Work order generation method and device, intelligent work order filling system and work order processing system

Abstract

The invention provides a work order generation method and device, an intelligent work order filling system and a work order processing system.The work order generation method comprises the steps that a reference work order with the highest relevancy with equipment basic fault information input by a user is obtained from an enterprise asset management system in a keyword matching search mode; if a reference work order with relatively high relevancy with the equipment basic fault information input by the user cannot be obtained from the enterprise asset management system, determining a historical work order with the highest similarity in a vector database as the reference work order according to vector data corresponding to the equipment basic fault information input by the user; and according to at least one of the confirmation information and the modification information of the user on the reference work order, generating a current work order, the current work order comprising the equipment parameters and the processing method information. According to the method, traditional keyword matching search and large language vector retrieval technologies are ingeniously combined, so that the problems of long time consumption and unstable quality existing in manual work order filling in the prior art are solved.

Description

Technical Field

[0001] The present application belongs to the technical field of work order generation, and specifically relates to a work order generation method and device, an intelligent work order filling system, and a work order processing system. Background Art

[0002] In traditional nuclear power plants, when users report equipment for repair, they simply describe the cause of the failure. Although there is a digital enterprise asset management (EAM) system that supports online operations, maintenance personnel still need to make a lot of preparations when they receive a maintenance work order. Maintenance personnel need to complete the unit, plant information, sensitive areas, system version, equipment classification, nuclear safety plan, quality plan, defect analysis, tools and consumables, work steps and dozens of other fields based on the basic work order information submitted, such as equipment code and equipment name, and then conduct manual cross-checks to confirm that the work order information and operation steps are complete before going to the site for maintenance. The entire maintenance preparation is time-consuming and lengthy. It often takes more than 2 hours to fill in the relevant information for a work order; and there is a certain degree of personalization and human bias in manual compilation, which affects the stability of the quality of work order writing. Summary of the invention

[0003] In view of this, the embodiments of the present application are committed to providing a work order generation method and device, an intelligent work order filling system and a work order processing system, which cleverly combine traditional keyword matching search and large language vector retrieval technology to solve the problems of long time consumption and unstable quality in the existing manual filling of work orders.

[0004] The first aspect of the present application provides a work order generation method, which includes: using a keyword matching search method to obtain a reference work order with the highest relevance to the basic equipment fault information input by the user from the enterprise asset management system. The basic equipment fault information includes key fields that define the problem; if the reference work order with the highest relevance to the basic equipment fault information input by the user cannot be obtained from the enterprise asset management system, then according to the vector data corresponding to the basic equipment fault information input by the user, the historical work order with the highest relevance in the vector database is confirmed as the reference work order; according to at least one of the user's confirmation information and modification information of the reference work order, a current work order is generated, and the current work order includes equipment parameters and processing method information.

[0005] In a specific embodiment of the present application, before generating the current work order based on at least one of the user's confirmation information and modification information on the reference work order, the work order generation method also includes: if the reference work order is incomplete, obtaining device matching key information from the vector database based on vector similarity matching, and supplementing the reference work order based on the device matching key information.

[0006] In a specific embodiment of the present application, if the reference work order is incomplete, the device matching key information is obtained from the vector database based on vector similarity matching, and the reference work order is supplemented and improved according to the device matching key information. The work order generation method also includes: if the reference work order does not completely cover the work order information, a separate supplement rule is used to separately supplement the uncovered part of the work order information.

[0007] In a specific embodiment of the present application, after generating the current work order based on at least one of the user's confirmation information and modification information of the reference work order, the work order generation method also includes: writing the current work order back to the enterprise asset management system through an RPA robot so that the enterprise asset management system issues an order based on the current work order.

[0008] In a specific embodiment of the present application, after generating the current work order based on at least one of the user's confirmation information and modification information of the reference work order, the work order generation method also includes: regularly synchronizing the completed work order historical data of the enterprise asset management system, and completing and improving the current work order through key information query and vectorized correlation matching.

[0009] The second aspect of the present application provides a work order generation device, which includes an acquisition module, a confirmation module and a generation module. The acquisition module is used to use a keyword matching search method to obtain a reference work order with the highest relevance to the basic equipment fault information input by the user from the enterprise asset management system, and the basic equipment fault information includes the key fields of the problem definition, the hardware equipment number and the fault type; the confirmation module is used to confirm the historical work order with the highest relevance in the vector database as the reference work order according to the vector data corresponding to the basic equipment fault information input by the user when the reference work order with the highest relevance to the basic equipment fault information input by the user cannot be obtained from the enterprise asset management system; the generation module is used to generate a current work order according to at least one of the user's confirmation information and modification information of the reference work order, and the current work order includes equipment parameters and processing method information.

[0010] The third aspect of the present application provides an intelligent work order filling system, which includes a processor and a memory. The processor is used to execute the work order generation method provided in the first aspect of the present application. The memory is used to store executable instructions of the processor.

[0011] The fourth aspect of the present application provides a work order processing system, which includes: the intelligent work order filling system, the enterprise asset management system and the vector database provided in the third aspect of the present application. The enterprise asset management system is connected to the intelligent work order filling system. The vector database is connected to the intelligent work order filling system and the enterprise asset management system.

[0012] The fifth aspect of the present application provides a computer-readable storage medium on which computer executable instructions are stored. When the executable instructions are executed by a processor, the work order generation method provided in the first aspect of the present application is implemented.

[0013] The sixth aspect of the present application provides a computer program product, including a computer program / instructions, which, when executed by a processor, implements the work order generation method provided in the first aspect of the present application.

[0014] The beneficial effect of the technical solution of the present application is that: through this method based on vector similarity matching and large language model generation, precise keyword matching such as equipment code / task type precise query is used. If there is no output, it is combined with the historical work order vectorization retrieval matching to find the work order that meets the similarity requirements and fill in the reference, so that the most relevant processing experience can be automatically filled in for the new work order, and the operation steps and risk warnings are summarized and generated, which greatly improves the efficiency and accuracy of work order preparation and reduces the risk of human error caused by lack of experience or omissions. At the same time, this ability to automatically learn and transfer knowledge from massive historical work orders also enables the system to continuously self-learn and continuously iterate and optimize, continuously accumulate and precipitate work order processing expertise, thereby forming a closed loop to improve the accuracy and efficiency of work order generation. The embodiment of the present application combines the advantages of correlation retrieval and custom rule logic, cleverly combines traditional keyword matching search and large language vector retrieval technology, and uses the equipment work order data of the nuclear power EAM system to realize autonomous learning of historical work order rules and automatic generation of processing methods, thereby realizing intelligent generation of work order generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The figure is a flow chart of a work order generation method provided in one embodiment of the present application.

[0016] Figure 2 Shown is a schematic diagram of a vectorized multi-dimensional space provided by an embodiment of the present application.

[0017] Figure 3 Shown is a schematic diagram of work order keyword vectorization provided by an embodiment of the present application.

[0018] Figure 4 Shown is a schematic diagram of a work order processing system provided in one embodiment of the present application. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0020] At least one embodiment of the present application provides a work order generation method, and the execution subject of the work order generation method can be a processor or a server, etc. The following takes the execution subject of the work order generation method as a processor of an intelligent work order filling system as an example for illustration. Figure 1 , the work order generation method includes the following steps.

[0021] S10: Using a keyword matching search method, obtain a reference work order with the highest relevance to the basic equipment failure information input by the user from the enterprise asset management system. The basic equipment failure information includes key fields that define the problem.

[0022] Specifically, when a user reports a device for repair, the enterprise asset management system receives basic device fault information entered by the user (corresponding to Figure 4 In S1), the enterprise asset management system searches for and matches the reference work order (i.e., the original work order) with the highest relevance to the basic equipment failure information input by the user through the basic equipment failure information, and sends the reference work order to the intelligent work order filling system (corresponding to Figure 4 S2 in ).

[0023] It should be noted that the key fields for defining the problem include but are not limited to the faulty device number, hardware device number, device model, fault type, and fault handling method.

[0024] S20: If the reference work order with the highest relevance to the basic equipment fault information input by the user cannot be obtained from the enterprise asset management system, the historical work order with the highest relevance in the vector database is confirmed as the reference work order based on the vector data corresponding to the basic equipment fault information input by the user.

[0025] Specifically, if S10 has no output, the enterprise asset management system will use the vector data corresponding to the basic equipment fault information input by the user to quickly retrieve historical work orders in the vector database that are highly similar to the basic equipment fault information through semantic vector similarity matching. These highly similar historical work orders may involve the same or similar equipment models and fault phenomena, which contain valuable processing methods and work risk analysis experience. The enterprise asset management system will retrieve the most relevant historical work orders and output them to the intelligent work order filling system, so that the intelligent work order filling system can receive the historical work order and use it as a reference work order.

[0026] For example, refer to Figure 3 , the enterprise asset management system can obtain historical work orders, and convert the work order information of historical work orders into vector data according to the defined key fields, such as: equipment type, unit, engineering task title, production date, system, work order type], etc. into vector data, and the vector data format is such as [-0.005, 0.012, -0.014, 0.006, -0.013...], and store it in the vector database (corresponding to Figure 4 (S3).

[0027] It should be noted that the intelligent work order filling system can be integrated with the open source large language model GLM-4-9B capability to reintegrate and generate supplementary information for work orders.

[0028] In some embodiments, the most relevant historical work order may be a historical work order with a similarity not less than a minimum threshold and ranked first. The minimum threshold may be set according to actual needs, for example, the minimum threshold may range from 80% to 90%.

[0029] For example, the work order data stored in the vector database is in a multidimensional space. Figure 2 , the diagram can only show 3 dimensions, but actually there are about 10 dimensions of space according to the defined key fields. When searching for similarity, the points on the vector coordinates will be closer if the work order information is closer. Conversely, by vectorizing the current basic work order and then searching for the close vector points (work orders), similar work orders can be obtained, thereby extracting the original repair work order data.

[0030] S30: Generate a current work order according to at least one of the confirmation information and modification information of the reference work order by the user, wherein the current work order includes equipment parameters and processing method information.

[0031] Specifically, if the reference work order is correct, users such as maintenance personnel can confirm the reference work order on the intelligent work order filling system and generate the current work order. If the reference work order needs to be supplemented or is incorrect, the user can modify the reference work order on the intelligent work order filling system and generate the current work order (corresponding to Figure 4 S4).

[0032] According to the technical solution provided by the embodiment of the present application, through this method based on vector similarity matching and large language model generation, precise keyword matching such as equipment code / task type precise query is used. If there is no output, the historical work order vectorization retrieval matching is combined to find the work order that meets the similarity requirements and fill in the reference, so that the most relevant processing experience can be automatically filled in for the new work order, and the operation steps and risk warnings are summarized and generated, which greatly improves the efficiency and accuracy of work order processing and reduces the risks caused by lack of experience or omissions. At the same time, this ability to automatically learn and transfer knowledge from massive historical work orders also enables the system to continuously self-learn and continuously optimize, continuously accumulate and precipitate work order processing expertise, thereby forming a closed loop to improve the accuracy and efficiency of work order generation. The embodiment of the present application combines the advantages of correlation retrieval and custom rule logic, cleverly combines traditional keyword matching search and large language vector retrieval technology, and uses nuclear power scene equipment work order data to realize a system that can autonomously learn historical work order rules and automatically generate processing methods and problem analysis, and realize an intelligent work order generation method.

[0033] In the work order generation method provided in at least one embodiment of the present application, before the above S30, the work order generation method also includes S21.

[0034] S21: If the reference work order is incomplete, the key device matching information is obtained from the vector database based on the vector similarity matching, and the reference work order is supplemented and improved according to the key device matching information.

[0035] The vector database stores equipment regulations, manuals, and equipment information obtained from the equipment information system. Equipment information includes but is not limited to equipment type, equipment model, drawings, and experience feedback information. The vector database is stored on the server of the intelligent work order filling system.

[0036] In the embodiment of the present application, the incomplete information is combined with the equipment's procedural materials to verify and supplement key information, and further comprehensive application solves the problems of difficult work order filling and complex information.

[0037] In the work order generation method provided in at least one embodiment of the present application, after the above S21, the work order generation method also includes S22.

[0038] S22: If the reference work order does not completely cover the work order information, a separate supplement rule is used to separately supplement the uncovered portion of the work order information.

[0039] It should be noted that, generally, the work order information can be completely covered by the above S21. There are also some additional fields, which are separate supplementary rules. For example, if the identification type is A, the wrong information is B. This rule is maintained separately.

[0040] After the above steps, the information of the reference work order has been integrated and generated, waiting for manual confirmation.

[0041] In the work order generation method provided in at least one embodiment of the present application, after the above S30, the work order generation method also includes S40.

[0042] S40: The current work order is written back to the enterprise asset management system through the RPA robot, so that the enterprise asset management system issues an order based on the current work order.

[0043] For example, the current work order is directly filled in the same work order ID in the enterprise asset management system (corresponding to Figure 4 In S5), after success, the work order reaches the ready state. Maintenance personnel can issue orders based on the work order information and perform on-site maintenance (corresponding to Figure 4 (S6).

[0044] It should be noted that the RPA robot method adopts the Robotic Process Automation (RPA) method. RPA is a business process automation technology based on software robots and artificial intelligence (AI).

[0045] In the work order generation method provided in at least one embodiment of the present application, after the above S30, the work order generation method further includes S50.

[0046] S50: Regularly synchronize the historical data of completed work orders in the enterprise asset management system, and complete the current work orders through key information query and vectorized correlation matching.

[0047] Specifically, the intelligent work order filling system will regularly synchronize the historical data of completed work orders in the enterprise asset management system, and complete or update the information in the current work order such as equipment parameters and automatically generated processing method information through key information query and vectorized correlation matching (using Figures S10 to S30). This process can interact with the vector database to obtain the corresponding equipment parameters and processing method information.

[0048] In some embodiments, the enterprise asset management system periodically synchronizes the completed work order history data to the vector database. These historical work orders extract multiple (10+) key fields and content and store them in the multidimensional vector database Faiss. The vector database maps the data into a bunch of multidimensional data coordinates. The data with higher content similarity is closer in the corresponding coordinate space. Figure 2 The storage and mapping can provide data similarity matching when a new original completed work order is matched next time.

[0049] At least one embodiment of the present application further provides a work order generating device, which includes an acquisition module, a confirmation module and a generation module.

[0050] The acquisition module is used to obtain the reference work order with the highest relevance to the basic equipment fault information input by the user from the enterprise asset management system by using a keyword matching search method. The basic equipment fault information includes key fields that define the problem.

[0051] The confirmation module is used to confirm the historical work order with the highest relevance in the vector database as the reference work order based on the vector data corresponding to the basic equipment fault information input by the user if the reference work order with the highest relevance to the basic equipment fault information input by the user cannot be obtained from the enterprise asset management system.

[0052] The generation module is used to generate a current work order according to at least one of the confirmation information and modification information of the reference work order by the user, and the current work order includes equipment parameters and processing method information.

[0053] In the work order generation device provided in at least one embodiment of the present application, the work order generation device further includes a first supplementation and improvement module. The acquisition module is also used to obtain device matching key information from a vector database according to device regulations or manuals based on vector similarity matching if the reference work order is incomplete. The first supplementation and improvement module is used to supplement and improve the reference work order according to the device matching key information.

[0054] In the work order generation device provided in at least one embodiment of the present application, the first supplementation and improvement module is further used to use a separate supplementation rule to separately supplement the uncovered part of the work order information if the reference work order does not completely cover the work order information.

[0055] In the work order generation device provided in at least one embodiment of the present application, the work order generation device further includes a write-back module. The write-back module is used to write the current work order back to the enterprise asset management system through the RPA robot, so that the enterprise asset management system issues an order according to the current work order.

[0056] In the work order generation device provided in at least one embodiment of the present application, the work order generation device further includes a second supplementation and improvement module. The second supplementation and improvement module is used to regularly synchronize the completed work order history data of the enterprise asset management system, and to complete and improve the current work order through key information query and vectorized correlation matching.

[0057] It should be noted that the work order generation device includes the corresponding technical features of the work order generation method provided in the above embodiments of the present application, and can achieve the corresponding technical effects, which will not be repeated here.

[0058] At least one embodiment of the present application further provides an intelligent work order filling system, the intelligent work order filling system comprising a processor and a memory.

[0059] The intelligent work order filling system includes a processor, which further includes one or more processors, and a memory resource represented by a memory for storing instructions executable by the processor, such as an application. The application stored in the memory may include one or more modules each corresponding to a set of instructions. In addition, the processor is configured to execute the instructions to perform the above-mentioned work order generation method.

[0060] The intelligent work order filling system may also include a power supply component configured to perform power management of the intelligent work order filling system, a wired or wireless network interface configured to connect the intelligent work order filling system to the network, and an input / output (I / O) interface. The intelligent work order filling system may operate based on an operating system stored in the memory, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM or the like.

[0061] The intelligent work order filling system provided in the embodiment of the present application uses natural language processing technology to automatically extract key field information from historical work order texts, such as equipment code, equipment model, fault phenomenon, operation steps, work risks, etc., and semantically vectorizes these key fields and stores them in a vector database.

[0062] At least one embodiment of the present application further provides a work order processing system, which includes the intelligent work order filling system, the enterprise asset management system and the vector database in the above embodiment. The enterprise asset management system is connected to the intelligent work order filling system. The vector database is connected to the intelligent work order filling system and the enterprise asset management system.

[0063] Specifically, in the process of intelligent filling and generation of work orders, the enterprise asset management system receives the initial information of equipment, faults, etc. input by the user; next, the intelligent work order filling system uses the powerful semantic understanding and generation capabilities of the large language model to learn the processing steps and risk warnings of these similar historical work orders, and automatically generates detailed operation instructions and precautions for the current work order. In specific implementation, the key fields of similar work orders can be constructed into instructions (Prompt) in the form of natural language together with user input, and input into the pre-trained large language model. The above large model language can use an open source large language model such as GLM-4-9B, etc., and the model directly generates the text content required for the current work order.

[0064] It should be noted that users can enter user information on the integrated platform, and the intelligent work order filling system can implement single sign-on and feedback to the integrated platform after receiving the user information.

[0065] The work order processing system includes the intelligent work order filling system in the above-mentioned embodiment, and can implement the work order generation method in the above-mentioned embodiment, that is, it includes the corresponding technical features of the work order generation method in the above-mentioned embodiment, and can achieve the corresponding technical effects, which will not be repeated here.

[0066] At least one embodiment of the present application further provides a computer-readable storage medium having computer executable instructions stored thereon. When the instructions in the storage medium are executed by the processor of the above-mentioned intelligent work order filling system, the above-mentioned intelligent work order filling system can execute any one of the work order generation methods in the above-mentioned embodiments.

[0067] At least one embodiment of the present application further provides a computer program product, including a computer program / instruction, which implements any one of the work order generation methods in the above embodiments when the computer program / instruction is executed by a processor.

[0068] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0069] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program check codes.

[0070] Technical personnel in the relevant field can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device and intelligent work order filling system can refer to the logical process in the above-mentioned work order generation method embodiment, and will not be repeated here.

[0071] It should be noted that the combination of the various technical features in the embodiments of the present application is not limited to the combination described in the embodiments of the present application or the combination described in the specific embodiments, and all technical features described in the present application can be freely combined or combined in any way unless there is a contradiction between them.

[0072] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "comprises" and "includes" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0073] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0074] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A work order generation method, characterized in that: include: Using a keyword matching search method, obtain the reference work order with the highest relevance to the basic equipment fault information input by the user from the enterprise asset management system. The basic equipment fault information includes key fields that define the problem. If the reference work order with the highest relevance to the basic equipment fault information input by the user cannot be obtained from the enterprise asset management system, the historical work order with the highest relevance in the vector database is confirmed as the reference work order based on the vector data corresponding to the basic equipment fault information input by the user; A current work order is generated according to at least one of the confirmation information and the modification information of the reference work order by the user, and the current work order includes equipment parameters and processing method information.

2. The work order generation method according to claim 1, characterized in that: Before generating the current work order according to at least one of the confirmation information and the modification information of the reference work order by the user, the method further includes: If the reference work order is incomplete, the key device matching information is obtained from the vector database based on vector similarity matching, and the reference work order is supplemented and improved according to the key device matching information.

3. The work order generation method according to claim 2, characterized in that: If the reference work order is incomplete, the method further includes obtaining key device matching information from a vector database based on vector similarity matching, and supplementing and improving the reference work order according to the key device matching information. If the reference work order does not fully cover the work order information, a separate supplement rule is used to separately supplement the uncovered part of the work order information.

4. The work order generation method according to any one of claims 1 to 3, characterized in that: After generating the current work order according to at least one of the confirmation information and the modification information of the reference work order by the user, the method further includes: The current work order is written back to the enterprise asset management system through the RPA robot so that the enterprise asset management system can issue an order based on the current work order.

5. The work order generation method according to any one of claims 1 to 3, characterized in that: After generating the current work order according to at least one of the confirmation information and the modification information of the reference work order by the user, wherein the current work order includes the equipment parameters and the processing method information, the method further includes: Regularly synchronize the historical data of completed work orders in the enterprise asset management system, and complete the current work orders through key information query and vectorized correlation matching.

6. A work order generating device, characterized in that: include: An acquisition module is used to acquire a reference work order corresponding to the basic equipment fault information input by the user from the enterprise asset management system by using a keyword matching search method, wherein the basic equipment fault information includes the key fields of the problem definition, namely, the hardware equipment number and the fault type; A confirmation module, for confirming the historical work order with the highest similarity in the vector database as the reference work order based on the vector data corresponding to the basic equipment fault information input by the user if the reference work order corresponding to the basic equipment fault information input by the user cannot be obtained from the enterprise asset management system; The generation module generates a current work order according to at least one of the confirmation information and modification information of the reference work order by the user, wherein the current work order includes equipment parameters and processing method information.

7. An intelligent work order filling system, characterized in that: include: A processor, configured to execute the work order generation method according to any one of claims 1 to 5; as well as A memory is used to store executable instructions of the processor.

8. A work order processing system, characterized in that: include: The intelligent work order filling system as claimed in claim 7; An enterprise asset management system, which is in communication connection with the intelligent work order filling system; The vector database is communicatively connected with the intelligent work order filling system and the enterprise asset management system.

9. A computer-readable storage medium having computer executable instructions stored thereon, characterized in that: When the executable instructions are executed by the processor, the work order generation method described in any one of claims 1 to 5 is implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the work order generation method according to any one of claims 1 to 5 is implemented.

Images