Machine fault question-answering method and system based on industrial knowledge graph

Abstract

The invention relates to a machine fault questioning and answering method and system based on an industrial knowledge graph, and belongs to the technical field of fault diagnos.The method comprises the steps that maintenance data are obtained regularly, the maintenance data are processed, and a preset knowledge graph is updated based on the processed maintenance data; wherein the knowledge graph and the processed maintenance data at least comprise basic information of a fault object and a corresponding fault solution; fault inquiry information is received, the fault inquiry information is processed, and feedback content corresponding to the processed fault inquiry information is inquired from the knowledge graph; and outputting the feedback content obtained by query, so that a presenter who puts forward the fault inquiry information knows the feedback content. The method has the effect of improving the fault solving efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of fault diagnosis, and in particular to a machine fault question-answering method and system based on an industrial knowledge graph. Background Art

[0002] With the rapid development of industrial automation and intelligent manufacturing technology, the complexity of machinery and equipment continues to increase, and machine fault diagnosis and maintenance have become important links to ensure the normal operation of the production line.

[0003] However, traditional fault diagnosis methods often rely on experienced engineers who use their long-term accumulated experience to make fault judgments and repairs. However, this approach is prone to a situation where supply exceeds demand, that is, the number of faulty machines exceeds the number of engineers. At this time, it is not easy to solve the fault problems of all faulty machines in a timely manner, resulting in low fault solving efficiency, so it needs to be improved. Summary of the invention

[0004] In order to improve the efficiency of troubleshooting, the present application provides a machine fault question and answer method and system based on industrial knowledge graph.

[0005] In the first aspect, the present application provides a machine fault question-answering method based on an industrial knowledge graph, which adopts the following technical solution: Periodically acquiring maintenance data, processing the maintenance data, and updating a preset knowledge graph based on the processed maintenance data; wherein both the knowledge graph and the processed maintenance data include at least basic information of the fault object and its corresponding fault solution; receiving fault inquiry information, processing the fault inquiry information, and searching the knowledge graph for feedback content corresponding to the processed fault inquiry information; The feedback content obtained by the query is outputted so that the person who has submitted the fault inquiry information can know the feedback content.

[0006] By adopting the above-mentioned technical solution, based on the correspondence between the relevant fault objects and fault solutions stored in the knowledge graph, when an industrial machine (i.e., the fault object) fails, the above-mentioned solution proposed in this application can be used to automatically query and feedback content corresponding to the fault inquiry information. The feedback content contains a fault solution, thereby helping to improve the efficiency of fault resolution.

[0007] Optionally, the processed fault inquiry information includes at least the fault object, and the feedback content includes a fault solution; The querying of feedback content corresponding to the processed fault inquiry information from the knowledge graph includes: Determine the importance of the process in which the fault object participates in industrial production and the maintenance urgency of the fault object according to the fault object corresponding to the fault inquiry information; Query the feedback content corresponding to the processed fault inquiry information from the knowledge graph, and adjust the arrangement order of the fault solutions in the feedback content based on the maintenance urgency of the fault object and the maintenance duration of each fault solution in the feedback content; among them, the fault solution with a higher arrangement order is preferentially output.

[0008] By adopting the above technical solution, since industrial machines participate in industrial production and are generally applied to a specific process, when an industrial machine fails, it is likely to affect industrial production efficiency, and even cause the entire industrial production line to be paralyzed. Therefore, this application specifically determines the maintenance urgency according to the importance of the process in which the fault object participates, and sorts the order of the output fault solutions, and the sorting basis is the maintenance duration of each fault solution. For example, the fault solution with a shorter maintenance duration has a higher arrangement order, so as to help speed up the speed of fault resolution.

[0009] Optionally, the processed fault inquiry information includes at least a fault object. The output of the queried feedback content includes: If the interaction method when receiving the fault inquiry information is the MR interaction method, obtain the MR interaction scenario, and when the fault object included in the fault inquiry information appears in the MR interaction scenario, add the feedback content to the MR interaction scenario.

