Intelligent detection method and system for tensile strength of solder bars based on multiple factors
By constructing an intelligent detection method for the tensile strength of solder bars based on multiple factors, the problem of insufficient comprehensive consideration of environmental factors in the aging test of the tensile strength of solder bars is solved, and the accurate prediction and control of the tensile strength of solder bars is realized, ensuring the stability of the welding process.
Patent Information
- Application Number
- CN202410764470.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-14
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-06-14
AI Technical Summary
Existing methods for testing the tensile strength aging of solder bars lack comprehensive consideration of various environmental factors, making it difficult to achieve accurate prediction and control.
By identifying multiple environmental factors affecting the tensile strength of solder bars, defining value ranges, conducting sample value combinations and aging experiments, constructing an intelligent prediction model, and combining the storage environment to accurately predict tensile strength.
It achieves precise control over the tensile strength of solder bars, ensuring the strength of the welding process.
Smart Images

Figure CN118364294B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of solder bar performance management, and in particular to a multi-factor based intelligent detection method and system for the tensile strength of solder bars. Background Art
[0002] Tin soldering rods are welding rods used for soldering. Tin soldering can be used for sealed metal welding without requiring high temperature and high pressure. Solder is a commonly used welding material, and its main component is tin. The tensile strength of the solder bar refers to the maximum force it can withstand under tension. The tensile strength of the solder bar directly affects the strength of the welding process, so it is crucial to control the tensile strength of the solder bar.
[0003] Since the aging of the tensile strength of the solder bar is the result of the interaction between various environmental factors, the existing aging test method for the tensile strength of the solder bar lacks comprehensive consideration of various environmental factors, making it difficult to achieve accurate prediction and control of the tensile strength of the solder bar. Summary of the invention
[0004] In order to solve the above technical problems, a method and system for intelligent detection of the tensile strength of solder bars based on multiple factors are provided. This technical solution solves the problem that the existing aging test method for the tensile strength of solder bars lacks comprehensive consideration of various environmental factors, making it difficult to accurately predict and control the tensile strength of solder bars.
[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:
[0006] An intelligent detection method for tensile strength of solder bars based on multiple factors, comprising:
[0007] Determine at least one environmental factor that affects the decrease in the tensile strength of the solder bar, and record it as an influencing factor;
[0008] Determine the value range of each influencing factor based on the actual environment;
[0009] Within the value range of the influencing factor, several values are taken according to the set parameter gradient to form several sample values corresponding to the influencing factor;
[0010] Take one value from a number of sample values corresponding to each influencing factor and combine them into a sample parameter set;
[0011] Traverse all sample values of all influencing factors until no new sample parameter set can be formed;
[0012] The relationship between the decrease in tensile strength of solder bars and time under each set of sample parameter sets was tested in the test box;
[0013] Based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets, an intelligent prediction model for the decrease in the tensile strength of the solder bar is constructed;
[0014] The storage environment of the solder bar is obtained, and the tensile strength of the solder bar is determined based on the storage environment of the solder bar and the intelligent prediction model of the decrease value of the tensile strength of the solder bar.
[0015] Preferably, within the value range of the influencing factor, a plurality of values are taken according to a set parameter gradient, and the plurality of sample values corresponding to the influencing factor also include:
[0016] A number of samples corresponding to each influencing factor are divided into model sample values and model verification values, wherein there is at least one model verification value between any two adjacent model sample values.
[0017] Preferably, the relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets tested in the test box specifically includes:
[0018] Set at least two practice test durations;
[0019] Based on the set number of simulation test durations, for each set of sample parameter sets, a corresponding number of solder bar test samples are prepared;
[0020] Set the parameters in the test box to the parameters corresponding to the sample parameter set, and perform an environmental aging test on each solder bar test sample according to the simulated test duration;
[0021] Perform a tensile strength reduction test on each solder bar test sample after the environmental aging test, and obtain the tensile strength reduction value of the solder bar under each simulated test duration corresponding to the sample parameter set;
[0022] In a two-dimensional rectangular coordinate system, with the simulation test time as the horizontal axis and the decrease in the tensile strength of the solder bar as the vertical axis, the coordinate points of the decrease in the tensile strength of the solder bar - simulation test time corresponding to the sample parameter set are marked.
[0023] Preferably, the constructing of the intelligent prediction model for the decrease in tensile strength of solder bars based on the change relationship between the decrease in tensile strength of solder bars and time corresponding to all sample parameter sets specifically includes:
[0024] Determine a sample parameter set in which all elements in the sample parameter set are model sample values, record it as the model training sample parameter set, and record the non-model training sample parameter set in the sample parameter set as the model verification sample parameter set;
[0025] Based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set-simulation test duration coordinate point, at least one known function is called for fitting to obtain at least one fitting function corresponding to the model training sample parameter set;
[0026] Two sets of model training sample parameters between corresponding elements of which no other model sample values exist are used as a set of fitting model training sample parameter sets;
[0027] Determine all model validation sample parameter sets between each fitting model training sample parameter set group;
[0028] Combining all the fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs to obtain a plurality of fitting function groups;
[0029] Calculate the loss values of all fitting function groups for the model validation sample parameter set respectively;
[0030] The fitting function group with the smallest loss value is selected, and the two fitting functions corresponding to the fitting function group are used as the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group;
[0031] Based on the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to all fitting model training sample parameter set groups, an intelligent prediction model for the tensile strength decrease value of the solder bar is constructed.
