Visual large model cue word engineering compiling method for drawing substation terminal strip wiring identification, storage medium and equipment

Abstract

The invention discloses a visual large model cue word engineering compiling method for drawing transformer substation terminal strip wiring identification, a storage medium and equipment, and the method comprises the steps: collecting transformer substation drawings, carrying out the function module division of the transformer substation terminal strip wiring on each drawing, carrying out the information link sorting of the divided function modules, and carrying out the information link sorting of the divided function modules; compiling a cue word project of the visual large model; inputting the collected transformer substation drawing into the identification large model, and identifying the terminal strip wiring on the transformer substation drawing by using the compiled cue word project; comparing the identified terminal strip wiring with the terminal strip wiring on the corresponding transformer substation drawing, and if the comparison result is inconsistent, adjusting the cue word project of the visual large model; otherwise, solidifying the compiled prompt word project; and digitalizing the wiring of the terminal strip on the transformer substation drawing by using the solidified compiled prompt word project. According to the invention, the identification accuracy of terminal strip wiring on the transformer substation drawing is improved, and the digitalization process of the transformer substation drawing is accelerated.

Description

Technical Field

[0001] The present invention relates to the technical field of prompt word engineering, and in particular to a visual large-scale model prompt word engineering compilation method, storage medium and device for identifying terminal block wiring in a drawing of a transformer substation. Background Art

[0002] With the deepening digital transformation of power systems, there is an urgent need to build high-precision data models of electrical connection relationships to support the construction of intelligent operation and maintenance systems. Consequently, the need for digital processing of substation terminal block wiring drawings is becoming increasingly urgent. Current mainstream solutions utilize convolutional neural network-based image recognition technologies, such as YOLO and OCR. While these technologies can locate basic components and extract text on substation drawings, they face significant technical bottlenecks in practical engineering applications. First, image recognition technology is limited by its shallow network architecture, making it difficult to effectively analyze the topological relationships between drawing elements. Second, when designing different substation drawings following the same drafting specifications, there are still differences in layout style, annotation habits, and other dimensions, resulting in insufficient generalization capabilities of image recognition technology.

[0003] The visual macro model is an intelligent recognition framework capable of hierarchical feature parsing. It integrates 3D convolutional kernel stacking technology, a dynamic attention mechanism, and a cross-modal feature fusion module. Using a deep neural network with over 10 billion parameters, it can understand the semantics of multi-level drawings. It can also transfer and learn layout features from drawings of different design institutes and infer electrical connection logic, effectively overcoming the traditional algorithm's strong dependence on fixed drawing formats. The Chinese patent application, entitled "Visual Question Answering Method and Related Apparatus Based on Multimodal Macro Model," with publication number CN118607644A, encodes the original image's features to form a sequence of feature vectors, reconstructs the image, and then combines the knowledge content information in a multimodal knowledge base to predict questions and answers based on the reconstructed image and the question information. This visual question answering method requires images to have distinct image features and is generally suitable for general object recognition. However, the terminal block wiring on substation drawings is highly specialized, with high element repetition and unclear features. This method cannot guide the visual macro model to understand specialized terminal block wiring drawings, resulting in reduced drawing recognition accuracy.

[0004] Therefore, when applying visual large-scale models to recognize diverse terminal block wiring diagrams, refined and efficient prompt word engineering is crucial. Accurate prompt word engineering can efficiently guide the visual large-scale model to understand, reason, and analyze the content of substation drawings, reducing recognition errors. However, there is currently no established methodology for compiling prompt word engineering for visual large-scale model terminal block wiring diagram recognition applications. Summary of the Invention

[0005] In response to the problems existing in the prior art, the present invention provides a method, storage medium and device for compiling a visual large model prompt word project for identifying terminal block wiring in drawings. By compiling the prompt word project, the visual large model is guided to understand the diverse information of terminal block wiring on substation drawings, thereby improving the recognition accuracy of terminal block wiring on substation drawings, adapting to diverse substation drawings, and accelerating the digitization process of substation drawings.