[0010] By adopting the above technical solution, a multi-modal interaction method is provided, especially including the MR interaction method, that is, the user can wear an MR helmet, and the feedback content will be correspondingly displayed in the MR interaction scenario, and the virtual and real combination method is used to more vividly and conveniently obtain the fault solution of the fault object.

[0011] Optionally, the feedback content includes a fault solution, and the fault solution includes several specific steps; The adding of the feedback content to the MR interaction scenario includes: According to the specific steps corresponding to the fault solution in the feedback content, broadcast the specific steps in voice form, and determine the maintenance process in real time according to the picture displayed in the MR interaction scenario, and adjust the voice broadcast speed according to the maintenance process, so that the specific steps are synchronized with the maintenance process; Whenever a specific step is broadcast, play an operation animation in the MR interaction scenario, and the operation animation is a video animation of a virtual character performing the specific step.

[0012] By adopting the above-mentioned technical solution, the present application proposes to assist users in executing fault solution solutions in the form of voice broadcast, and further, it will help users execute each specific step of the fault solution in a more vivid way by playing operation animations, thereby further improving the efficiency of fault resolution.

[0013] Optionally, the processed fault inquiry information also includes the fault type; The method further comprises: At every designated period, according to the fault type in the received fault inquiry information, determine the fault similarity between the fault types corresponding to different fault objects; Determine a similar fault set, and add the fault solution corresponding to each fault type in the similar fault set to the fault solutions corresponding to other fault types in the similar fault set in the knowledge graph; wherein the similar fault types included in the similar fault set are fault information having a similarity higher than a preset similarity value.

[0014] By adopting the above technical solution, different fault objects are associated. Specifically, for all fault types whose fault information similarity exceeds a preset similarity, their corresponding fault solutions can be shared, so that when the feedback content is displayed, different solutions can be provided to users, thereby broadening users' fault-solving ideas.

[0015] Optionally, the processing of the fault inquiry information and searching the knowledge graph for feedback content corresponding to the processed fault inquiry information may include: When receiving the fault inquiry information, determining and generating a communication channel, wherein the communication channel includes at least a single-link channel and a group chat channel; The step of outputting the feedback content obtained by the query includes: If the determined communication channel is a single-link channel, the feedback content obtained by the query is output through the single-link channel, so that only the person who has raised the fault inquiry information can know the feedback content; If the determined communication channel is a group chat channel, the person who raises the fault inquiry information will be added to a pre-built fault group, and the inquiry information raised by each user in the fault group and the feedback content obtained from the corresponding query will be displayed in the fault group so that all users in the fault group can be informed; wherein, the inquiry information at least includes the fault inquiry information.

[0016] By adopting the above technical solution, when a user submits fault inquiry information to seek a fault solution, the present application provides two ways to communicate with the user. One is to conduct one-on-one Q&A with the user through a single-link channel, and the other is to add the user to a fault group through a group chat channel, so that all users in the fault group can learn the inquiry information submitted by other users, realizing multi-party communication, so that users can exchange experiences with each other and improve the efficiency of fault resolution.

[0017] Optionally, the processed fault inquiry information further includes a fault type. The method further includes: Whenever a fault inquiry information is received, according to the fault type included in the fault inquiry information, evaluate and output the health status of the corresponding fault object, and determine whether there is a secondary fault. If so, query the fault solution corresponding to the secondary fault from the knowledge graph, and generate and output a prevention plan with the fault solution.

[0018] By adopting the above technical solution, the health status of the fault object is evaluated according to the fault type corresponding to the fault inquiry information, so that the user can independently judge the subsequent operation stability of the fault object. In addition, when one fault occurs, it may trigger other faults (i.e., secondary faults). The present application predicts whether a secondary fault will occur in the subsequent operation of the fault object and outputs a corresponding prevention plan.

