Prediction method and system for tensile test result of waterproof roll
Through big data mining technology and regular tuning models, the tensile test results of waterproof coils are predicted, and the problems of high experimental costs and long production cycles in the existing technology are solved, achieving the effect of rapid screening and cost reduction.
Patent Information
- Application Number
- CN202411523849.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-30
- Publication Date
- 2025-02-18
AI Technical Summary
The prior art is difficult to effectively predict the tensile test results of waterproof coils, resulting in high experimental costs and long production cycles.
By developing a prediction method and system, using big data mining technology, the tensile test results are predicted based on the material, production process and related experimental data of coil materials, and the prediction accuracy reaches more than 95% through regular tuning models.
The preliminary evaluation of the tension performance of the waterproof coil before the experiment was achieved, and the materials that meet the requirements were quickly screened out, which shortened the material screening and testing time, and reduced the experimental cost and R&D cost.
Smart Images
Figure CN119470011A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of building material testing, and in particular to a method and system for predicting waterproof coiled material tensile test results. Background Art
[0002] The purpose of the waterproof membrane tensile test is to simulate the tensile conditions in actual use, measure the key performance indicators of the membrane such as tensile strength, breaking strength and elongation, so as to evaluate its ability to withstand mechanical loads. Each batch of waterproof membrane must pass the tensile test before it can be put into use. However, except for new materials, the experimental data of common waterproof membranes are very rich. If the experimental results can be predicted based on the material, production process and related experimental data of the membrane based on big data mining technology, a lot of experimental costs can be saved, and the cycle from production to release can be shortened, thereby achieving the effect of improving productivity. Summary of the invention
[0003] The present invention provides a method for predicting the tensile test result of a waterproof coiled material, comprising:
[0004] Step 1. Import the material data and production process data of the waterproof membrane to be tested into the prediction system, and configure the simulation test environment in the configuration file of the prediction system;
[0005] Step 2: The prediction system processes the imported data into a format recognizable by the waterproofing membrane tensile test result prediction model based on the format conversion function;
[0006] Step 3. Initialize the hyperparameters of the waterproof membrane tensile test result prediction model according to the configured simulation test environment, input the processed imported data into the model, and output the prediction results;
[0007] Step 4. Explain the output prediction results as a waterproof membrane tensile test result report;
[0008] Step 5. Regularly tune the waterproof membrane tensile test result prediction model based on the actual tensile test records to ensure that its prediction accuracy remains above 95%.
[0009] The method for predicting the tensile test result of waterproofing coiled material as described above, wherein the prediction system processes the imported data into a format recognizable by the tensile test result prediction model of waterproofing coiled material based on the format conversion function, specifically comprises the following sub-steps:
[0010] Create a format conversion function;
[0011] Pass the imported data into the format conversion function and return the processed data;
[0012] Organize the processed data into a set.
[0013] The method for predicting the tensile test result of a waterproofing coiled material as described above, wherein the hyper-parameters of the prediction model for the tensile test result of the waterproofing coiled material are initialized according to the configured simulation test environment, specifically comprising the following sub-steps:
[0014] Extract the environmental parameters of the simulation test in the configuration file;
[0015] According to the extracted environmental parameters, the corresponding model hyperparameter values are queried from the mapping table;
[0016] In the waterproof membrane tensile test result prediction model, the relevant hyper-parameters of the test environment are adjusted to the queried model hyper-parameter values.
[0017] In the above-mentioned method for predicting the tensile test result of waterproof coiled material, the training process of the prediction model of the tensile test result of waterproof coiled material includes the following sub-steps:
[0018] Extract various test data and test results from historical waterproof membrane tensile test records and organize them into training data sets;
[0019] The training data set is divided into training batches according to the test environment, and the waterproof membrane tensile test result prediction model is trained in batches using the divided training data set to determine the optimal solution for each hyperparameter of the model;
[0020] Store different test environment parameters and their corresponding model hyperparameter values in a mapping table.
