CN118169547B

CN118169547B - Single-use circuit detection method and system for electric anastomat

Info

Publication number: CN118169547B
Application number: CN202410589676.2A
Authority: CN
Inventors: 赵虎成; 杨帅; 沈彬
Original assignee: Changzhou Yilong Information Technology Co ltd
Current assignee: Changzhou Yilong Information Technology Co ltd
Priority date: 2024-05-13
Filing date: 2024-05-13
Publication date: 2024-08-02
Anticipated expiration: 2044-05-13
Also published as: CN118169547A

Abstract

The invention relates to the technical field of circuit detection, in particular to a single-use circuit detection method and a system of an electric anastomat, wherein the method comprises the following steps: determining components in a design circuit diagram, acquiring connection relations among the components, and generating a detection reference template; the method comprises the steps that a detection reference template is in one-to-one correspondence with components and connection relations in an actual circuit board, circuit sample information of the actual circuit board is obtained, the sample information comprises component sample information and connection relation sample information, the circuit sample information is compared with the detection reference template according to detection item labels, and component comparison results and connection comparison results are respectively obtained; and analyzing the component comparison result and the connection comparison result through the defect detection model to obtain a circuit detection defect result. According to the invention, the quality and the safety of the single-use circuit are ensured to a certain extent under the condition of no passage, the single-use circuit is detected in batches in an intelligent mode, and the detection efficiency is improved.

Description

Single-use circuit detection method and system for electric anastomat

Technical Field

The invention relates to the technical field of circuit detection, in particular to a single-use circuit detection method and system of an electric anastomat.

Background

The electric anastomat nail bin is disposable equipment, can not be reused once suturing and cutting are completed in the operation process, and if repeated operation or repeated use is carried out, the electric anastomat nail bin can cause serious operation accidents such as tissue damage, bleeding and the like because the electric anastomat nail bin is not sutured after tissue cutting.

However, the single-use circuit is required to face detection problems when leaving the factory, and because the equipment has the disposable characteristic, the circuit is not easy to carry out path detection in actual detection, the quality of products is required to be strictly ensured before the circuit is put into use, and the use safety is improved.

The information disclosed in this background section is only for enhancement of understanding of the general background of the disclosure and is not to be taken as an admission or any form of suggestion that this information forms the prior art that is well known to a person skilled in the art.

Disclosure of Invention

The invention provides a single-use circuit detection method and a system for an electric anastomat, which can effectively solve the problems in the background technology.

In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:

a single use circuit detection method of an electric stapler, the method comprising:

Obtaining a design circuit diagram of an electric anastomat, determining components in the design circuit diagram, obtaining connection relations among the components, and generating a detection reference template;

The detection reference templates are in one-to-one correspondence with the components and the connection relations in the actual circuit board, and detection item labels are arranged on the components and the connection relations in the detection reference templates;

obtaining circuit sample information of an actual circuit board, wherein the sample information comprises component sample information and connection relation sample information, comparing the circuit sample information with the detection reference template according to the detection item label, and respectively obtaining a component comparison result and a connection comparison result;

and analyzing the component comparison result and the connection comparison result through a defect detection model to obtain a circuit detection defect result.

Further, obtaining circuit sample information of the actual circuit board includes:

Image acquisition is carried out on the actual circuit board, a circuit sample image is obtained, and circuit detection characteristics are extracted from the circuit sample image;

Classifying the circuit detection features according to the components and the connection relation in the detection reference template, and respectively integrating the circuit detection features into detection feature sets;

the detection feature set comprises position information, state information and welding point information of the components.

Further, preprocessing the circuit sample image, including:

Converting the circuit sample image into a gray level image, identifying edge features in the gray level image, and performing edge enhancement operation on the identified edge features;

Converting the enhanced gray level image into a binary image, selecting a proper threshold value, setting a welding spot area and a component area in the binary image as a first color, and setting a background area of the actual circuit board as a second color;

and performing integrity filling on the binary graph by adopting morphological operation.