[0019] In a second aspect, the present application provides a machine fault Q&A system based on an industrial knowledge graph, including A knowledge graph update module, configured to periodically obtain maintenance data, process the maintenance data, and update a preset knowledge graph based on the processed maintenance data; wherein, both the knowledge graph and the processed maintenance data at least include basic information of a fault object and its corresponding fault solution; A fault communication and diagnosis module, configured to receive fault inquiry information, process the fault inquiry information, and query feedback content corresponding to the processed fault inquiry information from the knowledge graph; A solution feedback module, configured to output the queried feedback content, so that the submitter of the fault inquiry information can learn the feedback content.

[0020] In a third aspect, the present application provides a machine fault Q&A device based on an industrial knowledge graph, including a memory and a processor, and a computer program capable of being loaded and executed by the processor and performing any one of the methods described in the first aspect is stored on the memory.

[0021] Fourthly, the present application provides a computer-readable storage medium, which is characterized in that it stores a computer program that can be loaded and executed by a processor to perform any one of the methods described in the first aspect.

[0022] In summary, the present application includes at least one of the following beneficial technical effects: 1. In the present application, based on the correspondence relationship between the fault object and the fault solution stored in the knowledge graph, when a fault occurs in an industrial machine (i.e., the fault object), the above-mentioned solution proposed by the present application can be used to automatically query and feedback the feedback content corresponding to the fault inquiry information, and the feedback content contains the fault solution, thereby helping to improve the efficiency of fault resolution; 2. Further, the present application provides a multi-modal interaction method, especially including the MR interaction method, that is, the feedback content is correspondingly displayed in the MR interaction scene, and the virtual and real combination method is used to more vividly and conveniently know and execute the fault solution. Description of the Drawings

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0024] Figure 1 It is a schematic flowchart of a machine fault answering method based on an industrial knowledge graph disclosed in an embodiment of the present application.

[0025] Figure 2 It is a structural block diagram of a machine fault answering system based on an industrial knowledge graph disclosed in an embodiment of the present application.

[0026] Description of the reference numerals: 1. Knowledge graph update module; 2. Fault communication and diagnosis module; 3. Solution feedback module. Detailed Embodiments

[0027] The following will further elaborate on the present application in conjunction with the attached Figure 1-2 to further illustrate the present application in detail.

[0028] The embodiment of the present application discloses a machine fault answering method based on an industrial knowledge graph (hereinafter referred to as the fault answering method for short). The fault answering method is used to diagnose faults, query and output fault solutions when the user raises a fault inquiry, for the user's reference, so as to improve the efficiency of fault resolution. The execution subject of the fault answering method of the present application is a machine fault answering system based on an industrial knowledge graph (hereinafter referred to as the fault answering system for short). The following will be combined with the attached Figure 1Specifically elaborate on the process steps of the fault Q&A method.

[0029] S101, Regularly obtain maintenance data, process the maintenance data, and update a preset knowledge graph based on the processed maintenance data; wherein, both the knowledge graph and the processed maintenance data at least include the basic information of the fault object and its corresponding fault solution.

[0030] In implementation, the data sources of the maintenance data can specifically include machine manuals, historical maintenance records, and network channels related to machine repair (such as websites, forums); the acquisition method of the maintenance data disclosed in the exemplary embodiment of the present application is generally passive reception, that is, it is obtained by manual entry.

[0031] The operations for the fault Q&A system to process the maintenance data at least include data cleaning, format conversion, text extraction (such as using OCR technology to extract text information from images or scanned documents), annotation and classification (annotating key information, such as repair steps, fault types, fault solutions, etc., and also tagging the data for classification and retrieval).

[0032] The processed maintenance information at least includes the basic information of the fault object (such as machine name, model, function description, etc.), the fault type of the fault object, and the fault solution for each fault type. The fault Q&A system will update the knowledge graph according to the processed maintenance information. Correspondingly, the basic information, fault type, and fault solution of the fault object included in the maintenance information all exist in the knowledge graph in the form of entities, and by establishing an association relationship between the entities, the corresponding relationship between the fault type and the fault solution can be realized. The fault type can specifically include the specific performance parameter data when the machine fails. A graph database can also be used to store entities and relationships.

[0033] S102, Receive the fault inquiry information, process the fault inquiry information, and query the corresponding feedback content from the knowledge graph for the processed fault inquiry information.

[0034] S103, Output the queried feedback content so that the person who proposed the fault inquiry information can know the feedback content.

[0035] In implementation, the user can access the fault Q&A system in the form of a web page or an APP, and trigger a fault inquiry instruction by inputting the basic information of the fault object and the fault type. Correspondingly, the fault Q&A system will generate a fault inquiry message based on the content input by the user, and then process the fault inquiry message. The specific processing content includes semantic recognition, keyword extraction, etc. Then, the fault Q&A system queries the knowledge graph through a query engine, using the fault type and / or the basic information of the fault object as keywords, and queries the feedback content corresponding to the processed fault inquiry message. The feedback content can specifically include a fault solution. Finally, the fault Q&A system outputs the feedback content so that the user (i.e., the proposer of the fault inquiry message) can obtain the feedback content.

[0036] Optionally, the present application provides multiple interaction methods for the user to interact with the fault Q&A system. When triggering the fault inquiry instruction, the user can select an interaction method, and the interaction method can at least include the MR interaction method. Correspondingly, S103 will include the following sub-steps: If the interaction method when receiving the fault inquiry message is the MR interaction method, obtain the MR interaction scenario. When the fault object included in the fault inquiry message appears in the MR interaction scenario, add the feedback content to the MR interaction scenario; Among them, the above "adding the feedback content to the MR interaction scenario" further includes the following sub-steps: According to the specific steps corresponding to the fault solution in the feedback content, broadcast the specific steps in voice form, and determine the maintenance process in real time according to the picture displayed in the MR interaction scenario, and adjust the voice broadcast speed according to the maintenance process so that the specific steps are synchronized with the maintenance process; Whenever a specific step is broadcast, play an operation animation in the MR interaction scenario, and the operation animation is a video animation of a virtual character performing the specific steps.

[0037] In implementation, the fault Q&A system is pre-configured with an MR interaction device (MR helmet). The fault Q&A system will use the actual environment scenario that the user wearing the MR interaction device can view as the MR interaction scenario and display it on the mirror surface of the MR interaction device for the user to view; the fault Q&A system will, based on the basic information of the fault object included in the fault inquiry message, search for the picture of the fault object from a preset database, and compare the picture with the MR interaction scenario in real time. When the picture exists in the image corresponding to the MR interaction scenario, it is considered that the fault object included in the fault inquiry message appears in the MR interaction scenario, and at this time, the feedback content is added to the MR interaction scenario.

[0038] The adding method can be: displaying the feedback content in text form near the fault object in the MR interactive scene; or playing the specific steps of the fault solution contained in the feedback content in voice form with the help of the speaker of the MR interactive device. In addition, the playing speed of the specific steps can be pre-set, or as mentioned above: synchronized with the user's maintenance process.

[0039] Specifically, since a fault solution generally includes multiple specific steps, the fault question and answer system will stop playing after playing each specific step until a trigger instruction to continue playing is obtained from the user, and then play the next specific step until the playing is completed.

[0040] In addition, feedback content can be added by playing operation animations in the MR interactive scene. The operation animations are video animations of preset virtual characters performing all specific steps in the fault solution, so that users can perform all specific steps in the fault solution more intuitively and accurately.

[0041] In addition, in addition to the above-mentioned MR interaction method, the interaction method of the present application can also include a text question and answer interaction method, that is, the fault question and answer system creates a text question and answer interface, and displays the fault inquiry information raised by the user and the feedback content obtained by the fault question and answer system on the text question and answer interface. Among them, the feedback content can include text content and video content, and the video content can specifically be an operation video of the fault solution.

[0042] Optionally, when triggering a fault inquiry command, the user can also select a communication channel. Specifically, the communication channels include single-link channels and group chat channels. A single-link channel means that the user and the fault question and answer system will have a one-to-one interaction and communication; and a group chat channel means that the fault question and answer system will create a group chat, and the user can communicate and interact with the fault question and answer system and / or other users through the group chat.

[0043] Accordingly, before the step of “processing the fault inquiry information and searching the knowledge graph for feedback content corresponding to the processed fault inquiry information” in S102, the following steps are also included: When receiving fault inquiry information, determining and generating a communication channel, the communication channel at least includes a single-link channel and a group chat channel; S103 specifically includes the following contents: If the determined communication channel is a single-link channel, the feedback content obtained by the query is output through the single-link channel, so that only the person who raises the fault inquiry information can know the feedback content; If the determined communication channel is a group chat channel, add the originator of the fault inquiry information to the pre-constructed fault group, and display the inquiry information raised by each user in the fault group and the corresponding feedback content obtained by query in the fault group, so that all users in the fault group can know; among them, the inquiry information includes at least the fault inquiry information.

[0044] In implementation, the fault Q&A system can respectively establish corresponding fault groups according to different fault types and fault objects, and add users with the same fault type and fault object included in the raised fault inquiry information to the corresponding fault group. So-called a user communication group. Whenever any user in the fault group raises an inquiry information related to the fault type and fault object, and the communication channel selected by the user when raising the inquiry information is a group chat channel, the fault Q&A system will broadcast the inquiry information to all other users in the corresponding fault group, so that all users can know. Correspondingly, the feedback content related to the inquiry information obtained by the fault Q&A system will also be broadcast to all other users in the corresponding fault group to achieve efficient communication among multiple users.

[0045] Optionally, "query the feedback content corresponding to the processed fault inquiry information from the knowledge graph" in S102 further includes the following steps: Determine the importance degree of the process participated by the fault object in industrial production and the maintenance urgency degree of the fault object according to the fault object corresponding to the fault inquiry information; Query the feedback content corresponding to the processed fault inquiry information from the knowledge graph, and adjust the arrangement order of the fault solutions in the feedback content based on the maintenance urgency degree of the fault object and the maintenance duration of each fault solution in the feedback content; among them, the fault solution with a more forward arrangement order is preferentially output.

[0046] In implementation, the basic information corresponding to each fault object in the knowledge graph also includes the process participated by the corresponding fault object and the importance degree of the process, as well as the maintenance urgency degree of the fault object. The importance degree of the process is consistent with the maintenance urgency degree of the fault object. The importance degree of the process can be expressed in the form of a level. And the embodiments of the present application propose to judge the importance degree of the process by judging the position of the process participated by the fault object in the industrial production line. For example: in the whole industrial production line, the more subsequent processes of the process participated by the fault object (that is, the more forward the process participated by the fault object in the whole industrial production line), the higher the importance degree and the higher the maintenance urgency degree.

[0047] In addition, each fault solution corresponding to a fault object in the knowledge graph has a maintenance duration, which can be considered as the duration spent on executing the entire fault solution. Each fault solution also corresponds to the number of successful maintenance times. Specifically, whenever the fault Q&A system outputs feedback content, it will also output maintenance result feedback information for the user to fill in the maintenance result. The maintenance result specifically includes the fault solution that successfully solves the fault. If the fault solution is in the corresponding feedback content, the user can fill in the maintenance result in a selectable manner, and the number of successful maintenance times of the corresponding fault solution will be incremented by one; if it is not in the corresponding feedback content, the user needs to manually enter the specific content of the fault solution that successfully solves the fault.

[0048] The fault Q&A system is used to adjust the arrangement order of all fault solutions included in the feedback content according to the maintenance urgency. Specifically, for fault objects with a maintenance urgency higher than the specified level, the arrangement order of the fault solutions in the feedback content will be adjusted according to the rule that the shorter the maintenance duration, the higher the priority of the arrangement order. For fault objects with a maintenance urgency not higher than the specified level, the arrangement order of the fault solutions in the corresponding feedback content will be adjusted according to the rule that the higher the success rate, the higher the priority of the arrangement order.

[0049] Optionally, since the fault types specifically include specific performance parameter data during machine faults, such as voltage data, current data, temperature data, etc., the present application will establish associations between different fault objects with similar fault types based on the following scheme, so that different fault objects with similar fault types can share each other's fault solutions and broaden the user's thinking on fault solving.

[0050] Specifically, the fault Q&A method further includes the following steps: At regular intervals, according to the fault types in the received fault inquiry information, determine the fault similarity between the fault types corresponding to different fault objects; Determine the similar fault set, and add the fault solutions corresponding to each fault type in the similar fault set to the fault solutions corresponding to other fault types in the similar fault set in the knowledge graph; wherein, the similar fault types included in the similar fault set are fault information with a similarity higher than the preset similarity value.

[0051] In implementation, the fault Q&A system will calculate the similarity of different fault objects at specified intervals. The specific calculation method is as follows: for different fault types of different objects, the number of common performance parameters will be determined first, and then it will be determined whether the difference in the data values of the common performance parameters is within the specified difference range. If the following conditions are met: the number of common performance parameters is higher than the preset common number, and the data value difference is within the specified difference range, they are considered similar fault types to each other and added to the preset similar fault combination.

[0052] Exemplarily, for example, in fault type A1 of fault object A, there are parameters x, y, and z; in fault type B1 of fault object B, there are parameters m, x, and p. Then the common performance parameter between fault type A1 of fault object A and fault type B1 of fault object B is parameter x, and the number of common performance parameters is 1.

[0053] Optionally, the fault Q&A method further includes the following sub-steps: Whenever a fault inquiry message is received, according to the fault type included in the fault inquiry message, evaluate and output the health degree of the corresponding fault object, and determine whether there is a secondary fault. If there is, query the fault solution corresponding to the secondary fault from the knowledge graph, and generate and output a prevention plan with the fault solution.

[0054] In implementation, record the reception time of the fault inquiry message corresponding to each fault object, determine this reception time as the occurrence time of the corresponding fault type, and judge whether there is a sequence and secondary situation in terms of the occurrence time of different fault types of the same fault object according to the occurrence time of the fault type. If the time difference between the occurrence times corresponding to different fault types of the same fault object is less than the preset time difference, it is considered that the later-occurring fault type is a secondary fault of the earlier-occurring fault type, and it is considered that there is a concurrent relationship between the foregoing fault types; the fault Q&A system is used to generate and store each group of fault types with a concurrent relationship.

[0055] In addition, the basic information of each fault object included in the knowledge graph of the present application may further include the normal data range of the performance parameters of the fault object. Correspondingly, whenever the fault Q&A system receives a fault inquiry message, compare the specific performance parameter data in the fault type included in the fault inquiry message with the stored normal data range, specifically compare the difference between the performance parameter data in the fault type and the threshold of the normal range, and determine and output the health degree of the fault object according to the preset corresponding relationship table between the difference and the health degree.

[0056] The embodiment of the present application also discloses a machine fault Q&A system based on an industrial knowledge graph. Refer to Figure 2, the machine fault Q&A system based on the industrial knowledge graph includes: A knowledge graph update module 1, which is used to regularly obtain maintenance data, process the maintenance data, and update a preset knowledge graph based on the processed maintenance data; wherein, both the knowledge graph and the processed maintenance data at least include the basic information of the fault object and its corresponding fault solution; A fault communication and diagnosis module 2, which is used to receive fault inquiry information, process the fault inquiry information, and query the corresponding feedback content from the knowledge graph for the processed fault inquiry information; A solution feedback module 3, which is used to output the queried feedback content so that the person who proposed the fault inquiry information can know the feedback content.

[0057] Optionally, the fault communication and diagnosis module 2 is further used to determine the importance of the process in which the fault object participates in industrial production and the maintenance urgency of the fault object according to the fault object corresponding to the fault inquiry information; Query the corresponding feedback content from the knowledge graph for the processed fault inquiry information, and adjust the arrangement order of the fault solutions in the feedback content based on the maintenance urgency of the fault object and the maintenance duration of each fault solution in the feedback content; wherein, the fault solution with a higher arrangement order is preferentially output.

[0058] Optionally, the solution feedback module 3 is further used to, if the interaction method when receiving the fault inquiry information is the MR interaction method, obtain the MR interaction scenario, and add the feedback content to the MR interaction scenario when the fault object included in the fault inquiry information appears in the MR interaction scenario.

[0059] Optionally, the solution feedback module 3 is further used to broadcast the specific steps in the form of voice according to the specific steps corresponding to the fault solutions in the feedback content, and determine the maintenance process in real time according to the picture displayed in the MR interaction scenario, and adjust the voice broadcast speed according to the maintenance process so that the specific steps are synchronized with the maintenance process; whenever a specific step is broadcast, an operation animation is played in the MR interaction scenario, and the operation animation is a video animation of a virtual character executing the specific step.

[0060] Optionally, a similarity solution supplement module is configured to determine the fault similarity between the fault types corresponding to different fault objects according to the fault type in the received fault inquiry information every specified period; and is further configured to determine a set of similar faults, and add the fault solutions corresponding to each fault type in the set of similar faults to the fault solutions corresponding to other fault types in the set of similar faults in the knowledge graph; wherein, the similar fault types included in the set of similar faults are fault information with a similarity higher than a preset similarity value. Optionally, it further includes a communication channel selection module, configured to determine and generate a communication channel when receiving a fault inquiry information, and the communication channel includes at least a single-link channel and a group chat channel. The solution feedback module 3 is further configured to, if the determined communication channel is a single-link channel, output the query result feedback content through the single-link channel, so that only the sender of the fault inquiry information can know the feedback content; and is further configured to, if the determined communication channel is a group chat channel, add the sender of the fault inquiry information to a pre-constructed fault group, and display the inquiry information and the corresponding query result feedback content sent by each user in the fault group in the fault group, so that all users in the fault group can know; wherein, the inquiry information includes at least the fault inquiry information.

[0061] Optionally, it further includes a fault prediction module, configured to, whenever receiving a fault inquiry information, evaluate and output the health degree of the corresponding fault object according to the fault type included in the fault inquiry information, and determine whether there is a secondary fault. If so, query the fault solution corresponding to the secondary fault from the knowledge graph, and generate and output a prevention plan with the fault solution.

[0062] The embodiment of the present application also discloses a machine fault question-answering device based on an industrial knowledge graph. The machine fault question-answering device based on an industrial knowledge graph includes a memory and a processor. The memory stores a computer program that can be loaded and executed by the processor to perform the above-mentioned machine fault question-answering method based on an industrial knowledge graph.

[0063] The embodiment of the present application also discloses a computer-readable storage medium, which stores a computer program that can be loaded and executed by the processor to perform the above-mentioned machine fault question-answering method based on an industrial knowledge graph. The computer-readable storage medium includes, for example: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes.

[0064] It should be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0065] The above embodiments are only used to illustrate the technical solutions of the present application, rather than limiting the protection scope of the application. Obviously, the described embodiments are only partial embodiments of the present application, rather than all embodiments. Based on these embodiments, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope to be protected by the present application.

Claims

1. A machine fault question-answering method based on industrial knowledge graph, characterized in that: include: Periodically acquiring maintenance data, processing the maintenance data, and updating a preset knowledge graph based on the processed maintenance data; wherein both the knowledge graph and the processed maintenance data include at least basic information of the fault object and its corresponding fault solution; receiving fault inquiry information, processing the fault inquiry information, and searching the knowledge graph for feedback content corresponding to the processed fault inquiry information; The feedback content obtained by the query is outputted so that the person who has submitted the fault inquiry information can know the feedback content.

2. The machine fault question-answering method based on industrial knowledge graph according to claim 1 is characterized in that: The processed fault inquiry information at least includes the fault object, and the feedback content includes the fault solution; The querying of feedback content corresponding to the processed fault inquiry information from the knowledge graph includes: Determine, according to the fault object corresponding to the fault inquiry information, the importance of the process in which the fault object participates in industrial production, and the maintenance urgency of the fault object; The feedback content corresponding to the processed fault inquiry information is queried from the knowledge graph, and the arrangement order of the fault solutions in the feedback content is adjusted based on the maintenance urgency of the fault object and the maintenance duration of each fault solution in the feedback content; wherein, the fault solutions with higher arrangement order are output first.

