Electrical conduit flame retardant property detection system and method based on deep learning
By building a flame retardant performance detection system based on deep learning, and combining simulation technology to adjust material properties and analyze characteristics, the problem of traditional inefficiency is solved, and fast and accurate positioning of the cause of flame retardant performance is achieved.
Patent Information
- Application Number
- CN202510328063.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-19
- Publication Date
- 2025-07-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The flame retardant performance detection efficiency of traditional electrical casings is inefficient and it is difficult to quickly lock the specific reasons that affect the flame retardant performance.
Build an electrical casing flame retardant performance detection system based on deep learning. Through the flame retardant performance detection module, anomaly verification module and casing attribute detection module, combined with simulation technology, deep analysis of temperature characteristics, smoke characteristics and image characteristics, simulate material attribute adjustment, conduct similarity analysis, and determine actual material problems.
It has achieved efficient and accurate discovery of the reasons for insufficient flame retardant performance of electrical casings, improved detection efficiency, and ensured the targeted and accurate detection.
Smart Images
Figure CN120354094A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of flame retardancy performance testing of electrical conduit pipes, and more specifically, it relates to a detection system and method for the flame retardancy performance of electrical conduit pipes based on deep learning. Background Art
[0002] As an important protective material in electrical engineering, the flame retardancy performance of electrical conduit pipes is directly related to the safety and reliability of electrical systems. After the traditional electrical conduit pipes are tested for flame retardancy performance and it is found that the flame retardancy performance of the electrical conduit pipes does not meet the standards, it is necessary to detect the reasons for the insufficient flame retardancy performance. The current detection method is to check the material properties of the electrical conduit pipes one by one according to the conventional detection sequence. This method based on the conventional detection sequence, although to a certain extent, can find out the reasons for the insufficient flame retardancy performance, has obvious deficiencies: low detection efficiency: Since the material properties of electrical conduit pipes involve multiple aspects, detecting one by one according to the conventional sequence is not only time-consuming and laborious, but also easy to miss key information, resulting in low detection efficiency. It is difficult to quickly lock the problem: The traditional detection methods often lack pertinence and cannot quickly and accurately lock the specific reasons for the insufficient flame retardancy performance.
[0003] Therefore, there is an urgent need for a detection system and method for the flame retardancy performance of electrical conduit pipes based on deep learning. The present invention is proposed based on such a background, aiming to construct a detection system for the flame retardancy performance of electrical conduit pipes based on deep learning, and realize in-depth analysis of temperature characteristics, smoke characteristics, and image characteristics during the flame retardancy detection process of electrical conduit pipes, so as to accurately and efficiently determine the actual reasons affecting the flame retardancy performance of electrical conduit pipes. Summary of the Invention
[0004] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a detection system and method for the flame retardancy performance of electrical conduit pipes based on deep learning.
[0005] To achieve the above purpose, the present invention provides the following technical solutions:
[0006] A detection system for the flame retardancy performance of electrical conduit pipes based on deep learning, including a flame retardancy performance detection module, a flame retardancy performance abnormality verification module, and a conduit pipe attribute detection module;
[0007] The flame retardancy performance detection module: selects an electrical conduit pipe to be tested for flame retardancy performance, tests the flame retardancy performance of the electrical conduit pipe, and determines whether it is a conduit pipe with abnormal flame retardancy performance;
[0008] The flame retardancy performance abnormality verification module: when the electrical conduit pipe is marked as a conduit pipe with abnormal flame retardancy performance, constructs a conduit pipe simulation combustion model, the conduit pipe simulation combustion model generates multiple conduit pipe attribute modification strategies, and selects a potential abnormal method from the conduit pipe attribute modification strategies;
[0009] The casing attribute detection module: extracts the material attributes targeted by all potential abnormal methods and marks them as possible attributes, determines the flame retardancy performance influence index of each possible attribute, sorts all possible attributes in descending order according to the value of the flame retardancy performance influence index, and sequentially detects the possible attributes of the electrical casing according to the sorting order.
[0010] Furthermore, conduct a flame retardancy performance test on the electrical casing to determine whether it is a casing with abnormal flame retardancy performance: determine the model of the electrical casing, determine the combustion conditions corresponding to the electrical casing, burn the electrical casing under the combustion conditions, and after the combustion ends, determine the combustion performance index of the electrical casing. Set a combustion performance threshold index. When the combustion performance index of the electrical casing is less than the combustion performance threshold index, mark the electrical casing as a casing with abnormal flame retardancy performance.
[0011] Furthermore, the combustion performance index of the electrical casing is determined based on the following method: determine the comprehensive combustion characteristic set of the electrical casing, obtain a combustion performance analysis model, use the comprehensive combustion characteristic set as the input data of the combustion performance analysis model, and the combustion performance analysis model outputs to obtain the combustion performance index of the electrical casing.
[0012] Furthermore, the comprehensive combustion characteristic set of the electrical casing is determined based on the following method: obtain the temperature characteristics, smoke characteristics, and image characteristics of the electrical casing, and integrate the temperature characteristics, smoke characteristics, and image characteristics into a comprehensive combustion characteristic set in the form of a characteristic set.
[0013] Furthermore, the temperature characteristics of the electrical casing are obtained based on the following method: collect the temperature data of the electrical casing during the combustion process in real time. After the combustion ends, preprocess and extract features from the collected temperature data to obtain the temperature characteristics of the electrical casing;
[0014] The smoke characteristics of the electrical casing are obtained based on the following method: collect the smoke data of the electrical casing during the combustion process in real time. After the combustion ends, preprocess and extract features from the collected smoke data to obtain the smoke characteristics of the electrical casing;
[0015] The image characteristics of the electrical casing are obtained based on the following method: take videos of the electrical casing during the combustion process in real time. After the combustion ends, perform frame extraction on the video based on a fixed frame extraction interval to obtain multiple combustion images. Extract features from the multiple combustion images to obtain multiple single-frame features, and perform feature fusion processing on the multiple single-frame features to obtain the image characteristics of the electrical casing.
[0016] Further, select a potential abnormal method in the casing attribute modification strategy: determine the similarity index of the flame retardant performance of each casing attribute modification strategy, set the standard similarity index of the flame retardant performance, and when the similarity index of the flame retardant performance of the casing attribute modification strategy is greater than the standard similarity index of the flame retardant performance, mark this casing attribute modification strategy as a potential abnormal method.
[0017] Further, the similarity index of the flame retardant performance of the casing attribute modification strategy is determined based on the following method: the casing simulation combustion model conducts a simulated combustion test on the casing entity based on a casing attribute modification strategy. After the combustion test, obtain the comprehensive combustion characteristic set of the casing entity, and synchronously obtain the comprehensive combustion characteristic set of the casing with abnormal flame retardant performance. Input the comprehensive combustion characteristic set of the casing entity and the comprehensive combustion characteristic set of the casing with abnormal flame retardant performance into the flame retardant performance similarity analysis model synchronously. The flame retardant performance similarity analysis model outputs the similarity index of the flame retardant performance of this casing attribute modification strategy.
[0018] Further, the influence index of the flame retardant performance of the material attribute is determined based on the following method: select a possible attribute, mark the remaining possible attributes as comparison attributes, mark the total number of potential abnormal methods for this possible attribute as number(potential), obtain the attribute relationship graph, compare the possible attribute with all comparison attributes pairwise. When there is an association between the possible attribute and a comparison attribute in the attribute relationship graph, increase the material interaction number by one, and mark the material interaction number as number(effect). Calculate the influence index of the flame retardant performance of this material attribute index(flame), where a1 is the first coefficient and a2 is the second coefficient. Calculate the influence index of the flame retardant performance of this material attribute index(flame), where a1 is the first coefficient and a2 is the second coefficient.
[0019] Further, the method for detecting the flame retardant performance of electrical conduit pipes based on deep learning includes the following steps:
[0020] Step 1: Select an electrical conduit pipe to be tested for flame retardant performance, conduct a flame retardant performance test on the electrical conduit pipe, and determine whether it is a casing with abnormal flame retardant performance;
[0021] Step 2: When the electrical conduit pipe is marked as a casing with abnormal flame retardant performance, construct a casing simulation combustion model;
[0022] Step 3: The casing simulation combustion model generates multiple casing attribute modification strategies, and select potential abnormal methods from the casing attribute modification strategies;
[0023] Step 4: Extract the material attributes targeted by all potential abnormal methods and mark them all as possible attributes, and determine the influence index of the flame retardant performance of each possible attribute;
[0024] Step 5: Sort all possible attributes in descending order according to the values of the flame retardancy performance influence index, and detect the possible attributes of the electrical conduit in the sorted order.
[0025] Compared with the prior art, the present invention has the following beneficial effects:
[0026] 1. The system of the present invention deeply analyzes the temperature characteristics, smoke characteristics, and image characteristics in the flame retardancy detection process of electrical conduits through a flame retardancy performance detection module, a flame retardancy performance anomaly verification module, and a conduit attribute detection module. After determining that the flame retardancy performance of the electrical conduit does not meet the standard, a flame retardancy detection simulation environment for the electrical conduit is constructed by combining simulation technology. By adjusting the material attributes of the electrical conduit, the temperature characteristics, smoke characteristics, and image characteristics of the electrical conduit in the flame retardancy detection process are simulated and analyzed under different conditions of material attribute adjustment, and a similarity analysis is performed on the real flame retardancy detection characteristics and the simulated flame retardancy detection characteristics, and the actual possible material attribute problems of the electrical conduit are comprehensively selected;
[0027] 2. The method of the present invention further analyzes the independence and relevance of possible material attribute problems, and customizes the detection order of the material attributes of the electrical conduit to ensure accurate and efficient discovery and determination of the actual reasons affecting the flame retardancy performance of the electrical conduit. Description of the Drawings
[0028] Figure 1 It is a system module diagram of a flame retardancy performance detection system for electrical conduits based on deep learning;
[0029] Figure 2 It is a flow chart for determining the similarity index of the flame retardancy performance of the conduit attribute modification strategy;
[0030] Figure 3 It is a method flow chart of a flame retardancy performance detection method for electrical conduits based on deep learning. Detailed Embodiments
[0031] Example 1: Refer to Figure 1-2 , a flame retardancy performance detection system for electrical conduits based on deep learning, including a flame retardancy performance detection module, a flame retardancy performance anomaly verification module, and a conduit attribute detection module.
[0032] Flame retardancy performance detection module: Select an electrical conduit to be tested for flame retardancy performance, determine the model of the electrical conduit, determine the corresponding combustion conditions for the electrical conduit (the combustion conditions include combustion temperature, combustion duration, flame height, etc.), conduct combustion on the electrical conduit under the combustion conditions, and after the combustion ends, determine the combustion performance index of the electrical conduit. Set the combustion performance threshold index (the combustion performance threshold index is a preset index used for comparison with the combustion performance index). When the combustion performance index of the electrical conduit is less than the combustion performance threshold index, mark the electrical conduit as a conduit with abnormal flame retardancy performance (when the combustion performance index of the electrical conduit is greater than or equal to the combustion performance threshold index, no corresponding marking is performed).
[0033] The combustion performance index of the electrical conduit is determined based on the following method: Determine the comprehensive combustion characteristic set of the electrical conduit, obtain the combustion performance analysis model, use the comprehensive combustion characteristic set as the input data of the combustion performance analysis model, and the combustion performance analysis model outputs the combustion performance index of the electrical conduit.
[0034] The comprehensive combustion characteristic set of the electrical conduit is determined based on the following method: Obtain the temperature characteristics, smoke characteristics, and image characteristics of the electrical conduit, and integrate the temperature characteristics, smoke characteristics, and image characteristics into a comprehensive combustion characteristic set in the form of a characteristic set.
[0035] The temperature characteristics of the electrical conduit are obtained based on the following method: Real-time collect the temperature data of the electrical conduit during the combustion process. After the combustion ends, preprocess and extract features from the collected temperature data to obtain the temperature characteristics of the electrical conduit.
[0036] The smoke characteristics of the electrical conduit are obtained based on the following method: Real-time collect the smoke data of the electrical conduit during the combustion process. After the combustion ends, preprocess and extract features from the collected smoke data to obtain the smoke characteristics of the electrical conduit.
[0037] The image characteristics of the electrical conduit are obtained based on the following method: Real-time capture a video of the electrical conduit during the combustion process. After the combustion ends, perform frame extraction on the video based on a fixed frame extraction interval to obtain multiple combustion images. Extract features from the multiple combustion images to obtain multiple single-frame features, and perform feature fusion processing on the multiple single-frame features to obtain the image characteristics of the electrical conduit.
[0038] Combustion performance analysis model: A neural network model is constructed. The comprehensive combustion feature sets of m electrical conduit pipes are collected. Using the comprehensive combustion feature sets of the electrical conduit pipes as training data, the neural network model is trained. A combustion performance index is assigned to each training data. The value range of the combustion performance index is (1.0 - 20.0). The closer the value of the combustion performance index is to 20, the more compliant the combustion performance of the electrical conduit pipe is with the standard. The training data is divided into a training set and a validation set at a ratio of 70%:30%. Neural network iterative training is performed on the training set and the validation set. After training is completed, a combustion performance analysis model is obtained.
[0039] Flame retardant performance anomaly verification module: When an electrical conduit pipe is marked as a pipe with abnormal flame retardant performance, obtain the model and combustion conditions of the pipe with abnormal flame retardant performance. Based on the model and combustion conditions, construct a pipe simulation combustion model. The pipe simulation combustion model generates multiple pipe attribute modification strategies. Determine the similarity index of the flame retardant performance of each pipe attribute modification strategy. Set a standard index for the similarity of flame retardant performance (the standard index for the similarity of flame retardant performance is a preset index used to compare with the similarity index of the flame retardant performance). When the similarity index of the flame retardant performance of a pipe attribute modification strategy is greater than the standard index for the similarity of flame retardant performance, mark this pipe attribute modification strategy as a potentially abnormal method (when the similarity index of the flame retardant performance of a pipe attribute modification strategy is less than or equal to the standard index for the similarity of flame retardant performance, no corresponding marking is made).
[0040] Construct a pipe simulation combustion model based on the model and combustion conditions: Select simulation software such as ANSYS Fluent and COMSOL Multiphysics. Determine the geometric structure and material properties of the pipe with abnormal flame retardant performance according to the model. Create a pipe entity in the simulation software based on the geometric structure and material properties. Create a combustion environment in the simulation software based on the combustion conditions. The combustion environment in the simulation software can burn the pipe entity. The constructed pipe simulation combustion model can adjust the material properties of the pipe entity. Each adjustment is only for one material property, and each adjustment generates a pipe attribute modification strategy (for example, pipe attribute modification strategy a is to change the flame retardant additive amount from 15% to 12%, pipe attribute modification strategy b is to change the wall thickness from 1.60 mm to 1.4 mm, pipe attribute modification strategy c is to change the hardness of the electrical conduit pipe from 92 HRC to 80 HRC, pipe attribute modification strategy d is to change the flame retardant additive amount from 15% to 10%, pipe attribute modification strategy e is to adjust the specific heat capacity of the electrical conduit pipe from 2.0 J / (g·K) to 1.8 J / (g·K)). The pipe simulation combustion model can perform simulated combustion tests on the pipe entity based on the pipe attribute modification strategy.
[0041] The similarity index of the flame retardancy performance of the casing attribute modification strategy is determined as follows: The casing simulation combustion model conducts a combustion test on the casing entity based on a casing attribute modification strategy. After the combustion test, the comprehensive combustion characteristic set of the casing entity is obtained, and the comprehensive combustion characteristic set of the casing with abnormal flame retardancy performance is obtained synchronously. The comprehensive combustion characteristic set of the casing entity and the comprehensive combustion characteristic set of the casing with abnormal flame retardancy performance are input into the flame retardancy performance similarity analysis model synchronously, and the flame retardancy performance similarity analysis model outputs the similarity index of the flame retardancy performance of the casing attribute modification strategy.
[0042] Flame retardancy performance similarity analysis model: Construct a deep learning model, collect the comprehensive combustion characteristic sets of multiple groups of casing entities and the comprehensive combustion characteristic sets of casings with abnormal flame retardancy performance, and use the comprehensive combustion characteristic sets of the casing entities and the comprehensive combustion characteristic sets of casings with abnormal flame retardancy performance as training data to train the deep learning model. Assign a similarity index of flame retardancy performance to each training data. The value range of the similarity index of flame retardancy performance is (10.0 - 50.0). The closer the similarity index of flame retardancy performance is to 50.0, the more similar the comprehensive combustion characteristic set of the casing entity is to the comprehensive combustion characteristic set of the casing with abnormal flame retardancy performance. Divide the training data into a training set, a validation set, and a test set, with a ratio of 70%:15%:15%. Train the training set, the validation set, and the test set. After training is completed, the flame retardancy performance similarity analysis model is obtained.
[0043] Casing attribute detection module: Extract the material attributes targeted by all potential abnormal methods and mark them all as possible attributes (for example, if the casing attribute modification strategy a is to change the flame retardant addition amount from 15% to 12%, then the material attribute targeted by the casing attribute modification strategy a is the flame retardant addition amount; if the casing attribute modification strategy b is to change the wall thickness from 1.60 mm to 1.4 mm, then the material attribute targeted by the casing attribute modification strategy b is the wall thickness; if the casing attribute modification strategy c is to change the hardness of the electrical casing from 92 HRC to 80 HRC, then the material attribute targeted by the casing attribute modification strategy c is the hardness). Determine the flame retardancy performance influence index of each possible attribute, sort all possible attributes in descending order according to the value of the flame retardancy performance influence index, and detect the possible attributes of the electrical casing in the sorted order (for example, if the first in the sorting is the flame retardant addition amount, then first detect the flame retardant addition amount of the electrical casing).
[0044] The flame retardancy performance influence index of material properties is determined based on the following method: Select a possible property, mark the remaining possible properties as comparison properties, mark the total number of potential abnormal methods for this possible property as number(potential), obtain the property relationship graph, compare the possible property with all comparison properties pairwise. When there is an association between the possible property and a comparison property in the property relationship graph, increase the material interaction number by one, mark the material interaction number as number(effect), and calculate the flame retardancy performance influence index index(flame) of this material property through where a1 is the first coefficient, a2 is the second coefficient, the value of a1 is 1.17, and the value of a2 is 0.83.
[0045] The property relationship graph includes all material properties of the electrical conduit and shows the relationship between each material property in the form of a knowledge graph. If there is an interaction relationship between two material properties, there is an association between the two material properties in the property relationship graph. For example, there is an interaction relationship between the flame retardant addition amount and the hardness of the electrical conduit (because different types of flame retardants have different effects on hardness. Inorganic flame retardants may increase hardness, while organic flame retardants may decrease hardness, so there is an interaction relationship between them), then there is an association between the flame retardant addition amount and the hardness of the electrical conduit in the property relationship graph. For example, there is no interaction relationship between the specific heat capacity and the hardness of the electrical conduit (although both specific heat capacity and hardness are important physical properties of materials, there is no direct interaction relationship between them. Specific heat capacity mainly reflects the ability of a material to absorb or release heat, while hardness mainly reflects the ability of a material to resist deformation. These two properties are independent in physical nature, so there is no direct interdependence or influence relationship between them), then there is no association between the specific heat capacity and the hardness of the electrical conduit in the property relationship graph.
[0046] Through the flame retardancy performance detection module, the flame retardancy performance abnormality verification module, and the casing property detection module, deeply analyze the temperature characteristics, smoke characteristics, and image characteristics during the flame retardancy detection of the electrical conduit. After determining that the flame retardancy performance of the electrical conduit does not meet the standard, construct a flame retardancy detection simulation environment for the electrical conduit in combination with simulation technology. By adjusting the material properties of the electrical conduit, simulate and analyze the temperature characteristics, smoke characteristics, and image characteristics of the electrical conduit during the flame retardancy detection under different material property adjustments, and conduct a similarity analysis of the real flame retardancy detection characteristics and the simulated flame retardancy detection characteristics, and comprehensively select the possible material property problems of the electrical conduit.
[0047] Example 2: Refer to Figure 3 , a method for detecting the flame retardancy performance of electrical conduits based on deep learning, includes the following steps:
[0048] Step 1: Select an electrical conduit to be tested for flame retardancy performance, test the electrical conduit for flame retardancy performance, and determine whether it is a conduit with abnormal flame retardancy performance.
[0049] Step 2: When the electrical conduit is marked as a conduit with abnormal flame retardancy performance, construct a conduit simulation combustion model.
[0050] Step 3: The conduit simulation combustion model generates multiple conduit attribute modification strategies, and select a potential abnormal method from the conduit attribute modification strategies.
[0051] Step 4: Extract the material attributes targeted by all potential abnormal methods and mark them all as possible attributes, and determine the flame retardancy performance influence index of each possible attribute.
[0052] Step 5: Sort all possible attributes in descending order according to the numerical value of the flame retardancy performance influence index, and test the possible attributes of the electrical conduit in the sorted order.
[0053] The above method conducts further independence and relevance analysis on possible material attribute problems, customizes the material attribute detection order for electrical conduits, and ensures accurate and efficient discovery and determination of the actual reasons affecting the flame retardancy performance of electrical conduits.
[0054] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters in the formulas are set by technicians in this field according to the actual situation.
[0055] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.
[0056] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not imply the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.
[0057] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.
[0058] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.
[0059] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical, or other forms.
[0060] If the above-mentioned functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs and other various media that can store program codes.
[0061] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A flame retardancy performance detection system for electrical conduit based on deep learning, characterized in that, It includes a flame retardancy performance detection module, a flame retardancy performance anomaly verification module, and a sleeve attribute detection module; The flame retardancy performance detection module: Select an electrical conduit to be tested for flame retardancy performance, test the electrical conduit for flame retardancy performance, and determine whether it is a conduit with abnormal flame retardancy performance; The flame retardancy performance anomaly verification module: When the electrical conduit is marked as a conduit with abnormal flame retardancy performance, construct a conduit simulation combustion model. The conduit simulation combustion model generates multiple conduit attribute modification strategies, and select a potential abnormal method from the conduit attribute modification strategies; The sleeve attribute detection module: Extract the material attributes targeted by all potential abnormal methods and mark them all as possible attributes, determine the flame retardancy performance influence index of each possible attribute, sort all possible attributes in descending order according to the value of the flame retardancy performance influence index, and sequentially detect the possible attributes of the electrical conduit according to the sorting order.
2. The flame retardant performance detection system for electrical conduit based on deep learning according to claim 1, characterized in that, Test the electrical conduit for flame retardancy performance and determine whether it is a conduit with abnormal flame retardancy performance: Determine the model of the electrical conduit, determine the combustion conditions corresponding to the electrical conduit, burn the electrical conduit under the combustion conditions. After the combustion ends, determine the combustion performance index of the electrical conduit, set the combustion performance threshold index. When the combustion performance index of the electrical conduit is less than the combustion performance threshold index, mark the electrical conduit as a conduit with abnormal flame retardancy performance.
3. The flame retardancy detection system for electrical conduit based on deep learning according to claim 2, characterized in that, The combustion performance index of the electrical conduit is determined based on the following method: Determine the comprehensive combustion characteristic set of the electrical conduit, obtain a combustion performance analysis model, use the comprehensive combustion characteristic set as the input data of the combustion performance analysis model, and the combustion performance analysis model outputs the combustion performance index of the electrical conduit.
4. The flame retardancy detection system for electrical conduit based on deep learning according to claim 3, wherein, The comprehensive combustion characteristic set of the electrical conduit is determined based on the following method: Obtain the temperature characteristic, smoke characteristic, and image characteristic of the electrical conduit, and integrate the temperature characteristic, smoke characteristic, and image characteristic into a comprehensive combustion characteristic set in the form of a characteristic set.
5. The flame retardancy detection system for electrical conduit based on deep learning according to claim 4, characterized in that, The temperature characteristic of the electrical conduit is obtained based on the following method: Real-time collect the temperature data of the electrical conduit during the combustion process. After the combustion ends, preprocess and extract features from the collected temperature data to extract the temperature characteristic of the electrical conduit; The smoke characteristic of the electrical conduit is obtained based on the following method: Real-time collect the smoke data of the electrical conduit during the combustion process. After the combustion ends, preprocess and extract features from the collected smoke data to extract the smoke characteristic of the electrical conduit; The image characteristic of the electrical conduit is obtained based on the following method: Real-time shoot a video of the electrical conduit during the combustion process. After the combustion ends, perform frame extraction on the video based on a fixed frame extraction interval to obtain multiple combustion images. Extract features from the multiple combustion images to obtain multiple single-frame features, and perform feature fusion processing on the multiple single-frame features to process and obtain the image characteristic of the electrical conduit.
6. The flame retardant performance detection system for electrical conduit based on deep learning according to claim 1, wherein Select a potential abnormal method from the conduit attribute modification strategies: Determine the flame retardancy performance similarity index of each conduit attribute modification strategy, set the flame retardancy performance similarity standard index. When the flame retardancy performance similarity index of the conduit attribute modification strategy is greater than the flame retardancy performance similarity standard index, mark the conduit attribute modification strategy as a potential abnormal method.
7. The detection system for the flame retardant performance of electrical conduit based on deep learning according to claim 6, characterized in that, The similarity index of the flame retardancy performance of the casing attribute modification strategy is determined in the following manner: The casing simulation combustion model conducts a combustion test on the casing entity based on a casing attribute modification strategy. After the combustion test, the comprehensive combustion characteristic set of the casing entity is obtained, and the comprehensive combustion characteristic set of the casing with abnormal flame retardancy performance is synchronously obtained. The comprehensive combustion characteristic set of the casing entity and the comprehensive combustion characteristic set of the casing with abnormal flame retardancy performance are synchronously input into the flame retardancy performance similarity analysis model, and the flame retardancy performance similarity analysis model outputs the similarity index of the flame retardancy performance of the casing attribute modification strategy.
8. The flame retardancy detection system for electrical conduit based on deep learning according to claim 1, wherein The influence index of the flame retardant performance of material properties is determined based on the following method: Select a possible property, mark the remaining possible properties as comparison properties, mark the total number of potential abnormal methods for this possible property as number(potential), obtain the property relationship graph, compare the possible property with all comparison properties pairwise. When there is an association between the possible property and a comparison property in the property relationship graph, increase the material interaction number by one, mark the material interaction number as number(effect), and calculate the influence index of the flame retardant performance of this material property index(flame) through where a1 is the first coefficient and a2 is the second coefficient.
9. A method for detecting the flame retardancy of electrical conduit pipes based on deep learning, which is applied to the system for detecting the flame retardancy of electrical conduit pipes based on deep learning according to any one of claims 1-8, characterized in that, It includes the following steps: Step 1: Select an electrical casing to be tested for flame retardancy performance, test the flame retardancy performance of the electrical casing, and determine whether it is a casing with abnormal flame retardancy performance; Step 2: When the electrical casing is marked as a casing with abnormal flame retardancy performance, construct a casing simulation combustion model; Step 3: The casing simulation combustion model generates multiple casing attribute modification strategies, and select potential abnormal methods from the casing attribute modification strategies; Step 4: Extract the material attributes targeted by all potential abnormal methods and mark them all as possible attributes, and determine the flame retardancy performance influence index of each possible attribute; Step 5: Sort all possible attributes in descending order according to the value of the flame retardancy performance influence index, and test the possible attributes of the electrical casing in the sorted order.