Further, extracting circuit detection features from the circuit sample image includes:

converting the circuit sample image into a target matrix for convolutional neural network identification;

Generating a plurality of layers of convolution kernels of the convolutional neural network according to the detection reference template, wherein each convolution kernel identifies a single feature in the circuit sample image;

Setting the moving step length of the multi-layer convolution kernel, extracting the characteristics of the target matrix according to the multi-layer convolution kernel and the moving step length, and integrating the characteristics into the detection characteristic set respectively.

Further, analyzing the component comparison result and the connection comparison result through a defect detection model, including:

Acquiring historical single-use circuit data information, and acquiring a feedback result of the historical single-use circuit to establish a single-use circuit database;

deep learning is carried out on the single-use circuit database through a deep learning algorithm to obtain a component standard interval and a connection relation standard interval, wherein each circuit sorts the components and the connection relation according to the detection item label;

and respectively judging the positions of the component comparison result in the component standard interval and the connection comparison result in the connection relation standard interval to obtain a circuit detection defect result.

Further, generating a detection reference template includes:

Deep learning is carried out on the single-use circuit database, and fault categories and fault results are obtained;

integrating the fault category and the fault result with the detection reference template respectively based on the detection reference template to generate a plurality of fault circuit templates, wherein each fault circuit template corresponds to one fault category or fault result;

And matching the circuit sample information with the fault circuit template to obtain a fault circuit matching result, and feeding the fault circuit matching result back to the defect detection model.

Further, training the defect detection model includes:

Constructing a plurality of model training sets aiming at the same detection reference template through the single-use circuit database, and orderly identifying the model training sets;

Selecting one model training set as a model verification set according to the ordered recognition, and using the rest model training sets of the ordered recognition as training sets of the model verification set to obtain the performance evaluation of the current model verification set on the defect detection model;

Repeating the steps, so that each model training set is used as the model verification set, and optimizing the defect detection model through the performance evaluation of each model verification set.

Further, obtaining a component standard interval and a connection relation standard interval includes:

carrying out normalization calculation on the data of each feature of the components and the connection relation, and carrying out calculation on the mean value and standard deviation on the normalized data;

selecting a confidence level, and calculating the upper and lower boundaries of a standard interval according to the confidence level, the normalized data mean value and the standard deviation;

And converting the upper and lower boundaries of the standard interval of the normalized data into an original data space, and performing inverse normalization by using the mean value and the standard deviation in the original data space to obtain a component standard interval and a connection relation standard interval, wherein the original data space is a data space in a single-use circuit database.

Further, detecting the disposable fuse in the actual circuit board to obtain a fuse detection result, inputting the fuse detection result into the defect detection model, and analyzing the comparison result with the components and analyzing the connection comparison result to obtain a circuit detection defect result, including:

Acquiring material information and structure information of the disposable fuse, and determining fusing conditions of the fuse by the actual circuit board;

Establishing index management according to the material information classification information of the disposable fuse through a single-use circuit database;

and acquiring the corresponding relation between the various material information and the structural information of the fuse and the fusing condition through the deep learning algorithm, and analyzing the fuse in the actual circuit board according to the corresponding relation to acquire the fuse detection result.

A single use circuit detection system for an electric stapler, the system comprising:

the method comprises the steps of referring to a template generation module, obtaining a design circuit diagram of an electric anastomat, determining components in the design circuit diagram, obtaining connection relations among the components, and generating a detection reference template;

The detection label setting module is used for corresponding the detection reference templates to the components and the connection relations in the actual circuit board one by one, and setting detection item labels for the components and the connection relations in the detection reference templates;

The comparison result acquisition module is used for acquiring circuit sample information of the actual circuit board, wherein the sample information comprises component sample information and connection relation sample information, comparing the circuit sample information with the detection reference template according to the detection item label to respectively acquire a component comparison result and a connection comparison result;

And the circuit detection defect module is used for analyzing the component comparison result and the connection comparison result through a defect detection model to obtain a circuit detection defect result.

By the technical scheme of the invention, the following technical effects can be realized:

The quality and the security of single use circuit have been guaranteed to a certain extent to effective solution under the condition of no passageway, and adopt intelligent mode to detect single use circuit in batches, promote detection efficiency, reduce the unnecessary labour of manpower detection.

The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings may be obtained according to the drawings without inventive effort to those skilled in the art.

