Automatic identification method and system for electrical drawing

Abstract

The invention provides an electrical drawing automatic identification method and system, and relates to the field of electrical drawing identification, and the electrical drawing automatic identification method comprises the following steps: 1, image preprocessing: collecting a multi-source original new energy electrical drawing, converting the multi-source original new energy electrical drawing into a uniform format, eliminating noise, and enhancing features; step 2, identifying and marking symbols; step 3, topological relation modeling; step 4, parameter verification; according to the method, the topological structure diagram of the electrical system is automatically generated by analyzing the connection relation between the elements, manual intervention is not needed, the recognition result and the BOM database of the new energy vehicle are integrated, automatic input and management of element information are achieved, production efficiency can be improved, human errors are reduced, and the method is suitable for popularization and application. The method can be expanded to automatic identification of transformer substation wiring diagrams and industrial robot electrical drawings, and promotes digital transformation of the manufacturing industry.

Description

Technical Field

[0001] The present invention relates to the field of new energy electrical drawing recognition, and in particular to an electrical drawing automatic recognition method and system. Background Art

[0002] With the rapid development of the new energy vehicle industry, the complexity of its electrical systems has increased significantly. Electrical drawings for new energy vehicles contain numerous key components (such as relays, contactors, contacts, sensitive switches, and electromagnetic clutches). These components are represented by specific graphical symbols and form complex topological structures. For example, the graphical symbols for relays and contactors are used to control the on / off state of circuits, the symbols for sensitive switches and sensors are used to sense environmental changes, and the symbols for electromagnetic clutches and brakes involve motion control. The design and arrangement of these symbols directly impact the safety and efficiency of the electrical system.

[0003] However, the recognition and analysis of traditional electrical drawings mainly rely on manual operations, which has the following problems: manual review and annotation of electrical drawings is time-consuming, especially for the dense component symbols and complex topological relationships in the high-voltage system diagrams of new energy vehicles. Manual recognition is easily affected by factors such as fatigue and experience differences, resulting in misjudgment of component classification or connection relationships. In addition, traditional OCR (optical character recognition) and CV (computer vision) rely only on image features for recognition and cannot verify the rationality of component parameters (such as wire cross-sectional area and voltage drop) and topological relationships.

[0004] Therefore, it is necessary to provide a new electrical drawing automatic recognition method and system to solve the above technical problems. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides a method and system for automatic recognition of electrical drawings.

[0006] The present invention provides an automatic recognition method for electrical drawings, comprising the following steps:

[0007] Step 1: Image preprocessing: collect original new energy electrical drawings from multiple sources and convert them into a unified format to eliminate noise and enhance features;

[0008] Step 2: Symbol recognition and annotation: Combine geometric invariant moments with electromagnetic field features, use a symbol feature enhancement processor to enhance symbol features, and apply an improved YOLOv5 network to identify components in new energy electrical drawings;

[0009] Step 3: Topological relationship modeling: skeleton refinement of new energy electrical drawings, extraction of line segments and nodes in the circuit, construction of a hierarchical graph neural network, and analysis of the topological relationship in the circuit;

[0010] Step 4: Parameter verification: Calculate the component parameters in the circuit based on electromagnetic field theory and verify the current balance at each node to ensure that the current in the circuit satisfies Kirchhoff's current law. Verify the current balance by calculating the sum of the current flowing into and out of the node, calculate the path voltage drop, and verify whether it matches the power supply voltage.

[0011] Step 5: Data output: Integrate the recognition results with the new energy vehicle BOM database, automatically enter and manage component information, and generate editable files.

[0012] Furthermore, the method of constructing a hierarchical graph neural network includes: dynamically adjusting edge weights by combining Euclidean distance and semantic similarity, and updating node features by using gated graph convolution.

[0013] Further, according to the electromagnetic field theory, the parameters of components in the circuit are calculated, including obtaining them based on the cross-sectional area of ​​the wire, the resistance, length, resistivity of the wire and the contactor contact pressure.

[0014] In a further embodiment, the current balance is verified by Kirchhoff's current law.

[0015] Further in the method, the skeleton thinning includes iteratively thinning the binary image, preserving the center line.

[0016] In a further method, line segment extraction includes detecting straight line segments using Hough transform and extracting parameters of line segments connected by wires.

[0017] Furthermore, the multi-source original new energy electrical drawings include scanned copies, DXF and DWG.

[0018] Furthermore, the image preprocessing specifically includes the following steps:

[0019] S1. Multi-source input: Convert DXF and DWG files to high-resolution PNG images, preserving vector information.

[0020] S2. Scanning correction: Correct the tilt of the scanned document;

[0021] S3, Enhanced Features: Improve image contrast through the CLAHE algorithm and enhance the symbols and line features in the drawings;

[0022] S4. Use the Sobel operator to extract high-frequency details, calculate the gradient of the image in the x and y directions, and obtain the edge intensity map to further highlight the key information in the drawing.

[0023] Another aspect of the present invention provides an electrical drawing automatic recognition system, comprising: an acquisition module for acquiring multi-source new energy electrical drawings and performing image preprocessing;

[0024] The symbol recognition module is used to accurately identify components and symbols by combining geometric invariant moment calculation, electromagnetic field feature extraction, and an improved YOLOv5 network;

[0025] Topology modeling module, used to analyze the topological relationship of the circuit using skeleton refinement, line segment extraction and hierarchical graph neural network;

[0026] Verification module, used to verify circuit parameters based on the parameter calculation model of electromagnetic field theory;

[0027] The data output module integrates the recognition results with the BOM database and generates editable drawing files.

[0028] Compared with related technologies, the electrical drawing automatic recognition method and system provided by the present invention have the following beneficial effects:

[0029] 1. The present invention combines Ohm's law, Kirchhoff's law and electromagnetic field characteristics (such as wire cross-sectional area and contact resistance) to verify the rationality of recognition results and reduce misjudgments caused by image noise or symbol occlusion.

[0030] 2. The present invention automatically generates a topological diagram of the electrical system by analyzing the connection relationship between components (such as wire connection and node distribution) without human intervention. The recognition results are integrated with the BOM (bill of materials) database of new energy vehicles to realize the automatic entry and management of component information, which helps to improve production efficiency and reduce human errors. It can be extended to the automated recognition of substation wiring diagrams and industrial robot electrical drawings, and promote the digital transformation of the manufacturing industry.

[0031] 3. Based on electromagnetic field theory, the present invention calculates the parameters of components in the circuit, such as inductance and capacitance, taking into account factors such as the geometric shape and material properties of the components to improve the accuracy of parameter calculation, verifies the current balance at each node, ensures that the current in the circuit satisfies Kirchhoff's current law, verifies the current balance by calculating the sum of the current flowing into and out of the node, ensures the normal operation of the circuit, calculates the path voltage drop, verifies whether it matches the power supply voltage, and ensures the rationality and performance of the circuit design by calculating the voltage drop of each path in the circuit. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a structural diagram of the electrical drawing automatic recognition system provided by the present invention;

[0033] Figure 2 A flowchart of the method for automatic identification of electrical drawings provided by the present invention;

[0034] Figure 3 This is a flowchart of the image preprocessing provided by the present invention. DETAILED DESCRIPTION

[0035] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0036] Please refer to Figure 1 、 Figure 2 、 Figure 3 ,in, Figure 1 This is a structural diagram of the electrical drawing automatic recognition system provided by the present invention; Figure 2 A flowchart of the method for automatic identification of electrical drawings provided by the present invention; Figure 3 This is a flowchart of the image preprocessing provided by the present invention.

[0037] Example 1

[0038] In the specific implementation process, Figure 1 As shown, an electrical drawing automatic recognition system includes: an acquisition module for acquiring multi-source new energy electrical drawings and performing image preprocessing;

[0039] The symbol recognition module is used to accurately identify components and symbols by combining geometric invariant moment calculation, electromagnetic field feature extraction, and an improved YOLOv5 network;

[0040] Topology modeling module, used to analyze the topological relationship of the circuit using skeleton refinement, line segment extraction and hierarchical graph neural network;

[0041] Verification module, used to verify circuit parameters based on the parameter calculation model of electromagnetic field theory;

[0042] The data output module integrates the recognition results with the BOM database and generates editable drawing files.

[0043] Example 2

[0044] In the specific implementation process, Figure 2 as well as Figure 3 As shown, the present invention provides an automatic recognition method for electrical drawings, which includes the following steps:

[0045] Step 1: Image preprocessing: collect original new energy electrical drawings from multiple sources, including scanned copies, DXF, and DWG, and convert them into a unified format to eliminate noise and enhance features.

[0046] It should be noted that image preprocessing specifically includes the following steps:

[0047] S1. Multi-source input: Convert DXF and DWG files to high-resolution PNG images, preserving vector information (such as line width and color coding);

[0048] S2. Scan correction: Use the findHomography function to correct the tilt of the scanned document:

[0049] H = findHomography(pts_src, pts_dst), where H is a unidirectional matrix, pts_src is the corner point of the original image, and pts_dst is the corner point of the target image;

[0050] S3. Enhanced features: Improve image contrast through CLAHE algorithm and enhance the symbols and line features in the drawings: I CLAHE =CLAHE(I gray ,clipLimit=2.0);

[0051] S4. Use the Sobel operator to extract high-frequency details, calculate the gradient of the image in the x and y directions, and obtain an edge intensity map to further highlight the key information in the drawing:

[0052]

[0053] Final edge strength:

[0054] Step 2: Symbol recognition and annotation: Combine geometric invariant moments with electromagnetic field features, use a symbol feature enhancement processor to enhance symbol features, and apply an improved YOLOv5 network to identify components in new energy electrical drawings;

[0055] It should be noted that the geometric invariant moment Hu moment is calculated as follows:

[0056] Seven invariant moments η1 to η7 are extracted from the symbol area for symbol shape description and matching. These invariant moments are invariant to translation, rotation, and scaling, and can accurately describe the shape characteristics of the symbol. The formula is as follows:

[0057]

[0058] Among them, μ pq is the center distance;

[0059] Electromagnetic field feature extraction: Calculate the inductance L and capacitance C of the conductor area using the following formula:

[0060]

[0061] Where N is the number of turns, A is the area, l is the length, and d is the spacing;

[0062] Example:

[0063] For single-strand wire (N=1, A=5mm 2 , l = 1m), the inductance is:

[0064] Improved YOLOv5 network:

[0065] S101. Apply a four-scale detection mechanism to increase sensitivity to extremely small components.

[0066] S102. Introduce a feature transmission channel to integrate more features and improve component recognition accuracy.

[0067] S103. Use the K-means++ clustering algorithm to calculate the anchor box to improve positioning accuracy.

[0068] S104, using SIoU (SCYLLA-intersection over union) positioning loss function to speed up network convergence;

[0069] Step 3: Topological relationship modeling: skeleton refinement of new energy electrical drawings, extraction of line segments and nodes in the circuit, construction of a hierarchical graph neural network, and analysis of the topological relationship in the circuit;

[0070] It should be noted that skeleton refinement involves iteratively thinning the binary image to preserve the centerline;

[0071] Line segment extraction involves detecting straight line segments using Hough transform: ρ = xcosθ + ysinθ

[0072] Extract the parameters (ρ, θ) of the wire connection segment, such as ρ = 100px, θ = 45°;

[0073] Step 4: Parameter verification: Calculate the component parameters in the circuit based on electromagnetic field theory and verify the current balance at each node to ensure that the current in the circuit satisfies Kirchhoff's current law. Verify the current balance by calculating the sum of the current flowing into and out of the node, calculate the path voltage drop, and verify whether it matches the power supply voltage.

[0074] Step 5: Data output: Integrate the recognition results with the new energy vehicle BOM database, automatically enter and manage component information, and generate editable files.

[0075] Integrating the identification results with the new energy vehicle BOM (Bill of Materials) database enables automatic entry and management of component information, which helps improve production efficiency and reduce human errors. For example, it can automatically identify the connection paths between batteries, motors, and inverters and calculate the path voltage drop to ensure that the design meets engineering specifications.

[0076] In a specific implementation process, building a hierarchical graph neural network includes: dynamically adjusting edge weights by combining Euclidean distance and semantic similarity:

[0077] Example: If the Euclidean distance d between nodes i and j is ij =50px, semantic similarity cos(f i ,f j )=0.8, then the weight is:

[0078] Update node features using gated graph convolution:

[0079] It is further explained that, according to electromagnetic field theory, the parameters of components in the circuit are calculated including obtaining them based on the cross-sectional area A of the wire, the resistance R of the wire, the length L, the resistivity ρ and the contact pressure P of the contactor;

[0080] Furthermore, the formula for the cross-sectional area of ​​the conductor is as follows:

[0081] Example: If the wire resistance R = 0.1Ω, the length L = 1m, and the copper resistivity ρ = 1.72×10 -8 Ω·m, then the cross-sectional area is:

[0082] Furthermore, the contactor contact pressure formula is:

[0083] Example: If the contact resistance R c =0.01Ω, contact radius a=0.5mm, material resistivity ρ=1.72×10 -8 Ω·m, then the contact pressure is:

[0084] Furthermore, the current balance is verified using Kirchhoff's current law:

[0085]

[0086] Example: The current input at node i is Output is I out =10A, satisfying Kirchhoff's current law.

[0087] Further, voltage drop verification:

[0088] Example: If the total resistance of the path R total =0.2Ω, current I = 50A, power supply voltage V supply =400V, then the voltage drop is:

[0089] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, or of course, by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the relevant technology, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiment.

[0090] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, from all points of view, the embodiments should be regarded as illustrative and non-restrictive. The scope of the present invention is defined by the appended claims rather than the foregoing description. It is intended that all changes that come within the meaning and range of equivalents of the claims be included in the present invention, and any reference signs in the claims should not be construed as limiting the claims to which they relate.

[0091] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for automatic recognition of electrical drawings, characterized in that: The method comprises the following steps: Step 1: Image preprocessing: collect original new energy electrical drawings from multiple sources and convert them into a unified format to eliminate noise and enhance features; Step 2: Symbol recognition and annotation: Combine geometric invariant moments with electromagnetic field features, use a symbol feature enhancement processor to enhance symbol features, and apply an improved YOLOv5 network to identify components in new energy electrical drawings; Step 3: Topological relationship modeling: skeleton refinement of new energy electrical drawings, extraction of line segments and nodes in the circuit, construction of a hierarchical graph neural network, and analysis of the topological relationship in the circuit; Step 4: Parameter verification: Calculate the component parameters in the circuit based on electromagnetic field theory and verify the current balance at each node to ensure that the current in the circuit satisfies Kirchhoff's current law. Verify the current balance by calculating the sum of the current flowing into and out of the node, calculate the path voltage drop, and verify whether it matches the power supply voltage. Step 5: Data output: Integrate the recognition results with the new energy vehicle BOM database, automatically enter and manage component information, and generate editable files.

2. The method for automatic identification of electrical drawings according to claim 1, characterized in that: The construction of the hierarchical graph neural network includes: dynamically adjusting edge weights by combining Euclidean distance and semantic similarity, and updating node features by using gated graph convolution.

3. The method for automatic recognition of electrical drawings according to claim 2, characterized in that: Calculating the component parameters in the circuit according to the electromagnetic field theory includes obtaining them according to the wire cross-sectional area, wire resistance, length, resistivity and contactor contact pressure.

4. The method for automatic recognition of electrical drawings according to claim 3, characterized in that: The current balance is verified using Kirchhoff's current law.

5. The method for automatic recognition of electrical drawings according to claim 4, characterized in that: The skeleton refinement includes iteratively thinning the binary image and retaining the center line.

6. The method for automatic recognition of electrical drawings according to claim 5, characterized in that: Line segment extraction includes detecting straight line segments using Hough transform and extracting parameters of wire connecting line segments.

7. The method for automatic recognition of electrical drawings according to claim 6, characterized in that: The multi-source original new energy electrical drawings include scans, DXF and DWG.

8. The method for automatic recognition of electrical drawings according to claim 7, characterized in that: The image preprocessing specifically includes the following steps: S1. Multi-source input: Convert DXF and DWG files to high-resolution PNG images, preserving vector information. S2. Scanning correction: Correct the tilt of the scanned document; S3, Enhanced Features: Improve image contrast through the CLAHE algorithm and enhance the symbols and line features in the drawings; S4. Use the Sobel operator to extract high-frequency details, calculate the gradient of the image in the x and y directions, and obtain the edge intensity map to further highlight the key information in the drawing.

9. An electrical drawing automatic recognition system using the method according to any one of claims 1 to 8, characterized in that: include: Acquisition module, used to collect multi-source new energy electrical drawings and perform image preprocessing; The symbol recognition module is used to accurately identify components and symbols by combining geometric invariant moment calculation, electromagnetic field feature extraction, and an improved YOLOv5 network; Topology modeling module, used to analyze the topological relationship of the circuit using skeleton refinement, line segment extraction and hierarchical graph neural network; Verification module, used to verify circuit parameters based on the parameter calculation model of electromagnetic field theory; The data output module integrates the recognition results with the BOM database and generates editable drawing files.

Images