[0021] The method for predicting the tensile test result of a waterproofing membrane as described above, wherein the outputted prediction result is interpreted as a tensile test result report of the waterproofing membrane, specifically comprises the following sub-steps:
[0022] Create a waterproof membrane tensile test result report template;
[0023] Analyze the value of each test result data item according to the prediction results;
[0024] Fill in the test data of the waterproof membrane and the values of each test result data item into the template of the waterproof membrane tensile test result report.
[0025] The method for predicting the tensile test result of a waterproofing membrane as described above, wherein the prediction model for the tensile test result of the waterproofing membrane is regularly tuned based on the actual tensile test records, specifically comprises the following sub-steps:
[0026] Obtain the actual tensile test records from the last time the model was updated to the current time, and process them as a training data set;
[0027] The training data set is divided into training batches according to the test environment, and the training data sets of different training batches are used to perform incremental training on the waterproof membrane tensile test result prediction model to tune the various hyperparameters in the model;
[0028] Update the mapping table between test environment parameters and model hyperparameter values according to the tuning results.
[0029] The present invention also provides a prediction system for waterproof coiled material tensile test results, comprising: a coiled material data import module, an imported data processing module, a prediction result output module, a prediction result interpretation module, and a prediction model tuning module;
[0030] The coil material data import module is used to import the material data and production process data of the waterproof coil material to be tested into the system, and configure the simulation test environment in the configuration file;
[0031] The coiled material data processing module is used to process the imported data into a format recognizable by the waterproof coiled material tensile test result prediction model based on the format conversion function;
[0032] The prediction result output module is used to initialize the hyperparameters of the waterproof membrane tensile test result prediction model according to the configured simulation test environment, input the processed imported data into the model, and output the prediction results;
[0033] The prediction result interpretation module is used to interpret the output prediction result as a waterproof membrane tensile test result report;
[0034] The prediction model tuning module is used to regularly tune the waterproof membrane tensile test result prediction model based on the actual tensile test records to ensure that its prediction accuracy remains above 95%.
[0035] The beneficial effects achieved by the present invention are as follows: Through the prediction method, engineers and R&D personnel can make a preliminary evaluation of the tensile properties of waterproof membranes before the experiment, so as to quickly screen out materials that may meet the requirements, greatly shorten the time for material screening and testing, and improve decision-making efficiency. At the same time, it reduces the number of unnecessary experiments and avoids repeated testing of a large number of materials, which can not only save experimental equipment and labor costs, but also reduce material waste, thereby reducing overall R&D costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0037] Figure 1This is a flow chart of a method for predicting the tensile test results of a waterproofing membrane provided in Example 1 of the present application;
[0038] Figure 2 It is a schematic diagram of a prediction system for a waterproof membrane tensile test result provided in Example 2 of the present application. DETAILED DESCRIPTION
[0039] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0040] Embodiment 1
[0041] like Figure 1 As shown, the first embodiment of the present application provides a method for predicting the tensile test results of a waterproof coiled material, comprising:
[0042] Step S10: importing the material data and production process data of the waterproof membrane to be tested into the prediction system, and configuring the environment of the simulation test in the configuration file of the prediction system;
[0043] The main factors affecting the tensile test results of waterproof membranes are the material, production process and test environment of the membranes. The material data and production process data are imported into the prediction system, and the experimental environment is adjusted through the configuration file. Before prediction, it is necessary to ensure that these three types of data are true and accurate. Some verification rules can be set to verify whether the imported data is compliant. For example, the main components in the material data must account for a larger proportion than the additives, whether the main components are common materials, whether there are gaps in the production process data, and whether it is a common production process; and based on the verification results, the user is prompted to check whether the relevant data items are correct.
[0044] Step S20: the prediction system processes the imported data into a format recognizable by the waterproof membrane tensile test result prediction model based on the format conversion function;
[0045] The data just imported into the system are two structured data tables, one storing the material data of the waterproofing membrane, and the other storing the production process data of the waterproofing membrane. These data have different formats and measurement units and need to be processed before they can be recognized by the prediction model. Specifically:
[0046] Step S21: create a format conversion function;
[0047] The format conversion function F(C) is used to convert the input data into a format recognizable by the waterproof membrane tensile test result prediction model, which is expressed as:
[0048] Where C is the input parameter set, c k is the kth value in the input parameter set, c max is the maximum value in the input parameter set, c min is the minimum value in the input parameter set, is the mean of all data in the input parameter set, τ is the extreme value resolution coefficient, k ranges from 1 to w, and w is the total number of items in all data in the input parameter set. It should be noted that the value range of τ is [2,4], and the value of this embodiment is 2.5.
[0049] Step S22: passing the imported data into the format conversion function and returning the processed data;
[0050] The imported data is processed into the input parameter set of the format conversion function, and the format conversion function returns the processed data. If the imported data item is numeric, it is directly placed in the input parameter set. If the imported data item is character type, it is replaced by the numeric code corresponding to the character. For example, the numeric code of asphalt is 0.8. It should be noted that the value range of the numeric code is (0,1).
[0051] Step S23: Arrange the processed data into a set.
[0052] Step S30: Initialize the hyper-parameters of the waterproof membrane tensile test result prediction model according to the configured simulation test environment, input the processed imported data into the model, and output the prediction result;
[0053] The waterproofing membrane tensile test result prediction model provides multiple sets of simulation test environments for the prediction of test results. Before the prediction begins, the system will query the corresponding model hyperparameter values from the mapping table according to the simulation test environment parameters set in the configuration file, and adjust the relevant hyperparameters of the test environment in the waterproofing membrane tensile test result prediction model to the queried model hyperparameter values to complete the model initialization, and then input the processed set into the model to output the prediction results. The training process of the waterproofing membrane tensile test result prediction model includes the following sub-steps:
[0054] Step S31: extracting various test data and test results from historical waterproofing membrane tensile test records and arranging them into a training data set;
[0055] The historical waterproofing membrane tensile test records record the detailed data of the test membrane material, production process and test environment. These test data are processed into a unified data format using a format conversion function, and then the various data in the test results are fused into a comprehensive value using a fusion formula. Finally, the test data and test results in each historical test record are taken as an input-output pair and stored in the training data set. The fusion formula of the test results is expressed as: Where Y is the fused value, y d is the value of the dth item of data in the test results, θ d is the fusion ratio of the d-th data, d takes values from 1 to D, and D is the total number of data items included in the test results.
[0056] Step S32: Divide the training data set into training batches according to the test environment, and use the divided training data sets to train the waterproof membrane tensile test result prediction model in batches to determine the optimal solution of each hyperparameter of the model;
[0057] Since there are actually only a limited number of test environments for a manufacturer, training the model in batches according to the test environment can speed up the training of the model and make the model focus more on other test data, thereby achieving customized development of the model.
[0058] Prediction model of tensile test results of waterproof membrane It is expressed as:
[0059] Where X is the input set of the model, is the output weight of the fth layer of the model, b f is the output error of the fth layer of the model, f ranges from 1 to Γ, Γ is the total number of layers of the model, x i is the value of the i-th test data in the input set, μ i is the super parameter of the influence of the i-th test data on the coil tensile test results, i ranges from 1 to m, m is the number of data items in the input set except the test environment data item, e j represents the value of the jth test environment data in the input set, μ j It is the hyperparameter of the influence of the j-th test environment data on the coil tensile test results. j takes values of m+1~n, and n is the total number of all test data in the input set.
[0060] During the training process, the hyperparameters in the model are continuously iterated, and the change curve of the MES loss function is observed. Training is stopped until the curve no longer has an obvious downward trend, and the optimal solution for each hyperparameter of the model is obtained.
[0061] Step S33: Store different test environment parameters and their corresponding model hyperparameter values in a mapping table.
[0062] Step S40: interpreting the output prediction results as a waterproof membrane tensile test result report, which is specifically divided into the following sub-steps:
[0063] Step S41: creating a waterproof membrane tensile test result report template;
[0064] The template contains three filling boxes: original imported data of waterproof membrane, simulated test environment data, and test results.
[0065] Step S42: parsing the values of each test result data item according to the prediction result;
[0066] The fusion formula of the test result is back-calculated according to the fusion ratio used when the prediction result and the fusion test result data item are fused, so as to obtain the value of each test result data item.
[0067] Step S43: Fill the test data of the waterproof membrane and the values of each test result data item into the template of the waterproof membrane tensile test result report;
[0068] Fill in the data in the corresponding filling box of the template to form a complete test result report and return it to the user.
[0069] Step S50: regularly tuning the waterproofing membrane tensile test result prediction model based on the actual tensile test records to ensure that its prediction accuracy remains above 95%;
[0070] With the update of waterproof membrane materials, production processes or test environments, the prediction model of waterproof membrane tensile test results also needs to be adjusted accordingly to ensure that its accuracy meets the application standards. Specifically:
[0071] Step S51: obtaining the actual tensile test records from the last time the model was updated to the current time, and processing them as a training data set;
[0072] Step S52: Divide the training data set into training batches according to the test environment, use the training data sets of different training batches to perform incremental training on the waterproof membrane tensile test result prediction model, and tune each hyperparameter in the model;
[0073] The processing and division of the training set is consistent with the relevant steps in the model step S40. During the training process, each hyperparameter in the model is also tuned based on the minimum loss principle.
[0074] Step S53: Update the mapping table between the test environment parameters and the model hyperparameter values according to the optimization results.
[0075] Embodiment 2
[0076] like Figure 2 As shown, the second embodiment of the present application provides a prediction system for the tensile test results of waterproof coiled materials, including: a coiled material data import module 21, an imported data processing module 22, a prediction result output module 23, a prediction result interpretation module 24, and a prediction model tuning module 25;
[0077] The coiled material data import module 21 is used to import the material data and production process data of the waterproof coiled material to be tested into the system, and configure the environment of the simulation test in the configuration file;
[0078] The main factors affecting the tensile test results of waterproof membranes are the material, production process and test environment of the membranes. The material data and production process data are imported into the prediction system, and the experimental environment is adjusted through the configuration file. Before prediction, it is necessary to ensure that these three types of data are true and accurate. Some verification rules can be set to verify whether the imported data is compliant. For example, the main components in the material data must account for a larger proportion than the additives, whether the main components are common materials, whether there are gaps in the production process data, and whether it is a common production process; and based on the verification results, the user is prompted to check whether the relevant data items are correct.
[0079] The coiled material data processing module 22 is used to process the imported data into a format recognizable by the waterproof coiled material tensile test result prediction model based on the format conversion function; it includes: a data format conversion submodule and a data sorting submodule;
[0080] 1. Data format conversion submodule, used to pass the imported data into the format conversion function and return the processed data;
[0081] The format conversion function F(C) is used to convert the input data into a format recognizable by the waterproof membrane tensile test result prediction model, which is expressed as:
[0082] Where C is the input parameter set, c k is the kth value in the input parameter set, c max is the maximum value in the input parameter set, c min is the minimum value in the input parameter set, is the mean of all data in the input parameter set, τ is the extreme value resolution coefficient, k ranges from 1 to w, and w is the total number of items in all data in the input parameter set. It should be noted that the value range of τ is [2,4], and the value of this embodiment is 2.5.
[0083] The imported data is processed into the input parameter set of the format conversion function, and the format conversion function returns the processed data. If the imported data item is numeric, it is directly placed in the input parameter set. If the imported data item is character type, it is replaced by the numeric code corresponding to the character. For example, the numeric code of asphalt is 0.8. It should be noted that the value range of the numeric code is (0,1).
[0084] 2. Data sorting submodule, used to sort the processed data into a set.
[0085] The prediction result output module 23 is used to initialize the hyperparameters of the waterproofing membrane tensile test result prediction model according to the configured simulation test environment, input the processed imported data into the model, and output the prediction result; it includes: a prediction model training submodule, a prediction model initialization submodule, and a prediction model input and output submodule;
[0086] 1. The prediction model training submodule is used to train the prediction model of the waterproof membrane tensile test results and determine the optimal solution of each hyperparameter in the model;
[0087] The training process of the waterproof membrane tensile test result prediction model includes the following sub-steps:
[0088] I. Extract various test data and test results from historical waterproof membrane tensile test records and organize them into training data sets;
[0089] The historical waterproofing membrane tensile test records record the detailed data of the test membrane material, production process and test environment. These test data are processed into a unified data format using a format conversion function, and then the various data in the test results are fused into a comprehensive value using a fusion formula. Finally, the test data and test results in each historical test record are taken as an input-output pair and stored in the training data set. The fusion formula of the test results is expressed as: Where Y is the fused value, y d is the value of the dth item of data in the test results, θ d is the fusion ratio of the d-th data, d takes values from 1 to D, and D is the total number of data items included in the test results.
[0090] II. Divide the training data set into training batches according to the test environment, use the divided training data set to train the waterproof membrane tensile test result prediction model in batches, and determine the optimal solution for each hyperparameter of the model;
[0091] Since there are actually only a limited number of test environments for a manufacturer, training the model in batches according to the test environment can speed up the training of the model and make the model focus more on other test data, thereby achieving customized development of the model.
[0092] Prediction model of tensile test results of waterproof membrane It is expressed as:
[0093] Where X is the input set of the model, is the output weight of the fth layer of the model, b f is the output error of the fth layer of the model, f ranges from 1 to Γ, Γ is the total number of layers of the model, x i is the value of the i-th test data in the input set, μi is the super parameter of the influence of the i-th test data on the coil tensile test results, i ranges from 1 to m, m is the number of data items in the input set except the test environment data item, e j represents the value of the jth test environment data in the input set, μ j It is the hyperparameter of the influence of the j-th test environment data on the coil tensile test results. j takes values of m+1~n, and n is the total number of all test data in the input set.
[0094] During the training process, the hyperparameters in the model are continuously iterated, and the change curve of the MES loss function is observed. Training is stopped until the curve no longer has an obvious downward trend, and the optimal solution for each hyperparameter of the model is obtained.
[0095] II. Store different test environment parameters and their corresponding model hyperparameter values in a mapping table.
[0096] 2. Prediction model initialization submodule, used to initialize the prediction model according to the configured simulation environment parameters;
[0097] The waterproof membrane tensile test result prediction model provides multiple sets of simulated test environments for the prediction of test results. Before the prediction begins, the system will query the corresponding model hyperparameter values from the mapping table according to the simulated test environment parameters set in the configuration file, and adjust the relevant hyperparameters of the test environment in the waterproof membrane tensile test result prediction model to the queried model hyperparameter values to complete the model initialization.
[0098] 3. The prediction model input and output submodule is used to input the data processed by the coiled material data processing module 22 into the initialized waterproof coiled material tensile test result prediction model and output the prediction result.
[0099] The prediction result interpretation module 24 is used to interpret the output prediction result as a waterproofing membrane tensile test result report; it includes: a report template creation submodule, a test result analysis submodule, and a report filling submodule;
[0100] 1. The report template creation submodule is used to create a report template for the tensile test results of waterproof membranes;
[0101] The template contains three filling boxes: original imported data of waterproof membrane, simulated test environment data, and test results.
[0102] 2. Test result analysis submodule, used to analyze the values of each test result data item according to the prediction results output by the waterproof membrane tensile test result prediction model;
[0103] The fusion formula of the test result is back-calculated according to the fusion ratio used when the prediction result and the fusion test result data item are fused, so as to obtain the value of each test result data item.
[0104] 3. The report filling submodule is used to fill in the test data of the waterproof membrane and the values of each test result data item into the template of the waterproof membrane tensile test result report;
[0105] Fill in the data in the corresponding filling box of the template to form a complete test result report and return it to the user.
[0106] The prediction model tuning module 25 is used to regularly tune the waterproofing membrane tensile test result prediction model based on the actual tensile test records to ensure that its prediction accuracy remains above 95%;
[0107] With the update of waterproof membrane materials, production processes or test environments, the prediction model of waterproof membrane tensile test results also needs to be adjusted accordingly to ensure that its accuracy meets the application standards. Specifically:
[0108] I. Obtain the real tensile test records from the last time the model was updated to the current time, and process them as training data sets;
[0109] II. Divide the training data set into training batches according to the test environment, use the training data sets of different training batches to perform incremental training on the waterproof membrane tensile test result prediction model, and tune the various hyperparameters in the model;
[0110] The processing and division of the training set is consistent with the relevant steps in the model step S40. During the training process, each hyperparameter in the model is also tuned based on the minimum loss principle.
[0111] III. Update the mapping table between test environment parameters and model hyperparameters according to the tuning results.
[0112] Corresponding to the above embodiment, an embodiment of the present invention provides a computer storage medium, including: at least one memory and at least one processor;
[0113] The memory is used to store one or more program instructions;
[0114] The processor is used to run one or more program instructions to execute a method for predicting the tensile test result of a waterproof membrane.
[0115] Corresponding to the above embodiment, an embodiment of the present invention provides a computer-readable storage medium, the computer storage medium contains one or more program instructions, and the one or more program instructions are used by a processor to execute a method for predicting the tensile test results of a waterproof membrane.
[0116] The embodiment disclosed in the present invention provides a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are executed on a computer, the computer executes the above-mentioned method for predicting the tensile test results of a waterproof membrane.
[0117] In the embodiment of the present invention, the processor may be an integrated circuit chip having the ability to process signals. The processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0118] The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present invention can be directly embodied as a hardware decoding processor for execution, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The processor reads the information in the storage medium and completes the steps of the above method in combination with its hardware.
[0119] The storage medium may be a memory, which may be, for example, a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memory.
[0120] Among them, the non-volatile memory can be a read-only memory (ROM), a programmable ROM (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory.
[0121] The volatile memory may be a random access memory (RAM) which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDRSDRAM), enhanced synchronous DRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus DRAM (DRRAM).
[0122] The storage media described in the embodiments of the present invention are intended to include, but are not limited to, these and any other suitable types of memory.
[0123] Those skilled in the art will appreciate that in one or more of the above examples, the functions described in the present invention can be implemented using a combination of hardware and software. When software is used, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. Storage media can be any available media that can be accessed by general or special-purpose computers.
[0124] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for predicting the tensile test results of waterproof coiled materials, characterized in that: include: Step 1. Import the material data and production process data of the waterproof membrane to be tested into the prediction system, and configure the simulation test environment in the configuration file of the prediction system; Step 2: The prediction system processes the imported data into a format recognizable by the waterproofing membrane tensile test result prediction model based on the format conversion function; Step 3. Initialize the hyperparameters of the waterproof membrane tensile test result prediction model according to the configured simulation test environment, input the processed imported data into the model, and output the prediction results; Step 4. Explain the output prediction results as a waterproof membrane tensile test result report; Step 5. Regularly tune the waterproof membrane tensile test result prediction model based on the actual tensile test records to ensure that its prediction accuracy remains above 95%.
2. A method for predicting the tensile test results of waterproof coiled material according to claim 1, characterized in that: The prediction system processes the imported data into a format that can be recognized by the waterproof membrane tensile test result prediction model based on the format conversion function, which specifically includes the following sub-steps: Create a format conversion function; Pass the imported data into the format conversion function and return the processed data; Organize the processed data into a set.
3. The method for predicting the tensile test results of waterproof coiled material according to claim 1, characterized in that: According to the configured simulation test environment, the hyper-parameters of the waterproof membrane tensile test result prediction model are initialized, which specifically includes the following sub-steps: Extract the environmental parameters of the simulation test in the configuration file; According to the extracted environmental parameters, the corresponding model hyperparameter values are queried from the mapping table; In the waterproof membrane tensile test result prediction model, the relevant hyper-parameters of the test environment are adjusted to the queried model hyper-parameter values.
4. The method for predicting the tensile test results of waterproof coiled material according to claim 1, characterized in that: The training process of the waterproof membrane tensile test result prediction model includes the following sub-steps: Extract various test data and test results from historical waterproof membrane tensile test records and organize them into training data sets; The training data set is divided into training batches according to the test environment, and the waterproof membrane tensile test result prediction model is trained in batches using the divided training data set to determine the optimal solution for each hyperparameter of the model; Store different test environment parameters and their corresponding model hyperparameter values in a mapping table.
5. The method for predicting the tensile test results of waterproof coiled material according to claim 1, characterized in that: The output prediction results are interpreted as a waterproof membrane tensile test result report, which includes the following sub-steps: Create a waterproof membrane tensile test result report template; Analyze the value of each test result data item according to the prediction results; Fill in the test data of the waterproof membrane and the values of each test result data item into the template of the waterproof membrane tensile test result report.
6. The method for predicting the tensile test results of waterproof coiled material according to claim 1, characterized in that: Based on the actual tensile test records, the waterproof membrane tensile test result prediction model is regularly tuned, which includes the following sub-steps: Obtain the actual tensile test records from the last time the model was updated to the current time, and process them as a training data set; Divide the training data set into training batches according to the test environment, use the training data sets of different training batches to perform incremental training on the waterproof membrane tensile test result prediction model, and tune the various hyperparameters in the model; Update the mapping table between test environment parameters and model hyperparameter values according to the tuning results.
7. A prediction system for tensile test results of waterproofing membranes, characterized in that: include: Coil data import module, imported data processing module, prediction result output module, prediction result interpretation module, prediction model tuning module; The coil material data import module is used to import the material data and production process data of the waterproof coil material to be tested into the system, and configure the simulation test environment in the configuration file; The coiled material data processing module is used to process the imported data into a format recognizable by the waterproof coiled material tensile test result prediction model based on the format conversion function; The prediction result output module is used to initialize the hyperparameters of the waterproof membrane tensile test result prediction model according to the configured simulation test environment, input the processed imported data into the model, and output the prediction results; The prediction result interpretation module is used to interpret the output prediction result as a waterproof membrane tensile test result report; The prediction model tuning module is used to regularly tune the waterproof membrane tensile test result prediction model based on the actual tensile test records to ensure that its prediction accuracy remains above 95%.
8. A prediction system for tensile test results of waterproofing membrane according to claim 7, characterized in that: The coil data import module includes: a data format conversion submodule and a data sorting submodule; The data format conversion submodule is used to pass the imported data into the format conversion function and return the processed data; The data sorting submodule is used to sort the processed data into a set.
9. A prediction system for tensile test results of waterproofing membrane according to claim 7, characterized in that: The prediction result output module includes a prediction model training submodule, a prediction model initialization submodule, and a prediction model input and output submodule; The prediction model training submodule is used to train the prediction model of the waterproof membrane tensile test results and determine the optimal solution of each hyperparameter in the model; The prediction model initialization submodule is used to initialize the prediction model according to the configured simulation environment parameters; The prediction model input and output submodule is used to input the data processed by the coiled material data processing module into the initialized waterproof coiled material tensile test result prediction model and output the prediction result.
10. A computer storage medium, characterized in that: include: at least one memory and at least one processor; A memory for storing one or more program instructions; A processor is used to run one or more program instructions to execute a method for predicting a waterproof membrane tensile test result as described in claims 1-6.