[0032] Preferably, the separately calculating the loss values of all fitting function groups for the model verification sample parameter set specifically includes:
[0033] Based on the strength fitting formula, the fitting value of the tensile strength reduction value for each fitting function group under each simulation test duration of the model verification sample parameter set between the fitting model training sample parameter set groups is calculated;
[0034] Retrieve the tensile strength reduction value of the solder bar under each simulation test duration corresponding to each model verification sample parameter set, and make a difference with the tensile strength reduction value fitting value of the model verification sample parameter set and take the absolute value to obtain the loss value corresponding to the model verification sample parameter set;
[0035] Add the loss values corresponding to all model validation sample parameter sets to obtain the loss value corresponding to the fitting function group;
[0036] Wherein, the intensity fitting formula is specifically: ,
[0037] In the formula, is a set of sample parameters for model validation. is the i-th element in the model validation sample parameter set, for For the fitting value of the tensile strength drop at time t of the fitting function group, are the two fitting functions corresponding to the fitting function group, for The i-th element in the corresponding model training sample parameter set, for The i-th element in the corresponding model training sample parameter set.
[0038] Preferably, the construction of the intelligent prediction model for the decrease in tensile strength of solder bars based on the decrease in tensile strength of solder bars-time variation function of the model training sample parameter sets corresponding to all the fitting model training sample parameter set groups specifically includes:
[0039] Determine an input environmental parameter set, and based on the input environmental parameter set, determine a model sample value corresponding to each influencing factor among the model sample values corresponding to each influencing factor, which is greater than the input environmental parameter and closest to the input environmental parameter, and record it as a first model sample value set;
[0040] Determine the model sample value corresponding to each influencing factor that is smaller than the input environmental parameter and closest to the input environmental parameter among the model sample values corresponding to each influencing factor, and record them as a second model sample value set;
[0041] Determine the solder bar tensile strength drop value-time variation function of the fitting model training sample parameter set group composed of the first model sample value set and the second model sample value set respectively;
[0042] The intelligent prediction model for the decrease in tensile strength of solder bars is specifically: ,
[0043] is the set of input environment parameters, The tensile strength drop value of the solder bar corresponding to the input environmental parameters - time variation function, is the solder bar tensile strength drop value-time variation function corresponding to the first model sample value set, is the solder bar tensile strength drop value-time variation function corresponding to the second model sample value set, is the i-th element in the first model sample value set, is the i-th element in the second model sample value set, is the i-th element in the input environment parameter set.
[0044] Preferably, the obtaining of the storage environment of the solder bar and determining the tensile strength of the solder bar based on the storage environment of the solder bar combined with the intelligent prediction model of the decrease value of the tensile strength of the solder bar specifically includes:
[0045] Based on the storage environment of the solder bar, the storage period of the solder bar is divided into several statistical periods according to the environmental changes, wherein the storage environment in each statistical period does not change, and the storage environment in two adjacent statistical periods changes;
[0046] Determine the solder bar tensile strength decrease value-time variation function corresponding to the storage environment corresponding to each statistical period;
[0047] Based on the duration of the first statistical period, the duration of the first statistical period is substituted into the solder bar tensile strength decrease value-time change function corresponding to the first statistical period to obtain the solder bar tensile strength decrease value of the first statistical period;
[0048] Calculate the cumulative decrease in the tensile strength of the solder bar according to the strength decrease cumulative formula, and subtract the cumulative decrease in the tensile strength of the solder bar from the tensile strength of the solder bar to obtain the tensile strength of the solder bar;
[0049] The strength reduction accumulation formula is specifically: ,
[0050] In the formula, is the cumulative decrease in the tensile strength of the solder bar in the first j statistical periods, is the initial equivalent time of the jth statistical period, is the length of the jth statistical period, is the cumulative decrease in the tensile strength of the solder bar. is the total number of statistical periods, It is the tensile strength decrease value of the solder bar corresponding to the jth statistical period - time variation function.
[0051] Furthermore, a multi-factor based intelligent detection system for the tensile strength of solder bars is proposed, which is used to implement the multi-factor based intelligent detection method for the tensile strength of solder bars as described above, including:
[0052] A sample value determination module, the sample value determination module is used to determine at least one environmental factor that affects the decrease in the tensile strength of the solder bar, recorded as an influencing factor, and based on the actual environment, determine the value range of each influencing factor, within the value range of the influencing factor, take a number of values according to the set parameter gradient, and form a number of sample values corresponding to the influencing factor, take a value from the several sample values corresponding to each influencing factor, and combine them into a sample parameter set, and traverse all sample values of all influencing factors until a new sample parameter set cannot be combined;
[0053] A testing device, the testing device is electrically connected to the sample value determination module, and the testing device is used to test the relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets in the test box;
[0054] A model analysis module, the model analysis module is electrically connected to the test device, and the model analysis module is used to build an intelligent prediction model for the decrease in the tensile strength of the solder bar based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets;
[0055] The tensile strength analysis module is electrically connected to the model analysis module, and is used to obtain the storage environment of the solder bar, and determine the tensile strength of the solder bar based on the storage environment of the solder bar and the intelligent prediction model of the tensile strength decrease value of the solder bar.