FIG. 1 is a schematic flow chart of a single use circuit detection method of an electric stapler;

FIG. 2 is a schematic flow chart of acquiring circuit sample information of an actual circuit board;

FIG. 3 is a schematic flow chart of analyzing component comparison results and connection comparison results;

FIG. 4 is a flow chart for training a defect detection model.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.

Example 1

As shown in fig. 1, the application provides a single-use circuit detection method of an electric anastomat, which comprises the following steps:

s10: obtaining a design circuit diagram of the electric anastomat, determining components in the design circuit diagram, obtaining connection relations among the components, and generating a detection reference template;

S20: the detection reference templates are in one-to-one correspondence with the components and the connection relations in the actual circuit board, and detection item labels are arranged on the components and the connection relations in the detection reference templates;

S30: obtaining circuit sample information of an actual circuit board, wherein the sample information comprises component sample information and connection relation sample information, comparing the circuit sample information with a detection reference template according to a detection item label, and respectively obtaining a component comparison result and a connection comparison result;

s40: and analyzing the component comparison result and the connection comparison result through the defect detection model to obtain a circuit detection defect result.

Specifically, the generation of the detection reference template is based on a design circuit diagram of the current production product, wherein important content of detection is to be embodied, such as components and connection relations, including resistors, capacitors, diodes, transistors, wires, welding, circuit boards, ground wires and the like, attention is paid to the fact that the structure of the design circuit diagram is consistent with the position of an actual circuit product besides determining detection items of the components and connection relations when the detection reference template is established, and detection item labels are arranged at corresponding positions, so that identification of defect identification is facilitated, and standard factory values in the reference template are clearly detected; in the detection of components and connection relations, sample information of an actual circuit board is obtained, wherein the sample information can comprise basic information and installation information of the components, for example, imaging technology, nondestructive detection, electronic test points and the like can be adopted, obtained data are subjected to one-to-one comparison according to detection item labels and detection reference templates, the compared data are input into a defect detection model to obtain a circuit defect result, the defect detection model is a data large model for evaluating defects of each single-use circuit product, the acquired data are subjected to primary data comparison and limitation through the detection reference templates, and then secondary analysis is performed through model large data, so that a more accurate analysis result is obtained.

According to the technical scheme, the quality and the safety of the single-use circuits are guaranteed to a certain extent under the condition of no passage, the single-use circuits are detected in batches in an intelligent mode, the detection efficiency is improved, and unnecessary labor force of manpower detection is reduced.

Further, as shown in fig. 2, obtaining circuit sample information of an actual circuit board includes:

S31: image acquisition is carried out on an actual circuit board, a circuit sample image is obtained, and circuit detection characteristics are extracted from the circuit sample image;

S32: classifying the circuit detection features according to the components and connection relations in the detection reference template, and respectively integrating the circuit detection features into detection feature sets;

s33: the detection feature set comprises position information, state information and welding point information of the components.

As a preference of the above embodiment, when acquiring an image of an actual circuit board, an acquisition position corresponding to the actual circuit board may be set according to a structure of the actual circuit board and a specific position of a component, image acquisition may be performed on the actual circuit board by using a camera or a scanner, etc. to ensure that a clear circuit board image is obtained, when acquiring an image, the whole circuit board is covered as much as possible, sufficient illumination is ensured to obtain a clear image, circuit detection features, such as a position, a shape, a size, a color, etc. of the component, a position and a state of a welding point, etc. are extracted, and the circuit detection features are classified according to the circuit detection features of the actual circuit board and the component and the connection relation in the detection reference template, so that detection features of similar components may be integrated into a set, such as putting features of all resistors together, and features of all connectors may be integrated into a set, or the same standard similar components may be integrated into a set, and at the same time, such as integrating and combining features of the connection relation, such as integrating features of the same kind and the connection relation, forming a feature set of connection relation, and converting the circuit detection features into a complex structure and a connection relation, and providing convenience for analyzing the circuit detection structure.

