In addition, the fault probability data is provided to the learning unit 204 for machine learning.

[0046] As a failure probability data 215, for example, Figure 3 The diagram shows the classification based on model, manufacturing number, main components, and type. The predicted failure probability is expressed as a percentage; the higher the number, the higher the probability of failure (in other words, the higher the urgency of the inspection). Furthermore, the types are categorized as "Yearly," "Sudden," and "Combined." "Yearly" indicates the probability of a time-related failure due to factors such as wear and tear; "Sudden" indicates the probability of signs of failure appearing based on recent operating conditions; and "Combined" considers both the "Yearly" and "Sudden" failure probabilities.

[0047] The inspection item acquisition unit 206, based on a request from the portable terminal 30, acquires the inspection item data 216 of the hydraulic excavator 10 stored in the data storage unit 210 and sends it to the portable terminal 30. As an example of the inspection item data 216, for example... Figure 4 As shown, items such as exhaust pipe, turbocharger, fuel filter, boom, etc. can be included, and these inspection items are divided according to the model.

[0048] The main component acquisition unit 207 acquires the main component data 217 of the hydraulic excavator 10 stored in the data storage unit 210 according to a request from the portable terminal 30, and sends it to the portable terminal 30. An example of the main component data 217 is, for example... Figure 5 As shown, they are divided according to machine model and inspection items.

[0049] [Regarding server control and processing]

[0050] The following is for reference Figure 6 Explain the control and processing of server 20. Figure 6 The control process shown is executed repeatedly, for example, according to a prescribed cycle.

[0051] First, in step S101, at least one of the following is acquired: the operation data history, the inspection data history, and the component replacement and repair data history of the hydraulic excavator 10. Furthermore, the information as "history" is not only based on the time the data was acquired, but can also be used to predict based on the history up to that point. The operation data acquisition unit 201 acquires the operation data sent from the hydraulic excavator 10 along with the identification number of the hydraulic excavator 10, and stores the acquired results in the data storage unit 210. When performing an inspection on any hydraulic excavator 10, the inspection data acquisition unit 202 acquires the inspection data sent from the portable terminal 30 along with the identification number of the hydraulic excavator 10, and stores the acquired results in the data storage unit 210. Additionally, as mentioned above, both the inspection data history and the component replacement and repair data history are information related to faults. This information is an indispensable variable in fault prediction; therefore, if it is included in the operation data history, it may not be referenced, but conversely, it will be referenced if it is not included.

[0052] When performing repairs on any hydraulic excavator 10, the component replacement and repair data acquisition unit 203 acquires the component replacement and repair data sent from the portable terminal 30 along with the identification number of the hydraulic excavator 10, and stores the acquired results in the data storage unit 210.

[0053] When the learning unit 204 receives feedback from the portable terminal 30 regarding the inspection and repair results of any hydraulic excavator 10 (acquisition of inspection data history and component replacement and repair data history), it updates the fault model generated based on at least one of the operation data history, the inspection data history stored in the data storage unit 210, and the component replacement and repair data history, and compares it with the fault probability data stored in the fault prediction unit 205. Furthermore, it generates deviation data 214 based on the difference between the fault probability data (predicted value) and the actual inspection result (actual performance value), and stores it in the data storage unit 210 along with the identification number of the hydraulic excavator 10. The deviation data 214 includes deviation content, deviation analysis results, and correction values ​​reflecting the analysis results, and is used for machine learning to predict fault probabilities.

[0054] In step S102, following step S101, a fault probability prediction is performed. At this time, the fault prediction unit 205 predicts the fault probability of each part of the hydraulic excavator 10 using the operation data history, inspection data acquired by the inspection data acquisition unit 202, component replacement and repair data acquired by the component replacement and repair data acquisition unit 203, and the fault model generated by the learning unit 204. Specifically, the fault model at the point in time when a fault is to be predicted is extracted from the data storage unit 210. The operation data history of the hydraulic excavator for which the fault is to be predicted is substituted into this model. The fault probability is then predicted based on the degree of similarity to the actual performance values ​​(operation data history) of hydraulic excavators that have previously experienced faults, and fault probability data is generated.

[0055] In step S103, following step S102, the fault prediction unit 205 stores the predicted result along with the identification number of the hydraulic excavator 10 in the data storage unit 210. This concludes the series of control processes.

[0056] Next, refer to Figures 7 to 12 This section explains the control and processing of each screen on the portable terminal 30.

[0057] [Screenshot of inspection items]

[0058] First, based on Figure 7 Screen 301 shows the inspection items of the portable terminal 30. Figure 7 This is a flowchart showing the control processing of the inspection items screen for portable terminals. Figure 7 The control process shown is performed through the cooperation of the portable terminal 30 and the server 20. More specifically, a portion of step S201 and steps S204 to S209 are performed by the server 20, while the other steps are performed by the portable terminal 30.

[0059] Figure 7 The control process shown begins, for example, when a maintenance worker or user selects an inspection item screen 301 displayed on the portable terminal 30. First, in step S201, the inspection item list is acquired. At this time, for example, a maintenance worker or user requests an inspection item list for the hydraulic excavator 10 from the server 20 by selecting (e.g., touching) the inspection item screen 301. The inspection item acquisition unit 206 of the server 20 acquires the inspection item list from the inspection item data 216 according to the request and sends it to the portable terminal 30.

[0060] Furthermore, the portable terminal 30 receives the inspection item list sent from the server 20 and displays the received content on the inspection item screen 301. The content displayed on the inspection item screen 301 includes, for example,... Figure 10 As shown, a list of inspection items can be compiled, such as "engine", "upper rotating body", and "lower traveling body".

[0061] Next, a cyclical process is performed on the inspection items shown in steps S202 to S210. Specifically, for example, when the maintenance personnel and user select one of the multiple inspection items displayed on the inspection item screen 301 (e.g., ... Figure 10 In the case of the “lower driving body” shown, the portable terminal 30 sends the selected inspection items to the server 20 and requests a list of major components from the server 20 (see step S203).

[0062] In step S204, following step S203, the main component acquisition unit 207 of server 20 acquires a list of main components from main component data 217 according to the request. For example... Figure 10 When the “lower running body” is selected as shown, the main component acquisition unit 207 acquires a list of main components associated with the “lower running body”, such as sprockets, central couplings, bogie frames, side frames, and bogie links.

[0063] Next, for these acquired major components, the cyclical processing of acquiring fault probabilities as shown in steps S205 to S207 is performed. That is, in step S205, the fault probabilities of the acquired major components are acquired sequentially (see step S206). The fault probabilities of the major components are predicted by the fault prediction unit 205 and stored as fault probability data 215 in the data storage unit 210.

[0064] In step S208, following step S207, the fault prediction unit 205 calculates the fault probability of the selected inspection item based on the fault probabilities of the main components obtained from steps S205 to S207. In step S209, following step S208, the server 20 transmits (in other words, sends) the fault probability of the inspection item predicted in step S208 back to the portable terminal 30.

[0065] The cyclic processing of inspection items shown in steps S202 to S210 is performed for each selected inspection item. For example, if "lower driving body" is selected in sequence, followed by "upper rotating body" and "engine", steps S203 to S209 are repeated in the order of "upper rotating body" and "engine".

[0066] In step S211, which follows step S210, the portable terminal 30 displays the received list of inspection items and the failure probability of each inspection item on the inspection item screen 301. Thus, the series of control processes related to the inspection item screen are completed.

[0067] [About the main components]

[0068] Next, based on Figure 8 Screenshots illustrating the main components of a portable terminal. Figure 8 This is a flowchart showing the control processing of the main components of a portable terminal. Figure 8 The control process shown is executed through the cooperation of the portable terminal 30 and the server 20. More specifically, steps S303 to S307 are processes executed by the server 20, while the other steps are processes executed by the portable terminal 30.

[0069] Figure 8 The control process shown begins, for example, based on the selection of the main component screen 302 displayed on the portable terminal 30 by the maintenance personnel or user. First, in step S301, the inspection items are selected. For example, if the maintenance personnel or user... Figure 11 If the "Front Part" inspection item is selected as shown, the portable terminal 30 will send the selected inspection item to the server 20 and request a list of the associated main components from the server 20 (see step S302).

[0070] In step S303, following step S302, the main component acquisition unit 207 of server 20 acquires a list of main components associated with the inspection items from the main component data 217 according to the request. Here, since "front part" is selected, the main component acquisition unit 207 acquires a list of main components such as buckets and linkages associated with "front part".

[0071] Next, for these main components, the loop process of obtaining the failure probability of the main components as shown in steps S304 to S306 is performed. That is, the failure probability of the main components obtained in step S303 is obtained sequentially (see step S305). The failure probability of the main components is predicted by the failure prediction unit 205 and stored in the data storage unit 210 as failure probability data 215.

[0072] In step S307, which follows step S306, server 20 sends back (in other words, transmits) the list of major components associated with the failure probability to portable terminal 30.

[0073] In step S308, following step S307, the portable terminal 30 displays the received list of main components and the failure probability of each component on the main component screen 302. For example, at this time... Figure 11 As shown, the failure probability is displayed for each link. Furthermore, if step S308 ends, the series of control processes related to the main component screen concludes.

[0074] [Screen showing probability of failure]

[0075] Next, based on Figure 9 This screen displays the probability of failure for portable terminals. Figure 9 This is a flowchart showing the control process for displaying the failure probability screen of a portable terminal. Figure 9 The control process shown is executed through the cooperation of the portable terminal 30 and the server 20. More specifically, steps S402 to S404 and steps S407 to S408 are processes executed by the server 20, while the other steps are processes executed by the portable terminal 30.

[0076] First, in step S401, information about the target hydraulic excavator is sent. At this time, for example, maintenance personnel or users input the identification number of the hydraulic excavator 10 for which they want to investigate the probability of failure into the input section 304 of the portable terminal 30, thus requesting the server 20.

[0077] In step S402, following step S401, it is determined whether the specified target hydraulic excavator is included in the managed objects. At this time, the server 20 determines whether there is a hydraulic excavator among the multiple hydraulic excavators being managed that matches the sent hydraulic excavator's identification number. If it is determined that there is no target hydraulic excavator, the control process ends. On the other hand, if it is determined that there is a target hydraulic excavator, the control process proceeds to step S403.

[0078] In step S403, server 20 determines whether it has inspection data for the target hydraulic excavator. If it determines that there is no inspection data, the control process ends. On the other hand, if it determines that there is inspection data, the control process proceeds to step S404.

[0079] In step S404, server 20 determines whether it has component replacement and repair data for the target hydraulic excavator. If it determines that component replacement and repair data for the target hydraulic excavator is not available, the control process ends. On the other hand, if it determines that component replacement and repair data is available, the control process proceeds to step S405.

[0080] In step S405, the main components are selected. If the maintenance personnel or user selects a main component displayed on the portable terminal 30, the portable terminal 30 will send the selected main component to the server 20, requesting a list of associated main components from the server 20 (see step S406). For example, the maintenance personnel or user may select a main component displayed on the portable terminal 30. Figure 11 Select the "bucket, linkage type" associated with "front part" and send it to server 20.

[0081] In step S407, following step S406, server 20 obtains a list of failure probabilities for major components from failure probability data 215 according to the request. In step S408, following step S407, server 20 transmits (in other words, sends) the obtained list of failure probabilities for major components back to portable terminal 30.

[0082] In step S409, following step S408, the portable terminal 30 displays the received list of failure probabilities for major components on the failure probability screen 303. In the failure probability screen 303, for example... Figure 11 As shown, the links in the multiple bucket linkages are categorized as "Link A" or "Link B" and are labeled with their respective model numbers. Furthermore, for each linkage, icons indicating the degree of inspection recommendation are displayed. Additionally, in... Figure 11 For ease of explanation, an alarm bell symbol is used, but it does not have to be an alarm bell if the inspection recommendation can be visually confirmed.

[0083] To make it easier for users to understand, "inspection recommendation" (see reference) is used instead of failure probability. Figure 12In other words, the inspection recommendation rate has the same meaning as the failure probability, and is a representation changed to make it easier for users to understand. Furthermore, the warning bell icon is preferably colored according to the range of the inspection recommendation rate. For example, the warning bell icon is red when the inspection recommendation rate is above 90%, yellow when it is between 60% and 90%, and blue when it is between 40% and 60%. This visual representation can effectively draw the attention of maintenance personnel or users.

[0084] Here, if an alarm icon is selected, the inspection recommendation level for each component it comprises will be further displayed using the alarm icon. Furthermore, if an alarm icon for each component is selected, then... Figure 12 As shown, the inspection recommendation rates for each type (i.e., "comprehensive", "yearly", and "emergency") are displayed through specific numerical values.

[0085] If step S409 ends, then the series of control processes related to the fault probability screen will end.

[0086] In the fault prediction system 1 configured as described above, the fault prediction unit 205 predicts the probability of failure for each part of the hydraulic excavator 10 based on the operation data acquired by the operation data acquisition unit 201, the inspection data acquired by the inspection data acquisition unit 202, and the component replacement and repair data acquired by the component replacement and repair data acquisition unit 203. By incorporating inspection data and component replacement and repair data into the operation data of the hydraulic excavator 10 to predict the failure probability, the accuracy of fault prediction can be improved.

[0087] In addition, deviation data 214 learned by the learning unit 204 is added to the operation data, inspection data and component replacement and repair data of the hydraulic excavator 10. Based on these data, the failure probability of each part of the hydraulic excavator 10 is predicted, thus improving the accuracy of failure prediction.

[0088] Furthermore, in this embodiment, the operation data acquisition unit 201, the inspection data acquisition unit 202, the component replacement and repair data acquisition unit 203, the learning unit 204, and the fault prediction unit 205 are described using the example of being installed on the server 20, but for example, Figure 13As shown, it can also be provided in the hydraulic excavator 10. For example, the inspection data acquisition unit 202, the component replacement and repair data acquisition unit 203, the learning unit 204, and the fault prediction unit 205 can be provided together with the data storage unit 218, which stores the inspection data 212, component replacement and repair data 213, deviation data 214, and fault probability data 215, in the control unit of the hydraulic excavator 10. In addition, in this case, the operation data collection unit 101 functions as the operation data acquisition unit. If this is done, the processing of the server 20 can be reduced, and the versatility of the fault prediction system 1 can be increased. Moreover, it is preferable that the hydraulic excavator 10 is provided with a display unit 102 that displays the fault probability. If this is done, the fault probability can be grasped not only on the portable terminal 30 side but also on the hydraulic excavator 10 side.

[0089] In addition, such as Figure 14 As shown, the operation data acquisition unit (operation information acquisition unit) 220, the learning unit 204, and the fault prediction unit 205 can be located together with the data storage unit 219 in the portable terminal 30. Here, the operation data acquisition unit 220, like the operation data acquisition unit 201 described above, acquires operation data sent from the hydraulic excavator 10, and stores and accumulates the acquired operation data along with the identification number of the hydraulic excavator 10 in the data storage unit 219. The data storage unit 219 stores operation data history 211, inspection data 212, component replacement and repair data 213, deviation data 214, and fault probability data 215. Furthermore, the inspection data 212 and the component replacement and repair data 213 can be directly obtained through input by maintenance personnel. This reduces the processing load on the server 20 and increases the versatility of the fault prediction system 1.

[0090] The embodiments of the present invention have been described in detail above. However, the present invention is not limited to the embodiments described above, and various design changes can be made without departing from the spirit of the present invention as described in the technical solution.

[0091] Explanation of reference numerals in the attached figures

[0092] 1: Fault Prediction System; 10: Hydraulic Excavator; 20: Server; 30: Portable Terminal; 40: Network; 101: Operation Data Collection Unit; 102: Display Unit; 201, 220: Operation Data Acquisition Unit (Operation Information Acquisition Unit); 202: Inspection Data Acquisition Unit (Inspection Information Acquisition Unit); 203: Component Replacement and Repair Data Acquisition Unit (Component Replacement and Repair Information Acquisition Unit); 204: Learning Unit; 205: Fault Prediction Unit; 206: Inspection Item Acquisition Unit; 207: Major Component Acquisition Unit; 210, 218, 219: Data Storage Unit; 211: Operation Data History; 212: Inspection Data; 213: Component Replacement and Repair Data; 214: Deviation Data; 215: Fault Probability Data; 216: Inspection Item Data; 217: Major Component Data; 301: Inspection Item Screen; 302: Major Component Screen; 303: Fault Probability Screen (Display Unit); 304: Input Unit.

Claims

1. A fault prediction system for predicting faults in operating machinery, characterized in that, have: An operation information acquisition unit that acquires the operation information of the operating machinery; Inspection information acquisition unit for acquiring inspection information of the operating machinery; A component replacement and repair information acquisition unit that acquires component replacement and repair information of the operating machinery; and The fault prediction unit predicts the failure probability of each part of the operating machinery based on the operating information obtained by the operating information acquisition unit, the inspection information obtained by the inspection information acquisition unit, the component replacement and repair information obtained by the component replacement and repair information acquisition unit, and deviation information. The deviation information is deviation data based on the difference between the failure probability of each part of the operating machinery stored in the storage unit and the actual inspection results of the operating machinery, and includes deviation content, deviation analysis results, and correction values ​​reflecting the analysis results.

2. The fault prediction system according to claim 1, characterized in that, It also includes a learning unit that learns the discrepancy information between the failure probability predicted by the failure prediction unit and the actual inspection results of the operating machinery. The fault prediction unit predicts the failure probability of each part of the operating machinery based on the deviation information learned by the learning unit.

3. The fault prediction system according to claim 2, characterized in that, It also has a server capable of communicating with the operating machinery. The operation information acquisition unit, the inspection information acquisition unit, the component replacement and repair information acquisition unit, the fault prediction unit, and the learning unit are located on the server.

4. The fault prediction system according to claim 3, characterized in that, It also has a portable terminal capable of communicating with the server. The portable terminal is provided with a display unit that displays the probability of failure predicted by the failure prediction unit.

5. The fault prediction system according to claim 2, characterized in that, It also has a portable terminal capable of communicating with the operating machinery. The operation information acquisition unit, the inspection information acquisition unit, the component replacement and repair information acquisition unit, the fault prediction unit, and the learning unit are located in the portable terminal.

6. The fault prediction system according to claim 5, characterized in that, The portable terminal is provided with a display unit that displays the probability of failure predicted by the failure prediction unit.

7. The fault prediction system according to claim 2, characterized in that, The operation information acquisition unit, the inspection information acquisition unit, the component replacement and repair information acquisition unit, the fault prediction unit, and the learning unit are located in the working machine.

8. The fault prediction system according to claim 7, characterized in that, The machine is equipped with a display unit that displays the probability of failure predicted by the failure prediction unit.

Images