3. The machine fault question-answering method based on industrial knowledge graph according to claim 1 is characterized in that: The processed fault inquiry information at least includes the fault object, The step of outputting the feedback content obtained by the query includes: If the interaction mode when receiving the fault inquiry information is the MR interaction mode, an MR interaction scene is acquired, and when the fault object included in the fault inquiry information appears in the MR interaction scene, the feedback content is added to the MR interaction scene.

4. The machine fault question-answering method based on industrial knowledge graph according to claim 3 is characterized in that: The feedback content includes a fault solution, and the fault solution includes several specific steps; The adding the feedback content to the MR interaction scene includes: According to the specific steps corresponding to the fault solution in the feedback content, the specific steps are broadcasted in voice form, and the maintenance progress is determined in real time according to the screen displayed by the MR interactive scene, and the voice broadcast speed is adjusted according to the maintenance progress so that the specific steps are synchronized with the maintenance progress; Whenever a specific step is broadcast, an operation animation is played in the MR interaction scene, and the operation animation is a video animation of a virtual character executing the specific step.

5. The machine fault question-answering method based on industrial knowledge graph according to claim 1 is characterized in that: The processed fault inquiry information also includes the fault type; The method further comprises: At every designated period, according to the fault type in the received fault inquiry information, determine the fault similarity between the fault types corresponding to different fault objects; Determine a similar fault set, and add the fault solution corresponding to each fault type in the similar fault set to the fault solutions corresponding to other fault types in the similar fault set in the knowledge graph; wherein the similar fault types included in the similar fault set are fault information having a similarity higher than a preset similarity value.

6. The machine fault question-answering method based on industrial knowledge graph according to claim 1 is characterized in that: The processing of the fault inquiry information and searching the knowledge graph for feedback content corresponding to the processed fault inquiry information include: When receiving the fault inquiry information, determining and generating a communication channel, wherein the communication channel includes at least a single-link channel and a group chat channel; The step of outputting the feedback content obtained by the query includes: If the determined communication channel is a single-link channel, the feedback content obtained by the query is output through the single-link channel, so that only the person who has raised the fault inquiry information can know the feedback content; If the determined communication channel is a group chat channel, the person who raises the fault inquiry information will be added to a pre-built fault group, and the inquiry information raised by each user in the fault group and the feedback content obtained from the corresponding query will be displayed in the fault group so that all users in the fault group can be informed; wherein, the inquiry information at least includes the fault inquiry information.

7. The machine fault question-answering method based on industrial knowledge graph according to claim 1 is characterized in that: The processed fault inquiry information also includes the fault type; The method further comprises: Whenever a fault inquiry message is received, the health of the corresponding fault object is evaluated and output according to the fault type contained in the fault inquiry message, and it is determined whether there is a secondary fault. If so, the fault solution corresponding to the secondary fault is queried from the knowledge graph, and a prevention plan with the fault solution is generated and output.

8. A machine fault question-answering system based on industrial knowledge graph, characterized in that: include, A knowledge graph updating module (1) is used to periodically acquire maintenance data, process the maintenance data, and update a preset knowledge graph based on the processed maintenance data; wherein the knowledge graph and the processed maintenance data both include at least basic information of a fault object and a corresponding fault solution; A fault communication diagnosis module (2) is used to receive fault inquiry information, process the fault inquiry information, and query feedback content corresponding to the processed fault inquiry information from the knowledge graph; The solution feedback module (3) is used to output the feedback content obtained by the query, so that the person who raises the fault inquiry information can know the feedback content.

9. A machine fault question-answering device based on industrial knowledge graph, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute any one of the methods according to claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and execute any one of the methods as claimed in claims 1 to 7.

Images