[0006] To achieve the above technical objectives, the present invention adopts the following technical solution: a method for compiling visual large model prompt words for identifying terminal block wiring in drawings of substations, comprising the following steps:

[0007] Step S1: Collect substation drawings and divide the substation terminal block wiring on each drawing into functional modules;

[0008] Step S2: sorting out information links of the divided functional modules;

[0009] Step S3: compile a prompt word project for the visual model based on the sorted information links;

[0010] Step S4: input the collected substation drawings into the visual macro model, and use the compiled prompt word project to identify the terminal block wiring on the substation drawings;

[0011] Step S5: Compare the identified terminal block wiring with the terminal block wiring on the corresponding power station drawing. If the comparison results are inconsistent, adjust the prompt word project of the visual large model; otherwise, solidify the compiled prompt word project;

[0012] Step S6: digitize the terminal block wiring on the substation drawing using the solidified edited prompt word project.

[0013] Furthermore, the functional modules divided in step S1 include: a terminal block identification module, a terminal identification module, a cable identification module, a loop identification module, a terminal short circuit identification module, a terminal jumper identification module and a connection relationship identification module.

[0014] Furthermore, step S2 includes: sorting out the relationship between the input and output of each functional module, and sorting out the working sequence of each functional module.

[0015] Furthermore, the working order of each functional module is: first identify the terminal row name in the terminal row identification module, then identify the terminal number in the terminal identification module, the loop number in the loop identification module and the cable number in the cable identification module, then judge the connection relationship between the terminal and the cable according to the connection relationship identification module, and finally identify the terminal short circuit situation through the terminal short circuit identification module and the terminal jump situation through the terminal jump identification module.

[0016] Furthermore, the prompt word project includes: compiling prompt words to identify the electrical content of each functional module, including: identifying cable identification, terminal block identification, terminal number identification, loop identification, terminal connection identification, terminal short circuit identification and jumper identification; compiling prompt words for anti-interference instructions of each functional module; compiling prompt words for the logic and connection relationship in each functional module; compiling prompt words for output examples.

[0017] Furthermore, the prompt words for compiling the output example include: terminal block number: XX; terminal number: XX; cable number: XX; loop number: XX; connection relationship: XX.

[0018] Furthermore, the process of adjusting the prompt word project of the visual large model is as follows: if one or more of the terminal block number, terminal number, cable number and loop number output by the visual large model are inconsistent with the corresponding strain station drawings, adjust the prompt words for identifying the electrical connotation of each functional module, and optimize the prompt words for the anti-interference instructions of each functional module; if the connection relationship output by the visual large model is inconsistent with the corresponding strain station drawings, adjust the prompt words for the logic and connection relationship in each functional module.

[0019] Furthermore, the present invention also provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the method for compiling visual large model prompt words for identifying wiring of a substation terminal block in a drawing.

[0020] Furthermore, the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, the method for engineering compilation of visual large model prompt words for identifying terminal block wiring in drawings is implemented.

[0021] Compared with the prior art, the present invention has the following beneficial effects: the visual large model prompt word engineering compilation method for identifying terminal block wiring in substation drawings of the present invention divides the functional modules of the terminal block wiring on the substation drawings and sorts out the information links, understands the design logic of the terminal block wiring, and compiles the prompt word engineering of the visual large model to guide the visual large model to understand the diversified information of the terminal block wiring on the substation drawings, and improves the visual large model's ability to understand highly professional and diversified substation drawings, thereby improving the recognition accuracy of the terminal block wiring on the substation drawings and speeding up the digitization process of the substation drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flow chart of the method for compiling visual large model prompt words for identifying substation terminal block wiring in drawings of the present invention. DETAILED DESCRIPTION

[0023] The technical solution of the present invention will be further explained below with reference to the accompanying drawings.

[0024] like Figure 1 This is a flow chart of a method for compiling visual large model prompt words for identifying terminal block wiring in a drawing of a substation according to the present invention. The method specifically includes the following steps:

[0025] Step S1: Collect substation drawings and divide the substation terminal block wiring on each drawing into functional modules. Each divided functional module represents a functional area or component in the substation drawing, including: terminal block identification module, terminal identification module, cable identification module, loop identification module, terminal short circuit identification module, terminal jumper identification module and connection relationship identification module, so as to ensure that the division of each functional module is clear and facilitate the subsequent compilation of prompt words.

[0026] Step S2, sorting out the information links of the divided functional modules, including: sorting out the relationship between the input and output of each functional module, sorting out the working order of each functional module, specifically, first identifying the terminal block name in the terminal block identification module, then identifying the terminal number in the terminal identification module, the loop number in the loop identification module and the cable number in the cable identification module, then judging the connection relationship between the terminal and the cable according to the connection relationship identification module, and finally identifying the terminal short circuit situation through the terminal short circuit identification module and the terminal jump situation through the terminal jump identification module.

[0027] Step S3: Compile a prompt word project for the visual big model based on the sorted information links. Guide the compilation of the prompt word project through the sorted information links, thereby guiding the visual big model to highly accurately identify the terminal block wiring on the substation drawing, thereby improving the recognition efficiency and accuracy of the terminal block wiring drawing. The prompt word project for the compiled visual large model should be specific and clear, and be able to guide the large model to accurately identify each element in the substation drawing, including: identification of terminal numbers, positioning of wiring positions, and judgment of connection relationships. Therefore, the compiled prompt word project needs to include: compiling prompt words to identify the electrical connotations of each functional module, including: identification of cable identification, terminal block identification, terminal number identification, loop identification, terminal connection identification, terminal short circuit identification and jumper identification; compiling prompt words for the anti-interference instructions of each functional module to prevent the visual large model from making mistakes or missing identification during the identification process of each functional module in the substation drawing; compiling prompt words for the logic and connection relationship in each functional module; compiling prompt words for output examples, including: terminal block number: XX; terminal number: XX; cable number: XX; loop number: XX; connection relationship: XX, so as to determine the output content and format of the visual large model and support the directional connection of the input and output information flows of each functional module.

[0028] Step S4: input the collected substation drawings into the visual macro model, and use the compiled prompt word project to identify the terminal block wiring on the substation drawings.

[0029] Step S5: Compare the identified terminal block wiring with the corresponding terminal block wiring on the power station drawing, detect the various electrical characteristics and the connection logic of the terminals and cables, and verify the validity of the prompt word project. If the comparison results are inconsistent, adjust the prompt word project of the visual large model; otherwise, solidify the compiled prompt word project.

[0030] In one technical solution of the present invention, the process of adjusting the prompt word project of the visual large model is as follows: if one or more of the terminal block number, terminal number, cable number and loop number output by the visual large model are inconsistent with the corresponding strain station drawing, the prompt words for identifying the electrical connotation of each functional module are adjusted, and the prompt words for the anti-interference instructions of each functional module are optimized; if the connection relationship output by the visual large model is inconsistent with the corresponding strain station drawing, the prompt words for the logic and connection relationship in each functional module are adjusted.

[0031] Step S6: digitize the terminal block wiring on the substation drawing using the solidified edited prompt word project.

[0032] The visual large model prompt word engineering compilation method for identifying terminal block wiring in substation drawings of the present invention divides the functional modules of the terminal block wiring on the substation drawings and sorts out the information links to understand the design logic of the terminal block wiring, thereby compiling the prompt word engineering of the visual large model, guiding the visual large model to understand the diversified information of the terminal block wiring on the substation drawings, and improving the visual large model's ability to understand highly professional and diversified substation drawings, thereby improving the recognition accuracy of the terminal block wiring on the substation drawings and accelerating the digitization process of the substation drawings.

[0033] In one technical solution of the present invention, a computer-readable storage medium is further provided, storing a computer program, wherein the computer program enables a computer to execute the method for engineering compilation of visual large model prompt words for identifying wiring of a substation terminal block in a drawing.

[0034] In one technical solution of the present invention, an electronic device is also provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for engineering compilation of visual large model prompt words for identifying wiring of substation terminal blocks in drawings is implemented.

[0035] In the embodiments disclosed herein, computer storage media can be tangible media that can contain or store programs for use by or in conjunction with an instruction execution system, device, or apparatus. Computer storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. More specific examples of computer storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0036] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this application 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.

[0037] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A visual large model prompt word engineering compilation method for identifying terminal block wiring in a drawing of a substation, characterized by: The steps include: Step S1: Collect substation drawings and divide the substation terminal block wiring on each drawing into functional modules; Step S2: sorting out information links of the divided functional modules; Step S3: compile a prompt word project for the visual model based on the sorted information links; Step S4: input the collected substation drawings into the visual macro model, and use the compiled prompt word project to identify the terminal block wiring on the substation drawings; Step S5: Compare the identified terminal block wiring with the terminal block wiring on the corresponding power station drawing. If the comparison results are inconsistent, adjust the prompt word project of the visual large model; otherwise, solidify the compiled prompt word project; Step S6: digitize the terminal block wiring on the substation drawing using the solidified edited prompt word project.

2. The method for compiling visual large model prompt words for identifying terminal block wiring in drawings according to claim 1 is characterized in that: The functional modules divided in step S1 include: a terminal block identification module, a terminal identification module, a cable identification module, a loop identification module, a terminal short circuit identification module, a terminal jumper identification module and a connection relationship identification module.

3. The method for compiling visual large model prompt words for identifying terminal block wiring in drawings of substations according to claim 2 is characterized in that: Step S2 includes: sorting out the relationship between the input and output of each functional module, and sorting out the working sequence of each functional module.

4. The method for compiling visual large model prompt words for identifying terminal block wiring in drawings of substations according to claim 3 is characterized in that: The working order of each functional module is as follows: first identify the terminal strip name in the terminal strip identification module, then identify the terminal number in the terminal identification module, the loop number in the loop identification module and the cable number in the cable identification module, then determine the connection relationship between the terminal and the cable according to the connection relationship identification module, and finally identify the terminal short circuit situation through the terminal short circuit identification module and the terminal jumper identification module.

5. The method for compiling visual large model prompt words for identifying terminal block wiring in drawings of substations according to claim 4 is characterized in that: The prompt word project includes: compiling prompt words to identify the electrical content of each functional module, including: identifying cable identification, terminal block identification, terminal number identification, loop identification, terminal connection identification, terminal short circuit identification and jumper identification; compiling prompt words for anti-interference instructions of each functional module; compiling prompt words for the logic and connection relationship in each functional module; and compiling prompt words for output examples.

6. The method for compiling visual large model prompt words for identifying terminal block wiring in drawings of substations according to claim 5 is characterized in that: The prompt words for compiling output examples include: terminal block number: XX; terminal number: XX; cable number: XX; loop number: XX; connection relationship: XX.

7. A visual large model prompt word engineering compilation method for identifying terminal block wiring in a drawing of a substation according to claim 6, characterized in that: The process of adjusting the prompt words of the visual large model is as follows: if one or more of the terminal block number, terminal number, cable number and loop number output by the visual large model are inconsistent with the corresponding power station drawings, adjust the prompt words for identifying the electrical connotation of each functional module and optimize the prompt words of the anti-interference instructions of each functional module; if the connection relationship output by the visual large model is inconsistent with the corresponding power station drawings, adjust the prompt words of the logic and connection relationship in each functional module.

8. A computer-readable storage medium storing a computer program, characterized in that: The computer program enables the computer to execute the visual large model prompt word engineering compilation method for drawing substation terminal block wiring identification as described in any one of claims 1 to 7.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for engineering compilation of visual large-scale model prompt words for identifying terminal block wiring in a drawing of a substation is implemented.

Images