[0056] Preferably, the model analysis module includes:
[0057] A preliminary fitting unit, the preliminary fitting unit is used to determine a sample parameter set in which all elements are model sample values, recorded as a model training sample parameter set, record a non-model training sample parameter set in the sample parameter set as a model verification sample parameter set, and based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set-simulation test duration coordinate point, call at least one known function for fitting, and obtain at least one fitting function corresponding to the model training sample parameter set;
[0058] A fitting verification unit, the fitting verification unit is used to use two model training sample parameter sets between which no other model sample values exist as fitting model training sample parameter set groups, determine all model verification sample parameter sets between each fitting model training sample parameter set group, combine all fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs, obtain a plurality of fitting function groups, respectively calculate the loss values of fitting all fitting function groups for the model verification sample parameter set, screen out the fitting function group with the smallest loss value, and use the two fitting functions corresponding to the fitting function group as the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group;
[0059] A model building unit is used to build an intelligent prediction model for the decrease in the tensile strength of a solder bar based on the decrease in the tensile strength of the solder bar-time variation function of the model training sample parameter set corresponding to all the fitting model training sample parameter set groups.
[0060] Preferably, the tensile strength analysis module comprises:
[0061] An initial calculation unit is used to calculate the Calculate the equivalent time length calculation unit for the decrease in tensile strength of the first solder bar, the equivalent time length calculation unit is used to calculate the equivalent time length calculation unit according to the formula Calculate the initial equivalent moment of the jth statistical period;
[0062] Iterative calculation unit, the iterative calculation unit is used to calculate the Calculate the cumulative decrease in tensile strength of the first j solder bars;
[0063] An accumulation unit, the accumulation unit is used to use the accumulated values of decrease in tensile strength of the solder bars in all statistical time periods calculated by the iterative calculation unit as the accumulated values of decrease in tensile strength of the solder bars;
[0064] The difference obtaining unit is used to obtain the tensile strength of the solder bar by subtracting the accumulated decrease in the tensile strength of the solder bar from the tensile strength of the solder bar.
[0065] Compared with the prior art, the present invention has the following beneficial effects:
[0066] The present invention proposes an intelligent detection scheme for the tensile strength of solder bars based on multiple factors. In the factory stage of each specification of solder bars, a parameter division and fitting method is used to construct an intelligent prediction model for the decrease in the tensile strength of the solder bars. A number of sample values are divided within the value range of the influencing factors, and an aging experiment of the tensile strength of the solder bars is performed on the sample values to obtain the tensile strength change standard of the solder bars corresponding to the sample values. When analyzing the tensile strength of the solder bars, the storage period of the solder bars is divided into a number of statistical periods, and the aging decrease in the tensile strength of the solder bars in each statistical period is iteratively accumulated. In this way, the tensile strength of the solder bars can be effectively controlled, thereby ensuring the strength of the welding process. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] Figure 1 A flow chart of the intelligent detection method for the tensile strength of solder bars based on multiple factors proposed by the present invention;
[0068] Figure 2 It is a flow chart of the method for testing the relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets in the test box in the present invention;
[0069] Figure 3 It is a flow chart of the overall method for constructing an intelligent prediction model for the decrease in tensile strength of solder bars in the present invention;
[0070] Figure 4A flow chart of a method for respectively calculating the loss values of all fitting function groups for a model verification sample parameter set in the present invention;
[0071] Figure 5 A flow chart of a method for constructing an intelligent prediction model for the decrease in tensile strength of solder bars based on the decrease in tensile strength of solder bars-time variation function of the model training sample parameter sets corresponding to all fitting model training sample parameter set groups in the present invention;
[0072] Figure 6 The present invention is a flowchart of a method for determining the tensile strength of a solder bar using the intelligent prediction model. DETAILED DESCRIPTION
[0073] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.
[0074] Reference Figure 1 As shown, a multi-factor based intelligent detection method for the tensile strength of solder bars includes:
[0075] Determine at least one environmental factor that affects the decrease in the tensile strength of the solder bar, and record it as an influencing factor;
[0076] Determine the value range of each influencing factor based on the actual environment;
[0077] Within the value range of the influencing factor, several values are taken according to the set parameter gradient to form several sample values corresponding to the influencing factor;
[0078] Take one value from a number of sample values corresponding to each influencing factor and combine them into a sample parameter set;
[0079] Traverse all sample values of all influencing factors until no new sample parameter set can be formed;
[0080] The relationship between the decrease in tensile strength of solder bars and time under each set of sample parameter sets was tested in the test box;
[0081] Based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets, an intelligent prediction model for the decrease in the tensile strength of the solder bar is constructed;
[0082] The storage environment of the solder bar is obtained, and the tensile strength of the solder bar is determined based on the storage environment of the solder bar and the intelligent prediction model of the decrease value of the tensile strength of the solder bar.
[0083] Among them, within the value range of the influencing factor, several values are taken according to the set parameter gradient, and the several sample values corresponding to the influencing factor also include:
[0084] A number of samples corresponding to each influencing factor are divided into model sample values and model verification values, wherein there is at least one model verification value between any two adjacent model sample values.
[0085] In this scheme, at the stage of each specification of solder bar leaving the factory, a parameter division and fitting method is used to construct an intelligent prediction model for the decrease in the tensile strength of the solder bar. A number of sample values are divided within the value range of the influencing factors, and the sample values are subjected to an aging experiment on the tensile strength of the solder bar to obtain the tensile strength change standard of the solder bar corresponding to the sample value. When analyzing the tensile strength of the solder bar, the storage period of the solder bar is divided into a number of statistical periods, and the aging decrease in the tensile strength of the solder bar in each statistical period is iteratively accumulated. In this way, the tensile strength of the solder bar can be effectively controlled.
[0086] Reference Figure 2 As shown in the figure, the relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets tested in the test box is as follows:
[0087] Set at least two practice test durations;
[0088] Based on the set number of simulation test durations, for each set of sample parameter sets, a corresponding number of solder bar test samples are prepared;
[0089] Set the parameters in the test box to the parameters corresponding to the sample parameter set, and perform an environmental aging test on each solder bar test sample according to the simulated test duration;
[0090] Perform a tensile strength reduction test on each solder bar test sample after the environmental aging test, and obtain the tensile strength reduction value of the solder bar under each simulated test duration corresponding to the sample parameter set;
[0091] In a two-dimensional rectangular coordinate system, with the simulation test time as the horizontal axis and the decrease in the tensile strength of the solder bar as the vertical axis, the coordinate points of the decrease in the tensile strength of the solder bar - simulation test time corresponding to the sample parameter set are marked.
[0092] Reference Figure 3 As shown in the figure, based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets, the intelligent prediction model for the decrease in the tensile strength of the solder bar is constructed, which specifically includes:
[0093] Determine a sample parameter set in which all elements in the sample parameter set are model sample values, record it as the model training sample parameter set, and record the non-model training sample parameter set in the sample parameter set as the model verification sample parameter set;
[0094] Based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set-simulation test duration coordinate point, at least one known function is called for fitting to obtain at least one fitting function corresponding to the model training sample parameter set;
[0095] Two sets of model training sample parameters between which no other model sample values exist are used as a set of fitting model training sample parameter sets;
[0096] Determine all model validation sample parameter sets between each fitting model training sample parameter set group;
[0097] Combining all the fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs to obtain a plurality of fitting function groups;
[0098] Calculate the loss values of all fitting function groups for the model validation sample parameter set respectively;
[0099] The fitting function group with the smallest loss value is selected, and the two fitting functions corresponding to the fitting function group are used as the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group;
[0100] Based on the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to all fitting model training sample parameter set groups, an intelligent prediction model for the tensile strength decrease value of the solder bar is constructed.
[0101] Reference Figure 4 As shown, the loss values of fitting all fitting function groups for the model validation sample parameter set are calculated separately, including:
[0102] Based on the strength fitting formula, the fitting value of the tensile strength reduction value for each fitting function group under each simulation test duration of the model verification sample parameter set between the fitting model training sample parameter set groups is calculated;
[0103] Retrieve the tensile strength reduction value of the solder bar under each simulation test duration corresponding to each model verification sample parameter set, and make a difference with the tensile strength reduction value fitting value of the model verification sample parameter set and take the absolute value to obtain the loss value corresponding to the model verification sample parameter set;
[0104] Add the loss values corresponding to all model validation sample parameter sets to obtain the loss value corresponding to the fitting function group;
[0105] Among them, the intensity fitting formula is specifically: ,
[0106] In the formula, is a set of sample parameters for model validation. is the i-th element in the model validation sample parameter set, for For the fitting value of the tensile strength drop at time t of the fitting function group, are the two fitting functions corresponding to the fitting function group, for The i-th element in the corresponding model training sample parameter set, for The i-th element in the corresponding model training sample parameter set.
[0107] Reference Figure 5 As shown, based on the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to all fitting model training sample parameter set groups, constructing the solder bar tensile strength decrease value intelligent prediction model specifically includes:
[0108] Determine an input environmental parameter set, and based on the input environmental parameter set, determine a model sample value corresponding to each influencing factor among the model sample values corresponding to each influencing factor, which is greater than the input environmental parameter and closest to the input environmental parameter, and record it as a first model sample value set;
[0109] Determine the model sample value corresponding to each influencing factor that is smaller than the input environmental parameter and closest to the input environmental parameter among the model sample values corresponding to each influencing factor, and record them as a second model sample value set;
[0110] Determine the solder bar tensile strength drop value-time variation function of the fitting model training sample parameter set group composed of the first model sample value set and the second model sample value set respectively;
[0111] The intelligent prediction model for the decrease in tensile strength of solder bars is as follows: ,
[0112] is the set of input environment parameters, The tensile strength drop value of the solder bar corresponding to the input environmental parameters - time variation function, is the solder bar tensile strength drop value-time variation function corresponding to the first model sample value set, is the solder bar tensile strength drop value-time variation function corresponding to the second model sample value set, is the i-th element in the first model sample value set, is the i-th element in the second model sample value set, is the i-th element in the input environment parameter set.
[0113] In this scheme, the model training sample parameter set group is used as the data anchor point of the intelligent prediction model of the tensile strength reduction value of the solder bar. The tensile strength reduction change of the solder bar between the model training sample parameter set groups is approximated as a linear change, and the function loss of each fitting function group when performing strength prediction is verified based on the measured values of the tensile strength reduction value of the solder bar of several model verification sample parameter sets between the model training sample parameter set groups. The fitting function group with the smallest function loss is screened out as the best fitting function of the model training sample parameter set group;
[0114] Among them, it can be understood that the same model training sample parameter set may have different solder bar tensile strength drop value-time variation functions in different model training sample parameter set groups. This is caused by the nonlinear correlation between the solder bar tensile strength drop value and each influencing parameter. Based on this, the actual solder bar tensile strength drop value-time variation function of the model training sample parameter set is calculated using the average value of multiple solder bar tensile strength drop value-time variation functions corresponding to the model training sample parameter set.
[0115] Reference Figure 6 As shown, obtaining the storage environment of the solder bar, and determining the tensile strength of the solder bar based on the storage environment of the solder bar combined with the intelligent prediction model of the decrease value of the tensile strength of the solder bar specifically includes:
[0116] Based on the storage environment of the solder bar, the storage period of the solder bar is divided into several statistical periods according to the environmental changes, wherein the storage environment in each statistical period does not change, and the storage environment in two adjacent statistical periods changes;
[0117] Determine the solder bar tensile strength decrease value-time variation function corresponding to the storage environment corresponding to each statistical period;
[0118] Based on the duration of the first statistical period, the duration of the first statistical period is substituted into the solder bar tensile strength decrease value-time change function corresponding to the first statistical period to obtain the solder bar tensile strength decrease value of the first statistical period;
[0119] Calculate the cumulative decrease in the tensile strength of the solder bar according to the strength decrease cumulative formula, and subtract the cumulative decrease in the tensile strength of the solder bar from the tensile strength of the solder bar to obtain the tensile strength of the solder bar;
[0120] The specific formula for the cumulative strength reduction is: ,
[0121] In the formula, is the cumulative decrease in the tensile strength of the solder bar in the first j statistical periods, is the initial equivalent time of the jth statistical period, is the length of the jth statistical period, is the cumulative decrease in the tensile strength of the solder bar. is the total number of statistical periods, It is the tensile strength decrease value of the solder bar corresponding to the jth statistical period - time variation function.
[0122] Among them, since the decrease value of the tensile strength of the solder bar is usually nonlinear with time, in this scheme, when calculating the cumulative value of the decrease in the tensile strength of the solder bar, an initial equivalent moment is introduced. The initial equivalent moment is the decrease value of the tensile strength of the solder bar in the current period - time variation function, and the time required to reach the cumulative value of the decrease in the tensile strength of the solder bar in the previous statistical period, so as to realize the accumulation of the tensile strength decrease value of the solder bar - time variation function corresponding to multiple nonlinear statistical time periods, and realize the accurate fitting of the tensile strength of the solder bar.
[0123] Further, based on the same inventive concept as the above-mentioned intelligent detection method for tensile strength of solder bars based on multiple factors, this solution proposes an intelligent detection system for tensile strength of solder bars based on multiple factors, including:
[0124] The sample value determination module is used to determine at least one environmental factor that affects the decrease in the tensile strength of the solder bar, recorded as an influencing factor, and determine the value range of each influencing factor based on the actual environment. Within the value range of the influencing factor, a number of values are taken according to the set parameter gradient to form a number of sample values corresponding to the influencing factor. A value is taken from the number of sample values corresponding to each influencing factor, and they are combined into a sample parameter set. All sample values of all influencing factors are traversed until a new sample parameter set cannot be combined;
[0125] A testing device, the testing device is electrically connected to the sample value determination module, and the testing device is used to test the relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets in the test box;
[0126] A model analysis module, the model analysis module is electrically connected to the test equipment, and the model analysis module is used to build an intelligent prediction model for the decrease in the tensile strength of the solder bar based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets;
[0127] The tensile strength analysis module is electrically connected to the model analysis module. The tensile strength analysis module is used to obtain the storage environment of the solder bar, and determine the tensile strength of the solder bar based on the storage environment of the solder bar and the intelligent prediction model of the tensile strength decrease value of the solder bar.
[0128] The model analysis module includes:
[0129] A preliminary fitting unit, the preliminary fitting unit is used to determine a sample parameter set in which all elements are model sample values, recorded as a model training sample parameter set, and a non-model training sample parameter set in the sample parameter set is recorded as a model verification sample parameter set, and based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set-simulation test duration coordinate point, at least one known function is called for fitting to obtain at least one fitting function corresponding to the model training sample parameter set;
[0130] A fitting verification unit, the fitting verification unit is used to use two model training sample parameter sets between which no other model sample values exist as fitting model training sample parameter set groups, determine all model verification sample parameter sets between each fitting model training sample parameter set group, combine all fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs, obtain several fitting function groups, respectively calculate the loss values of fitting all fitting function groups for the model verification sample parameter set, screen out the fitting function group with the smallest loss value, and use the two fitting functions corresponding to the fitting function group as the solder bar tensile strength reduction value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group;
[0131] The model building unit is used to build an intelligent prediction model for the decrease in the tensile strength of the solder bar based on the decrease in the tensile strength of the solder bar-time variation function of the model training sample parameter set corresponding to all the fitting model training sample parameter set groups.
[0132] The tensile strength analysis module includes:
[0133] Initial calculation unit, the initial calculation unit is used to calculate the Calculate the equivalent time calculation unit for the decrease in tensile strength of the first solder bar. The equivalent time calculation unit is used to calculate the equivalent time calculation unit according to the formula Calculate the initial equivalent moment of the jth statistical period;
[0134] Iterative calculation unit, iterative calculation unit is used to calculate the Calculate the cumulative decrease in tensile strength of the first j solder bars;
[0135] The accumulation unit is used to calculate the cumulative value of the decrease in the tensile strength of the solder bar in all statistical time periods based on the iterative calculation unit as the cumulative value of the decrease in the tensile strength of the solder bar;
[0136] The difference unit is used to subtract the accumulated decrease in the tensile strength of the solder bar from the tensile strength of the solder bar to obtain the tensile strength of the solder bar.
[0137] The process of using the above system is as follows:
[0138] Step 1: The sample value determination module determines at least one environmental factor that affects the decrease in the tensile strength of the solder bar, recorded as an influencing factor, and determines the value range of each influencing factor based on the actual environment. Within the value range of the influencing factor, a number of values are taken according to the set parameter gradient to form a number of sample values corresponding to the influencing factor. A value is taken from the number of sample values corresponding to each influencing factor, and they are combined into a sample parameter set. All sample values of all influencing factors are traversed until a new sample parameter set cannot be combined;
[0139] Step 2: The test equipment tests the relationship between the decrease in the tensile strength of the solder bar and the time under each set of sample parameter sets in the test box;
[0140] Step 3: The model analysis module calls the preliminary fitting unit to determine a sample parameter set in which all elements of the sample parameter set are model sample values, which is recorded as the model training sample parameter set, and the non-model training sample parameter set in the sample parameter set is recorded as the model verification sample parameter set, and based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set - the simulated test duration coordinate point, at least one known function is called for fitting to obtain at least one fitting function corresponding to the model training sample parameter set;
[0141] Step 4: The model analysis module calls the fitting verification unit to use two model training sample parameter sets between which no other model sample values exist as the fitting model training sample parameter set group, determine all model verification sample parameter sets between each fitting model training sample parameter set group, and combine all fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs to obtain several fitting function groups, respectively calculate the loss values of fitting all fitting function groups for the model verification sample parameter set, screen out the fitting function group with the smallest loss value, and use the two fitting functions corresponding to the fitting function group as the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group;
[0142] Step 5: The model building unit builds an intelligent prediction model for the decrease in tensile strength of the solder bar based on the decrease in tensile strength of the solder bar-time variation function of the model training sample parameter set corresponding to all the fitting model training sample parameter set groups;
[0143] Step 6: The tensile strength analysis module calls the initial calculation to use the formula Calculate the decrease in tensile strength of the first solder bar;
[0144] Step 7: The tensile strength analysis module cyclically calls the equivalent time calculation unit according to the formula Calculate the initial equivalent time and iterative calculation unit of the jth statistical period according to the formula Calculate the cumulative decrease in tensile strength of the first j solder bars until the last statistical period;
[0145] Step 8: The accumulation unit calculates the cumulative value of the decrease in the tensile strength of the solder bar in all statistical time periods based on the calculation by the iterative calculation unit, as the cumulative value of the decrease in the tensile strength of the solder bar;
[0146] Step 9: The difference unit subtracts the tensile strength of the solder bar from the accumulated decrease in the tensile strength of the solder bar to obtain the tensile strength of the solder bar.
[0147] In summary, the advantages of the present invention are that the tensile strength of the solder bar can be effectively controlled, thereby ensuring the strength of the welding process.
[0148] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.
Claims
1. An intelligent detection method for the tensile strength of solder bars based on multiple factors. It is characterized in that include: Determine at least one environmental factor that affects the decrease in the tensile strength of the solder bar, and record it as an influencing factor; Determine the value range of each influencing factor based on the actual environment; Within the value range of the influencing factor, several values are taken according to the set parameter gradient to form several sample values corresponding to the influencing factor; Take one value from a number of sample values corresponding to each influencing factor and combine them into a sample parameter set; Traverse all sample values of all influencing factors until no new sample parameter set can be formed; The relationship between the decrease in tensile strength of solder bars and time under each set of sample parameter sets was tested in the test box; Based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets, an intelligent prediction model for the decrease in the tensile strength of the solder bar is constructed; Obtain the storage environment of the solder bar, and determine the tensile strength of the solder bar based on the storage environment of the solder bar and the intelligent prediction model of the decrease value of the tensile strength of the solder bar; The relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets tested in the test box specifically includes: Set at least two practice test durations; Based on the set number of simulation test durations, for each set of sample parameter sets, a corresponding number of solder bar test samples are prepared; Set the parameters in the test box to the parameters corresponding to the sample parameter set, and perform an environmental aging test on each solder bar test sample according to the simulated test duration; Perform a tensile strength reduction test on each solder bar test sample after the environmental aging test, and obtain the tensile strength reduction value of the solder bar under each simulated test duration corresponding to the sample parameter set; In a two-dimensional rectangular coordinate system, with the simulation test time as the horizontal axis and the decrease in the tensile strength of the solder bar as the vertical axis, the coordinate points of the decrease in the tensile strength of the solder bar corresponding to the sample parameter set - simulation test time are marked; The construction of the intelligent prediction model for the decrease in the tensile strength of the solder bar based on the change relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets specifically includes: Determine a sample parameter set in which all elements in the sample parameter set are model sample values, record it as the model training sample parameter set, and record the non-model training sample parameter set in the sample parameter set as the model verification sample parameter set; Based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set-simulation test duration coordinate point, at least one known function is called for fitting to obtain at least one fitting function corresponding to the model training sample parameter set; Two sets of model training sample parameters between which no other model sample values exist are used as a set of fitting model training sample parameter sets; Determine all model validation sample parameter sets between each fitting model training sample parameter set group; Combining all the fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs to obtain a plurality of fitting function groups; Calculate the loss values of all fitting function groups for the model validation sample parameter set respectively; The fitting function group with the smallest loss value is selected, and the two fitting functions corresponding to the fitting function group are used as the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group; Based on the solder bar tensile strength reduction value-time variation function of the model training sample parameter set corresponding to all the fitting model training sample parameter set groups, an intelligent prediction model for the solder bar tensile strength reduction value is constructed; The loss values of fitting all fitting function groups for the model verification sample parameter set are calculated separately and specifically include: Based on the strength fitting formula, the fitting value of the tensile strength reduction value for each fitting function group under each simulation test duration of the model verification sample parameter set between the fitting model training sample parameter set groups is calculated; Retrieve the tensile strength reduction value of the solder bar under each simulation test duration corresponding to each model verification sample parameter set, and make a difference with the tensile strength reduction value fitting value of the model verification sample parameter set and take the absolute value to obtain the loss value corresponding to the model verification sample parameter set; Add the loss values corresponding to all model validation sample parameter sets to obtain the loss value corresponding to the fitting function group; Wherein, the intensity fitting formula is specifically: , In the formula, is a set of sample parameters for model validation. is the i-th element in the model validation sample parameter set, for For the fitting value of the tensile strength drop at time t of the fitting function group, are the two fitting functions corresponding to the fitting function group, for The i-th element in the corresponding model training sample parameter set, for The i-th element in the corresponding model training sample parameter set; The construction of the intelligent prediction model for the decrease in tensile strength of solder bars based on the decrease in tensile strength of solder bars-time variation function of the model training sample parameter set corresponding to all the fitting model training sample parameter set groups specifically includes: Determine an input environmental parameter set, and based on the input environmental parameter set, determine a model sample value corresponding to each influencing factor among the model sample values corresponding to each influencing factor, which is greater than the input environmental parameter and closest to the input environmental parameter, and record it as a first model sample value set; Determine the model sample value corresponding to each influencing factor that is smaller than the input environmental parameter and closest to the input environmental parameter among the model sample values corresponding to each influencing factor, and record them as a second model sample value set; Determine the solder bar tensile strength drop value-time variation function of the fitting model training sample parameter set group composed of the first model sample value set and the second model sample value set respectively; The intelligent prediction model for the decrease in tensile strength of solder bars is specifically: , is the set of input environment parameters, The tensile strength drop value of the solder bar corresponding to the input environmental parameters - time variation function, is the solder bar tensile strength drop value-time variation function corresponding to the first model sample value set, is the solder bar tensile strength drop value-time variation function corresponding to the second model sample value set, is the i-th element in the first model sample value set, is the i-th element in the second model sample value set, is the i-th element in the input environment parameter set.
2. According to claim 1, a multi-factor based intelligent detection method for tensile strength of solder bars, It is characterized in that Within the value range of the influencing factor, a number of values are taken according to the set parameter gradient, and the number of sample values corresponding to the influencing factor also includes: A number of samples corresponding to each influencing factor are divided into model sample values and model verification values, wherein there is at least one model verification value between any two adjacent model sample values.
3. According to claim 2, a multi-factor based intelligent detection method for tensile strength of solder bars, It is characterized in that The obtaining of the storage environment of the solder bar and determining the tensile strength of the solder bar based on the storage environment of the solder bar combined with the intelligent prediction model of the decrease value of the tensile strength of the solder bar specifically includes: Based on the storage environment of the solder bar, the storage period of the solder bar is divided into several statistical periods according to the environmental changes, wherein the storage environment in each statistical period does not change, and the storage environment in two adjacent statistical periods changes; Determine the solder bar tensile strength decrease value-time variation function corresponding to the storage environment corresponding to each statistical period; Based on the duration of the first statistical period, the duration of the first statistical period is substituted into the solder bar tensile strength decrease value-time change function corresponding to the first statistical period to obtain the solder bar tensile strength decrease value of the first statistical period; The cumulative decrease in the tensile strength of the solder bar is calculated according to the strength decrease cumulative formula, and the tensile strength of the solder bar is subtracted from the tensile strength of the solder bar to obtain the tensile strength of the solder bar; The strength reduction accumulation formula is specifically: , In the formula, is the cumulative decrease in the tensile strength of the solder bar in the first j statistical periods, is the initial equivalent time of the jth statistical period, is the length of the jth statistical period, is the cumulative decrease in the tensile strength of the solder bar. is the total number of statistical periods, It is the tensile strength decrease value of the solder bar corresponding to the jth statistical period - time variation function.
4. A multi-factor based intelligent detection system for the tensile strength of solder bars, used to implement the multi-factor based intelligent detection method for the tensile strength of solder bars as claimed in any one of claims 1 to 3, include: A sample value determination module, the sample value determination module is used to determine at least one environmental factor that affects the decrease in the tensile strength of the solder bar, recorded as an influencing factor, and based on the actual environment, determine the value range of each influencing factor, within the value range of the influencing factor, take a number of values according to the set parameter gradient, and form a number of sample values corresponding to the influencing factor, take a value from the several sample values corresponding to each influencing factor, and combine them into a sample parameter set, and traverse all sample values of all influencing factors until a new sample parameter set cannot be combined; A testing device, the testing device is electrically connected to the sample value determination module, and the testing device is used to test the relationship between the decrease value of the tensile strength of the solder bar and the time under each set of sample parameter sets in the test box; A model analysis module, the model analysis module is electrically connected to the test device, and the model analysis module is used to build an intelligent prediction model for the decrease in the tensile strength of the solder bar based on the relationship between the decrease in the tensile strength of the solder bar and time corresponding to all sample parameter sets; The tensile strength analysis module is electrically connected to the model analysis module, and is used to obtain the storage environment of the solder bar, and determine the tensile strength of the solder bar based on the storage environment of the solder bar and the intelligent prediction model of the tensile strength decrease value of the solder bar.
5. The intelligent detection system for tensile strength of solder bars based on multiple factors according to claim 4, It is characterized in that The model analysis module includes: A preliminary fitting unit, the preliminary fitting unit is used to determine a sample parameter set in which all elements are model sample values, recorded as a model training sample parameter set, record a non-model training sample parameter set in the sample parameter set as a model verification sample parameter set, and based on the solder bar tensile strength reduction value corresponding to the model training sample parameter set-simulation test duration coordinate point, call at least one known function for fitting, and obtain at least one fitting function corresponding to the model training sample parameter set; A fitting verification unit, the fitting verification unit is used to use two model training sample parameter sets between which no other model sample values exist as fitting model training sample parameter set groups, determine all model verification sample parameter sets between each fitting model training sample parameter set group, combine all fitting functions corresponding to each model training sample parameter set in the fitting model training sample parameter set group in pairs, obtain a plurality of fitting function groups, respectively calculate the loss values of fitting all fitting function groups for the model verification sample parameter set, screen out the fitting function group with the smallest loss value, and use the two fitting functions corresponding to the fitting function group as the solder bar tensile strength decrease value-time variation function of the model training sample parameter set corresponding to the fitting model training sample parameter set group; A model building unit is used to build an intelligent prediction model for the decrease in the tensile strength of a solder bar based on the decrease in the tensile strength of the solder bar-time variation function of the model training sample parameter set corresponding to all the fitting model training sample parameter set groups.
6. The intelligent detection system for tensile strength of solder bars based on multiple factors according to claim 4, It is characterized in that The tensile strength analysis module comprises: An initial calculation unit is used to calculate the Calculate the decrease in tensile strength of the first solder bar; The equivalent time length calculation unit is used to calculate the equivalent time length according to the formula Calculate the initial equivalent moment of the jth statistical period; Iterative calculation unit, the iterative calculation unit is used to calculate the Calculate the cumulative decrease in tensile strength of the first j solder bars; An accumulation unit, the accumulation unit is used to use the accumulated values of decrease in tensile strength of the solder bars in all statistical time periods calculated by the iterative calculation unit as the accumulated values of decrease in tensile strength of the solder bars; The difference obtaining unit is used to obtain the tensile strength of the solder bar by subtracting the accumulated decrease in the tensile strength of the solder bar from the tensile strength of the solder bar.