Further, preprocessing the circuit sample image includes:

converting the circuit sample image into a gray level image, identifying edge characteristics in the gray level image, and performing edge enhancement operation on the identified edge characteristics;

Converting the enhanced gray level image into a binary image, selecting a proper threshold value, setting a welding spot area and a component area in the binary image as a first color, and setting a background area of an actual circuit board as a second color;

As a preferred embodiment, the image processing software or programming language is used to convert the color image into a gray-scale image, the edge detection algorithm is used to identify the edge features in the image, the edge enhancement operation is performed on the identified edge features, a filter or enhancement algorithm can be used to enhance the contrast and definition of the edge, the image is binarized according to the enhanced gray-scale image by using a proper threshold method, the image is divided into black and white two colors, and each region in the binary image is respectively marked with different colors according to the characteristics of the welding spot, the component and the background, for example, the welding spot region and the component region are set to be white and the background region is set to be black. The morphological operation to supplement the integrity of the binary pattern may include an expansion operation, i.e., filling small holes that may exist in the solder joints, increasing the connectivity and integrity of the solder joints, helping to connect the dispersed solder joints, and allowing the solder joints to form a connected region; open operation, corrosion can eliminate small spots and isolated pixels, and re-expansion can fill holes in welding spots; closing operation, namely expanding to fill small holes and gaps in welding spots, and then corroding to eliminate irregular edges; when morphological operation is performed, the size of the core determines the operation effect, for example, for welding spot detection, the proper size of the core is selected according to the size and the distance of welding spots so as to achieve the optimal filling and connecting effect, the morphological operation can remove noise and isolated small areas in the image, and simultaneously fill holes and cracks in the image, so that the image is more continuous and complete, the processed image is helpful for accurately identifying and analyzing components and connection relations in the circuit board, and the stability and reliability of image processing are improved.

Generating a plurality of layers of convolution kernels of the convolution neural network according to the detection reference template, wherein each convolution kernel identifies a single feature in a circuit sample image;

setting a moving step length of the multi-layer convolution kernel, extracting features of the target matrix according to the multi-layer convolution kernel and the moving step length, and integrating the feature extraction into detection feature sets respectively.

In this embodiment, the circuit sample image is adjusted to an image with a specified size, usually square or rectangular, the image is converted to a format suitable for being input by a convolutional neural network, usually the image is converted to tensor, normalization processing is performed, so that the pixel value falls between 0 and 1, multiple layers of convolutional kernels of the convolutional neural network are designed according to components and connection relations in a detection reference template, each convolutional kernel is responsible for identifying a single feature in the circuit sample image, the size, depth and parameters of the convolutional kernel are designed, so that the features in the image, such as edges, textures, shapes and the like, can be effectively captured, the moving step length of the convolutional kernel, namely the moving distance of each time on a target matrix, the designed convolutional kernel is applied to the target matrix, the features of the image are extracted through convolution operation, the target matrix is slid according to the moving step length, the features of different positions of the image are sequentially extracted, the features extracted by the convolutional kernels are integrated, different features extracted by the convolutional kernels are formed, the multiple layers of the convolutional kernels are designed, the features in the multiple layers of the convolutional kernels are extracted, the various features in the detection feature set can be comprehensively captured, and the various features in the detection image can be represented by the multiple layers, and the robustness is improved, and the accuracy is improved.

Further, as shown in fig. 3, analyzing the component comparison result and the connection comparison result through the defect detection model includes:

S41: acquiring historical single-use circuit data information, and acquiring a feedback result of a historical single-use circuit to establish a single-use circuit database;

s42: deep learning is carried out on the single-use circuit database through a deep learning algorithm to obtain a component standard interval and a connection relation standard interval, wherein each circuit sorts the components and the connection relation according to the detection item label;

S43: and respectively judging the positions of the component comparison result in the component standard interval and the connection comparison result in the connection relation standard interval to obtain a circuit detection defect result.

Based on the above embodiment, a large amount of data information of the historical single-use circuit is collected, including parameters, connection relation, detection results and the like of components, the collected historical data information is sorted and stored, a single-use circuit database is built for subsequent analysis and learning, a deep learning algorithm such as a convolutional neural network, a cyclic neural network and the like is used for deep learning of the single-use circuit database, the data information of the historical single-use circuit is used as a training sample, the parameters, connection relation and detection results of the components are included, a proper neural network structure is designed, so that the model can accurately learn characteristics and defects of the circuit, the component comparison result and the connection comparison result are compared with a component standard interval and a connection relation standard interval, the positions of the component comparison result and the connection comparison result in the standard interval are judged according to the comparison result, whether abnormality or defect exists is determined if the component comparison result or the connection comparison result exceeds the standard interval range, namely the existence of the defect is determined, and otherwise the abnormality is determined to be normal.

Further, generating the detection reference template includes:

deep learning is carried out on the single-use circuit database, and fault types and fault results are obtained;

In this embodiment, the built single-use circuit database is utilized to perform deep learning, a deep learning algorithm is used to train circuit samples in the database, fault types and fault results of the circuit samples are marked, the model can accurately identify various fault types and fault results in the circuit through training and learning of a large number of samples, the acquired fault types and fault results are integrated according to information such as components and connection relations in the detection reference templates, a plurality of fault circuit templates are generated according to integrated data, each template corresponds to one fault type or fault result, when the templates are generated, the templates are ensured to be capable of covering various possible fault conditions, the templates are associated with specific fault types or fault results, the acquired circuit sample information is matched with the plurality of fault circuit templates, the matching results are fed back to the defect detection model according to the matching degree of the circuit samples and the templates, the corresponding fault types or fault results are used for subsequent defect detection and analysis, the single-use data are used as detection reference templates, the fault conditions can be accurately detected by the multiple fault source detection models, and the fault models can be comprehensively identified, and the fault conditions can be accurately detected in the circuit database.

Further, as shown in fig. 4, training the defect detection model includes:

s401: constructing a plurality of model training sets aiming at the same detection reference template through a single-use circuit database, and orderly identifying the model training sets;

s402: selecting one model training set as a model verification set according to the ordered recognition, and taking the rest model training sets of the ordered recognition as training sets of the model verification set to obtain performance evaluation of the current model verification set on the defect detection model;

S403: repeating the steps to enable each model training set to serve as a model verification set, and optimizing the defect detection model through performance evaluation of each model verification set.

As a preference of the above embodiment, a plurality of circuit samples are selected from the single-use circuit database as training sets, a plurality of model training sets are constructed for the same detection reference template according to the selected training sets, each model training set includes a set of circuit samples whose characteristics match with the corresponding detection reference template, one of the constructed model training sets is selected as model verification set, the remaining training sets are used as training sets, performance evaluation is performed on the defect detection model by using the selected verification set, including calculation of indexes such as accuracy, recall, F1 value, etc., performance of the model is determined according to the evaluation result, evaluation indexes are recorded, and the above steps are iterated until the model performance reaches the expected requirement or convergence.

Carrying out normalization calculation on the data of each feature of the components and the connection relation, and carrying out calculation on the mean value and the standard deviation on the normalized data;

Selecting a confidence level, and calculating the upper and lower boundaries of the standard interval according to the confidence level, the normalized data mean value and the standard deviation;

And converting the upper and lower boundaries of the standard interval of the normalized data into an original data space, and performing inverse normalization by using the mean value and the standard deviation in the original data space to obtain a component standard interval and a connection relation standard interval, wherein the original data space is the data space in the single-use circuit database.

As a preference to the above embodiment, the raw data of each feature of the components and connection relationships is normalized, and a common normalization method includes scaling the data to 0,1] range or using a z-score normalization method, and for the normalized data, calculating the mean and standard deviation for each feature, the mean value and standard deviation of the normalized data are used as the standard interval calculation basis of the components and connection relation, the confidence level and calculation standard interval are selected and used as the statistical analysis basis, and the upper and lower boundaries of the standard interval are calculated according to the confidence level, the mean value and standard deviation of the normalized data and the z-score value corresponding to the confidence level, wherein the upper and lower boundaries can be calculated according to the following formula:

upper boundary = mean + z-score standard deviation

Lower boundary = mean-z-score standard deviation

And converting the upper and lower boundaries of the standard interval of the normalized data into an original data space, and performing inverse normalization operation by using the mean value and standard deviation of the data space in the single-use circuit database to obtain the component standard interval and the connection relation standard interval. Wherein a confidence level is selected, commonly comprising 95% and 99.7%, the confidence level representing the degree of confidence in the data distribution, and for a 95% confidence level, the boundary of the standard interval is typically the mean plus or minus two standard deviations; for a 99.7% confidence level, the boundaries of the standard interval are typically the mean plus or minus three standard deviations.

Further, detecting the disposable fuse in the actual circuit board to obtain a fuse detection result, inputting the fuse detection result into a defect detection model, and analyzing the comparison result with the components and analyzing the connection comparison result to obtain a circuit detection defect result, including:

Acquiring material information and structure information of a disposable fuse, and determining fusing conditions of an actual circuit board on the fuse;

Establishing index management according to the material information classification information of the disposable fuses through a single-use circuit database;

And obtaining the corresponding relation between various material information and structural information of the fuse and the fusing condition through a deep learning algorithm, and analyzing the fuse in the actual circuit board according to the corresponding relation to obtain a fuse detection result.

In this embodiment, materials and structural information of the disposable fuse including material type, size, appearance characteristics, and the like are determined through a design drawing of the circuit board or physical inspection of the actual circuit board, and at the same time, fusing conditions of the disposable fuse including rated current, overload protection, and the like are determined according to design requirements or related standards of the circuit board.

Example two

Based on the same inventive concept as the single-use circuit detection method of the electric stapler in the foregoing embodiment, the invention also provides a single-use circuit detection system of the electric stapler, which comprises:

The reference template generation module is used for obtaining a design circuit diagram of the electric anastomat, determining components in the design circuit diagram, obtaining connection relations among the components and generating a detection reference template;

The detection label setting module is used for corresponding the detection reference templates to the components and the connection relations in the actual circuit board one by one and setting detection item labels for the components and the connection relations in the detection reference templates;

The comparison result acquisition module is used for acquiring circuit sample information of the actual circuit board, wherein the sample information comprises component sample information and connection relation sample information, comparing the circuit sample information with a detection reference template according to a detection item label to respectively acquire a component comparison result and a connection comparison result;

and the circuit detection defect module is used for analyzing the component comparison result and the connection comparison result through the defect detection model to obtain a circuit detection defect result.

The adjusting system can effectively realize the single-use circuit detection method of the electric anastomat, and has the technical effects as described in the embodiment, and the description is omitted herein.

Although the application has been described in connection with specific features and embodiments thereof, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the application. Accordingly, the specification and figures are merely exemplary illustrations of the present application as defined in the appended claims and are considered to cover any and all modifications, variations, combinations, or equivalents that fall within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the scope of the application. Thus, the present application is intended to include such modifications and alterations insofar as they come within the scope of the application or the equivalents thereof.

Claims

1. A method for detecting a single-use circuit of an electric stapler, the method comprising:

Obtaining circuit sample information of an actual circuit board comprises the following steps:

The detection feature set comprises position information, state information and welding point information of the components;

Analyzing the component comparison result and the connection comparison result through a defect detection model to obtain a circuit detection defect result, wherein the circuit detection defect result comprises:

respectively judging the positions of the component comparison result in the component standard interval and the connection comparison result in the connection relation standard interval to obtain a circuit detection defect result;

Generating a detection reference template, comprising:

2. The single-use circuit detection method of an electric stapler according to claim 1, wherein preprocessing the circuit sample image includes:

and performing integrity filling on the binary image by adopting morphological operation.

3. The single-use circuit detection method of an electric stapler according to claim 1 or 2, wherein extracting a circuit detection feature from the circuit sample image includes:

4. The single use circuit detection method of an electric stapler of claim 1, wherein training the defect detection model comprises:

and circularly selecting each training set as a model verification set, and using the rest orderly-identified model training sets as training sets of the current model verification set, so that each model training set is used as the model verification set, and optimizing the defect detection model through the performance evaluation of each model verification set.

5. The single-use circuit detection method of an electric stapler according to claim 1, wherein obtaining a component standard interval and a connection relation standard interval comprises:

6. The single-use circuit inspection method of an electric stapler according to claim 1, wherein inspecting the disposable fuse in the actual circuit board to obtain a fuse inspection result, and inputting the fuse inspection result to the defect inspection model, analyzing the comparison result with the components and analyzing the connection comparison result to obtain a circuit inspection defect result, comprises:

7. A single use circuit detection system for an electric stapler according to claim 1, wherein the single use circuit detection method for an electric stapler according to claim 1 is adopted, and the system comprises: