CN116861189A - Method for constructing equipment fault diagnosis maintenance knowledge base based on large language model - Google Patents
Method for constructing equipment fault diagnosis maintenance knowledge base based on large language model Download PDFInfo
- Publication number
- CN116861189A CN116861189A CN202310918204.2A CN202310918204A CN116861189A CN 116861189 A CN116861189 A CN 116861189A CN 202310918204 A CN202310918204 A CN 202310918204A CN 116861189 A CN116861189 A CN 116861189A
- Authority
- CN
- China
- Prior art keywords
- maintenance
- fault
- diagnosis
- knowledge base
- cloud platform
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012423 maintenance Methods 0.000 title claims abstract description 235
- 238000003745 diagnosis Methods 0.000 title claims abstract description 106
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000012549 training Methods 0.000 claims abstract description 8
- 230000008439 repair process Effects 0.000 claims description 22
- 238000005457 optimization Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 13
- 230000000694 effects Effects 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 5
- 239000013589 supplement Substances 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- 101100493820 Caenorhabditis elegans best-1 gene Proteins 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 2
- 230000008901 benefit Effects 0.000 abstract description 5
- 238000009825 accumulation Methods 0.000 description 7
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
- G06F16/252—Integrating or interfacing systems involving database management systems between a Database Management System and a front-end application
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of cloud computing, in particular to a method for constructing a device fault diagnosis maintenance knowledge base based on a large language model. By using the method, the historical maintenance data of each tenant can be fully utilized. On the one hand, a diagnosis maintenance knowledge base is constructed to store the historical maintenance data; when the tenant equipment has actual faults, a maintenance case with similarity higher than a preset value can be recommended for the tenant equipment in a matching mode for reference, so that the labor cost and the time cost of maintenance of the tenant equipment are reduced. On the other hand, after the historical data is processed, the historical data is used for training a diagnosis model, when the combined equipment fails and cannot be matched with a maintenance case with the similarity higher than a preset value, a reference scheme can be obtained through the diagnosis model, and the reference scheme is used as a reference when the tenant maintains the actual failure of the equipment. The method can reduce the dependence on manual specialists and improve the efficiency and economic benefit of enterprise equipment management and operation and maintenance.
Description
Technical Field
The invention belongs to the technical field of cloud computing, and particularly relates to a method for constructing a device fault diagnosis maintenance knowledge base based on a large language model.
Background
At present, in an equipment management system of an industrial production site, equipment fault maintenance work mainly depends on manual experience and a rule-based diagnosis system, and the manual experience is difficult to realize effective accumulation and multiplexing, and systematic management and sharing are realized, so that knowledge (maintenance experience) is lost and repeated construction is caused.
In order to ensure the efficiency and effect of equipment diagnosis and maintenance, some technicians put forward the technical thought of an equipment customized diagnosis and maintenance system, however, the fault diagnosis and maintenance system based on customized development cannot realize complex fault diagnosis because the fault type of diagnosis is too specific and limited. Moreover, many maintenance experiences only exist in individual enterprises, information islands exist, and industry experience sharing is difficult to achieve.
Thus, at present, equipment failure diagnosis and maintenance of enterprises still highly depend on personal experience and capabilities of human experts.
In summary, how to reduce the dependence on human expert and improve the efficiency and economic benefit of enterprise equipment management and operation and maintenance becomes the current urgent problem to be solved.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method for constructing a device fault diagnosis maintenance knowledge base based on a large language model, which can reduce the dependence on manual specialists and improve the efficiency and economic benefit of enterprise device management and operation and maintenance.
In order to solve the technical problems, the invention adopts the following technical scheme:
the method for constructing the equipment fault diagnosis maintenance knowledge base based on the large language model comprises the following steps:
s1, building a cloud platform, and building a diagnosis maintenance knowledge base and a diagnosis model on the cloud platform; the diagnosis model is a neural network language model;
s2, after the tenant applies to join the cloud platform, the historical maintenance data are synchronized to the cloud platform, the cloud platform processes the historical maintenance data to obtain maintenance cases, the maintenance cases are stored in a diagnosis maintenance knowledge base, and the diagnosis model is trained by using the maintenance cases;
s3, after the actual fault of the tenant equipment occurs, fault repair is carried out on the cloud platform, and actual fault information is input; the cloud platform queries the knowledge base, judges whether a maintenance case with similarity higher than a preset value exists, if so, sends the maintenance case to the tenant for reference, and goes to S5; if not, the process goes to S4;
s4, the cloud platform calls a diagnosis model to infer according to actual fault information, so that the most probable fault cause is obtained, a maintenance scheme suggestion is generated, the fault cause and the maintenance scheme suggestion are packaged into a reference scheme and sent to a corresponding tenant, and then S5 is carried out;
s5, after the maintenance personnel of the tenant combine the maintenance case or the reference scheme sent by the cloud platform to maintain the actual fault of the equipment, feeding back the actual maintenance scheme and the maintenance result to the cloud platform;
s6, if the maintenance personnel refer to a recommended reference scheme of the language model, the cloud platform compares the fed-back actual maintenance scheme and maintenance effect with the reference scheme, analyzes the difference between the fed-back actual maintenance scheme and the maintenance effect, stores the comparison analysis result in a diagnosis maintenance knowledge base, and is used for providing data support for optimization of the diagnosis model; if the maintenance personnel refer to the maintenance case and change from the actual maintenance scheme, the cloud platform processes the actual maintenance scheme and stores the processed actual maintenance scheme into a diagnosis maintenance knowledge base.
Preferably, in S2, the historical repair data includes a plurality of fault data, each fault data including device information, fault description, fault type, fault cause, and repair scheme; when the diagnosis model is trained, the training content comprises diagnosing the fault reason according to the equipment information and the fault description and generating maintenance proposal.
Preferably, in S2, the processing of the historical maintenance data includes filtering, sorting and normalizing.
Preferably, in S3, the actual fault information includes a fault description and a device parameter of an actual fault.
Preferably, in S3, when the actual fault information is entered, the fault description is filled in or selected from the diagnostic repair knowledge base.
Preferably, S3 comprises:
s301, after actual faults occur in equipment of tenants, fault repair is conducted on the cloud platform, equipment information is input, and fault description is filled in or selected from a knowledge base;
s302, the cloud platform acquires maintenance cases with similarity higher than a preset value from a diagnosis maintenance knowledge base; if not, go to S4, if yes, go to S303;
s303, if the number of the acquired maintenance cases is 1, the maintenance cases are sent to the tenant to serve as references, and S5 is carried out; if the number of the acquired maintenance cases is greater than 1, the process goes to S304;
and S304, guiding a repair staff to further screen the acquired maintenance cases according to the prompt and supplement fault description, sending the screened best 1 maintenance cases to the tenant as a reference, and turning to S5.
Preferably, S5 comprises:
s501, a maintenance person detects and judges the fault equipment, and confirms the accuracy of a maintenance case or a reference scheme sent by the cloud platform; if so, go to S502, and if not, go to S503;
s502, performing fault maintenance on the related faults by maintenance personnel according to a maintenance case or a reference scheme, and turning to S504 after the maintenance is completed;
s503, revising the maintenance case or the reference scheme by a maintenance person, or manually revising the diagnosis, determining a final maintenance scheme, performing fault maintenance according to the final maintenance scheme, and turning to S504 after the maintenance is completed;
s504, feeding back an actual maintenance scheme and maintenance results to the cloud platform.
Preferably, after S6, S7 is further included, and usage evaluation feedback of different tenants is collected periodically, so as to perform iterative optimization on the diagnostic model and other functional modules of the cloud platform.
Compared with the prior art, the invention has the following beneficial effects:
1. by using the method, the historical maintenance data of each tenant (enterprise) can be fully utilized. On the one hand, a diagnosis maintenance knowledge base is constructed to store the historical maintenance data; when the tenant equipment has actual faults, a maintenance case with similarity higher than a preset value can be recommended for the tenant equipment in a matching mode for reference, so that the labor cost and the time cost of maintenance of the tenant equipment are reduced. On the other hand, after the historical data is processed, the historical data is used for training a diagnosis model, when the combined equipment fails and cannot be matched with a maintenance case with the similarity higher than a preset value, a reference scheme (namely a failure cause and a maintenance scheme suggestion) can be obtained through the diagnosis model, and the maintenance scheme is used as a reference when a tenant maintains the actual failure of the equipment, so that the labor cost and the time cost of maintenance of the equipment can be reduced.
By the mode, accurate diagnosis results can be provided for equipment faults and maintenance to serve as references, the maintenance accuracy can be improved, the maintenance efficiency and quality are improved, the maintenance cost is saved, the resource utilization rate is improved, continuous accumulation and innovation of equipment maintenance experience accumulation are promoted, and the sustainability of digital upgrading of an equipment management system is enhanced. In addition, the processing method applies a large-scale language model to the field of equipment fault diagnosis: the large-scale language model is utilized to realize equipment fault diagnosis and maintenance scheme recommendation, a traditional rule-based fault diagnosis mode is broken through to a certain extent, and the artificial intelligence technology is applied to the key field of manufacturing industry.
In conclusion, the method can reduce the dependence on manual specialists and improve the efficiency and economic benefit of enterprise equipment management and operation and maintenance.
2. According to the invention, the cloud platform is used for deploying the public knowledge base, namely the diagnosis maintenance knowledge base, so that the collection of maintenance cases and experiences of different enterprises is realized, the information island can be broken, and the sharing and integration of industry knowledge are promoted.
3. In the application process of the cloud platform, the method can carry out iterative optimization on the diagnosis model and other functional modules of the cloud platform based on actual feedback (using experience feedback and maintenance actual result feedback) of the tenant, so that the diagnosis model can be ensured to meet actual fault analysis requirements more and more, and users can have better operation experience when using the cloud platform.
4. The method not only realizes unidirectional optimization improvement of the diagnosis model, but also supplements and perfects the content of the knowledge base through the reasoning result (and the practical application and modification of maintenance personnel) of the diagnosis model, so that the knowledge base is continuously enriched along with user feedback and case accumulation, and the collaborative development of knowledge and the model is realized.
Drawings
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart of an embodiment;
FIG. 2 is a flow chart of step 1 in the embodiment;
FIG. 3 is a flow chart of step 3 in the embodiment;
fig. 4 is a flowchart of step 5 in the embodiment.
Detailed Description
The following is a further detailed description of the embodiments:
embodiment one:
as shown in fig. 1, the embodiment discloses a method for constructing a device fault diagnosis and maintenance knowledge base based on a large language model, which comprises the following steps:
s1, building a cloud platform, and building a diagnosis maintenance knowledge base and a diagnosis model on the cloud platform; the diagnostic model is a neural network language model.
As shown in fig. 2, in the implementation, S1 includes:
s101, building an infrastructure of a cloud platform, wherein the infrastructure comprises computing resources, storage resources, network resources and a security mechanism;
s102, developing a device fault diagnosis maintenance knowledge base system on the cloud platform, wherein the system comprises a diagnosis maintenance knowledge base and a diagnosis model, and supports fault repair, device maintenance, acceptance check and intelligent fault diagnosis.
S103, formulating a user use protocol, and defining a user data acquisition range and a use rule.
And S2, after the tenant applies to join the cloud platform, synchronizing the historical maintenance data to the cloud platform, processing the historical maintenance data by the cloud platform to obtain maintenance cases, storing the maintenance cases in a diagnosis maintenance knowledge base, and training the diagnosis model by using the maintenance cases.
In particular implementations, the historical repair data includes a plurality of fault data, each fault data including equipment information, fault description, fault type, fault cause, and repair scheme; when the diagnosis model is trained, the training content comprises diagnosing the fault reason according to the equipment information and the fault description and generating maintenance proposal. The processing of the historical maintenance data includes filtering, sorting and normalizing.
S3, after the actual fault of the tenant equipment occurs, fault repair is carried out on the cloud platform, and actual fault information is input; the actual fault information comprises fault description and equipment parameters of an actual fault; and when the actual fault information is recorded, filling in or selecting fault description from the diagnosis maintenance knowledge base. The cloud platform queries the knowledge base, judges whether a maintenance case with similarity higher than a preset value exists, if so, sends the maintenance case to the tenant for reference, and goes to S5; if not, the process goes to S4;
as shown in fig. 3, in the implementation, S3 includes:
s301, after actual faults occur in equipment of tenants, fault repair is conducted on the cloud platform, equipment information is input, and fault description is filled in or selected from a knowledge base;
s302, the cloud platform acquires maintenance cases with similarity higher than a preset value from a diagnosis maintenance knowledge base; if not, go to S4, if yes, go to S303;
s303, if the number of the acquired maintenance cases is 1, the maintenance cases are sent to the tenant to serve as references, and S5 is carried out; if the number of the acquired maintenance cases is greater than 1, the process goes to S304;
and S304, guiding a repair staff to further screen the acquired maintenance cases according to the prompt and supplement fault description, sending the screened best 1 maintenance cases to the tenant as a reference, and turning to S5.
S4, the cloud platform calls a diagnosis model to infer according to actual fault information, so that the most probable fault cause is obtained, a maintenance scheme suggestion is generated, the fault cause and the maintenance scheme suggestion are packaged into a reference scheme and sent to a corresponding tenant, and then S5 is carried out;
s5, after the maintenance personnel of the tenant combine the maintenance case or the reference scheme sent by the cloud platform to maintain the actual fault of the equipment, feeding back the actual maintenance scheme and the maintenance result to the cloud platform;
as shown in fig. 4, in the implementation, S5 includes:
s501, a maintenance person detects and judges the fault equipment, and confirms the accuracy of a maintenance case or a reference scheme sent by the cloud platform; if so, go to S502, and if not, go to S503;
s502, performing fault maintenance on the related faults by maintenance personnel according to a maintenance case or a reference scheme, and turning to S504 after the maintenance is completed;
s503, revising the maintenance case or the reference scheme by a maintenance person, or manually revising the diagnosis, determining a final maintenance scheme, performing fault maintenance according to the final maintenance scheme, and turning to S504 after the maintenance is completed;
s504, feeding back an actual maintenance scheme and maintenance results to the cloud platform.
S6, if the maintenance personnel refer to a recommended reference scheme of the language model, the cloud platform compares the fed-back actual maintenance scheme and maintenance effect with the reference scheme, analyzes the difference between the fed-back actual maintenance scheme and the maintenance effect, stores the comparison analysis result in a diagnosis maintenance knowledge base, and is used for providing data support for optimization of the diagnosis model; if the maintenance personnel refer to the maintenance case and change from the actual maintenance scheme, the cloud platform processes the actual maintenance scheme and stores the processed actual maintenance scheme into a diagnosis maintenance knowledge base.
And S7, periodically collecting the use evaluation feedback of different tenants, and performing iterative optimization on the diagnosis model and other functional modules of the cloud platform.
By using the method, the historical maintenance data of each tenant (enterprise) can be fully utilized. On the one hand, a diagnosis maintenance knowledge base is constructed to store the historical maintenance data; when the tenant equipment has actual faults, a maintenance case with similarity higher than a preset value can be recommended for the tenant equipment in a matching mode for reference, so that the labor cost and the time cost of maintenance of the tenant equipment are reduced. On the other hand, after the historical data is processed, the historical data is used for training a diagnosis model, when the combined equipment fails and cannot be matched with a maintenance case with the similarity higher than a preset value, a reference scheme (namely a failure cause and a maintenance scheme suggestion) can be obtained through the diagnosis model, and the maintenance scheme is used as a reference when a tenant maintains the actual failure of the equipment, so that the labor cost and the time cost of maintenance of the equipment can be reduced. By the mode, accurate diagnosis results can be provided for equipment faults and maintenance to serve as references, the maintenance accuracy can be improved, the maintenance efficiency and quality are improved, the maintenance cost is saved, the resource utilization rate is improved, continuous accumulation and innovation of equipment maintenance experience accumulation are promoted, and the sustainability of digital upgrading of an equipment management system is enhanced. In addition, the processing method applies a large-scale language model to the field of equipment fault diagnosis: the large-scale language model is utilized to realize equipment fault diagnosis and maintenance scheme recommendation, a traditional rule-based fault diagnosis mode is broken through to a certain extent, and the artificial intelligence technology is applied to the key field of manufacturing industry.
According to the invention, the cloud platform is used for deploying the public knowledge base, namely the diagnosis maintenance knowledge base, so that the collection of maintenance cases and experiences of different enterprises is realized, the information island can be broken, and the sharing and integration of industry knowledge are promoted.
Embodiment two:
in order to facilitate a better understanding and implementation of the method by those skilled in the art, the present embodiment will be described by way of an example. By using the method, the cloud platform is constructed, used and optimized as follows.
1. Constructing a cloud platform: by adopting the B/S architecture, a user accesses the cloud platform through a browser, and the cloud platform can deploy components such as an equipment maintenance module, an experience management module, a diagnosis model, computing resources and the like. After the user is verified, the historical maintenance cases synchronously uploaded by the user are used for training a diagnosis model, and the data of the historical maintenance cases are processed and stored in a diagnosis maintenance knowledge base.
2. User access: after finding out real-time equipment faults, a user logs in the cloud platform and enters an equipment maintenance module; and inputting information such as fault equipment, use environment, fault description and the like, and submitting the information to the cloud platform.
3. Query history cases: the cloud platform queries the knowledge base, checks whether a history case and a maintenance scheme (namely a maintenance case) with similar reading larger than a preset value exist, and if so, displays the related content to a user for reference; if multiple cases exist in the same fault description, guiding a user to add description through options, screening out the one with the highest similar reading, and pushing the one to the user for reference; if not, the next step is carried out.
4. Diagnosis is performed using a diagnostic model: user information is input into a diagnosis model, the diagnosis model integrates factors such as equipment type, use environment, fault description and the like, and the most probable fault diagnosis result and maintenance scheme (namely a reference scheme) are recommended and displayed for the user to refer to.
5. Monitoring and maintaining: pushing a maintenance case or a reference scheme to maintenance personnel, initially verifying the accuracy through manual detection equipment, and if the accuracy is high, repairing the maintenance case or the reference scheme; if the two types of the data are different, re-diagnosing and repairing; and filling in an actual implementation scheme after maintenance is completed.
6. And (3) comparing and storing: the cloud platform compares the actual execution scheme filled by the maintenance personnel with the recommended result of the diagnosis model, analyzes the difference between the actual execution scheme and the recommended result of the diagnosis model, stores the compared result to a knowledge base, and provides data support for model optimization.
7. And (5) user acceptance: the user checks and evaluates the comparison result and the final maintenance effect, scores and fills in comments; and the cloud platform collects the evaluation content and stores the evaluation content into a knowledge base.
8. Model and rule optimization: the experience management module periodically analyzes user feedback and new cases stored in the knowledge base, identifies the defects of the diagnosis model and knowledge rules, and performs optimization adjustment by using feedback content.
9. And (3) system optimization: the cloud platform periodically checks user feedback and system operation data, upgrades the diagnosis model and other components, and continuously optimizes system performance and user experience.
10. And (3) new case verification: the optimized diagnosis model is used for new case diagnosis, and the obtained diagnosis result and maintenance effect are evaluated by the user again, and the cycle is re-entered, so that continuous improvement is realized.
According to the method, in the application process of the cloud platform, iteration optimization can be performed on the diagnosis model and other functional modules of the cloud platform based on actual feedback (using experience feedback and maintenance actual result feedback) of the tenant, so that the diagnosis model can be ensured to meet actual fault analysis requirements more and more, and users can have better operation experience when using the cloud platform. In addition, the method not only realizes unidirectional optimization improvement of the diagnosis model, but also supplements and perfects the knowledge base content through the reasoning result (and the practical application and modification of maintenance personnel) of the diagnosis model, so that the knowledge base is also continuously enriched along with user feedback and case accumulation, and the collaborative development of knowledge and the model is realized.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the technical solution, and those skilled in the art should understand that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the present invention, and all such modifications and equivalents are included in the scope of the claims.
Claims (8)
1. The method for constructing the equipment fault diagnosis maintenance knowledge base based on the large language model is characterized by comprising the following steps of:
s1, building a cloud platform, and building a diagnosis maintenance knowledge base and a diagnosis model on the cloud platform; the diagnosis model is a neural network language model;
s2, after the tenant applies to join the cloud platform, the historical maintenance data are synchronized to the cloud platform, the cloud platform processes the historical maintenance data to obtain maintenance cases, the maintenance cases are stored in a diagnosis maintenance knowledge base, and the diagnosis model is trained by using the maintenance cases;
s3, after the actual fault of the tenant equipment occurs, fault repair is carried out on the cloud platform, and actual fault information is input; the cloud platform queries the knowledge base, judges whether a maintenance case with similarity higher than a preset value exists, if so, sends the maintenance case to the tenant for reference, and goes to S5; if not, the process goes to S4;
s4, the cloud platform calls a diagnosis model to infer according to actual fault information, so that the most probable fault cause is obtained, a maintenance scheme suggestion is generated, the fault cause and the maintenance scheme suggestion are packaged into a reference scheme and sent to a corresponding tenant, and then S5 is carried out;
s5, after the maintenance personnel of the tenant combine the maintenance case or the reference scheme sent by the cloud platform to maintain the actual fault of the equipment, feeding back the actual maintenance scheme and the maintenance result to the cloud platform;
s6, if the maintenance personnel refer to a recommended reference scheme of the language model, the cloud platform compares the fed-back actual maintenance scheme and maintenance effect with the reference scheme, analyzes the difference between the fed-back actual maintenance scheme and the maintenance effect, stores the comparison analysis result in a diagnosis maintenance knowledge base, and is used for providing data support for optimization of the diagnosis model; if the maintenance personnel refer to the maintenance case and change from the actual maintenance scheme, the cloud platform processes the actual maintenance scheme and stores the processed actual maintenance scheme into a diagnosis maintenance knowledge base.
2. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 1, wherein: in S2, the historical maintenance data comprises a plurality of fault data, wherein each fault data comprises equipment information, fault description, fault type, fault reason and maintenance scheme; when the diagnosis model is trained, the training content comprises diagnosing the fault reason according to the equipment information and the fault description and generating maintenance proposal.
3. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 2, wherein: in S2, the processing of the historical maintenance data includes filtering, sorting and normalizing.
4. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 3, wherein: in S3, the actual fault information includes a fault description and an equipment parameter of the actual fault.
5. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 4, wherein: and S3, filling in or selecting fault description from a diagnosis maintenance knowledge base when the actual fault information is recorded.
6. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 5, wherein: s3 comprises the following steps:
s301, after actual faults occur in equipment of tenants, fault repair is conducted on the cloud platform, equipment information is input, and fault description is filled in or selected from a knowledge base;
s302, the cloud platform acquires maintenance cases with similarity higher than a preset value from a diagnosis maintenance knowledge base; if not, go to S4, if yes, go to S303;
s303, if the number of the acquired maintenance cases is 1, the maintenance cases are sent to the tenant to serve as references, and S5 is carried out; if the number of the acquired maintenance cases is greater than 1, the process goes to S304;
and S304, guiding a repair staff to further screen the acquired maintenance cases according to the prompt and supplement fault description, sending the screened best 1 maintenance cases to the tenant as a reference, and turning to S5.
7. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 6, wherein: s5 comprises the following steps:
s501, a maintenance person detects and judges the fault equipment, and confirms the accuracy of a maintenance case or a reference scheme sent by the cloud platform; if so, go to S502, and if not, go to S503;
s502, performing fault maintenance on the related faults by maintenance personnel according to a maintenance case or a reference scheme, and turning to S504 after the maintenance is completed;
s503, revising the maintenance case or the reference scheme by a maintenance person, or manually revising the diagnosis, determining a final maintenance scheme, performing fault maintenance according to the final maintenance scheme, and turning to S504 after the maintenance is completed;
s504, feeding back an actual maintenance scheme and maintenance results to the cloud platform.
8. The method for building a device fault diagnosis and repair knowledge base based on a large language model as claimed in claim 1, wherein: and S7, periodically collecting the use evaluation feedback of different tenants, and performing iterative optimization on the diagnosis model and other functional modules of the cloud platform.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310918204.2A CN116861189A (en) | 2023-07-25 | 2023-07-25 | Method for constructing equipment fault diagnosis maintenance knowledge base based on large language model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310918204.2A CN116861189A (en) | 2023-07-25 | 2023-07-25 | Method for constructing equipment fault diagnosis maintenance knowledge base based on large language model |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116861189A true CN116861189A (en) | 2023-10-10 |
Family
ID=88224964
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310918204.2A Pending CN116861189A (en) | 2023-07-25 | 2023-07-25 | Method for constructing equipment fault diagnosis maintenance knowledge base based on large language model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116861189A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117130943A (en) * | 2023-10-26 | 2023-11-28 | 北京一平方科技有限公司 | Test case generation and operation and maintenance data analysis method based on large language model |
CN117575009A (en) * | 2023-11-22 | 2024-02-20 | 浙江杉工智能科技有限公司 | Bridge disease diagnosis and maintenance measure recommendation method based on large language model |
-
2023
- 2023-07-25 CN CN202310918204.2A patent/CN116861189A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117130943A (en) * | 2023-10-26 | 2023-11-28 | 北京一平方科技有限公司 | Test case generation and operation and maintenance data analysis method based on large language model |
CN117130943B (en) * | 2023-10-26 | 2024-02-20 | 北京一平方科技有限公司 | Test case generation and operation and maintenance data analysis method based on large language model |
CN117575009A (en) * | 2023-11-22 | 2024-02-20 | 浙江杉工智能科技有限公司 | Bridge disease diagnosis and maintenance measure recommendation method based on large language model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110929918B (en) | 10kV feeder fault prediction method based on CNN and LightGBM | |
CN116861189A (en) | Method for constructing equipment fault diagnosis maintenance knowledge base based on large language model | |
CN112859822B (en) | Equipment health analysis and fault diagnosis method and system based on artificial intelligence | |
CN110674189A (en) | Method for monitoring secondary state and positioning fault of intelligent substation | |
CN106533754A (en) | Fault diagnosis method and expert system for college teaching servers | |
CN116771576B (en) | Comprehensive fault diagnosis method for hydroelectric generating set | |
CN113627032B (en) | Intelligent decision-making method for equipment design/maintenance scheme based on digital twinning | |
CN110110095A (en) | A kind of power command text matching technique based on shot and long term memory Recognition with Recurrent Neural Network | |
CN112612819A (en) | Big data analysis and mining method and system for pumped storage power station | |
CN113627784A (en) | Enterprise asset management intelligent decision-making system based on industrial internet | |
CN116307641A (en) | Digital power plant-oriented resource collaborative scheduling management method and system | |
CN117827750A (en) | Personnel file automatic archiving method and system | |
CN113010624B (en) | Intelligent power distribution inspection system | |
CN116611813B (en) | Intelligent operation and maintenance management method and system based on knowledge graph | |
CN112783865A (en) | Semi-supervised man-machine combined operation and maintenance fault library generation method and system | |
Long et al. | 6G comprehensive intelligence: network operations and optimization based on Large Language Models | |
CN117291575A (en) | Equipment maintenance method, equipment maintenance device, computer equipment and storage medium | |
CN114819925B (en) | Industrial Internet of things system based on event sequence analysis and prediction and control method thereof | |
WO2024065776A1 (en) | Method for data processing, apparatus for data processing, electronic device, and storage medium | |
CN114091325A (en) | Power distribution station load prediction method and system based on multi-model layered learning | |
CN118520978B (en) | Traffic scene-oriented AI productivity platform framework system and method | |
CN114817258B (en) | Network source data compliance verification method, device and system | |
CN118365476B (en) | Intelligent gas-based enterprise information supervision method and device and Internet of things system | |
US20230401461A1 (en) | Intelligent Collaborative Decision Generation System with Link Prediction Assisted by User Feedback | |
CN109670641B (en) | Line loss uncovered transformer area prediction alarm method based on acquisition success rate |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |