Bonding copper wire conductivity analysis method and system based on neural network
By monitoring the acoustic emission signals and multi-stress data in real time, and combining deep neural networks for conductivity analysis and threshold adjustment, the shortcomings of bonded copper wire conductivity analysis in the existing technology are solved, and its reliability is improved in a multi-stress environment and performance.
Patent Information
- Application Number
- CN202410995296.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-07-24
AI Technical Summary
The existing neural network-based bonded copper wire conductivity analysis method is difficult to monitor and analyze in real time in a multi-stress environment, and the threshold setting and dynamic adjustment lack flexibility, so it cannot adapt to complex and changeable practical application scenarios.
By monitoring the acoustic emission signals in real time and combining real-time data of mechanical, thermal and electrical stress, a real-time performance evaluation model is constructed using deep neural networks, dynamically adjusting the alert threshold, and making adjustment suggestions for intermetallic compound control based on the final conductive performance evaluation results.
Real-time prediction and dynamic adjustment of the conductivity of bonded copper wires is achieved, which improves its reliability and performance in multi-stress environments and ensures the stability and reliability of solder joints.
Smart Images

Figure CN118549742B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of copper bonding wire detection, and in particular to a method and system for analyzing the conductive properties of copper bonding wires based on a neural network. Background Art
[0002] In the field of electronic packaging, the conductivity analysis of copper bonding wires has always been one of the key technologies to ensure the reliability and performance of electronic equipment. With the rapid development of microelectronics technology, the performance requirements for copper bonding wires are increasing, especially the stability under multiple stress environments. Traditional conductivity analysis methods mainly rely on static testing and finite element analysis. Although these methods can evaluate the performance of copper bonding wires to a certain extent, they have obvious deficiencies in real-time monitoring and comprehensive analysis of multiple stresses. For example, static testing cannot capture the impact of dynamic stress changes on conductivity, while finite element analysis is limited by model simplification and computing resources, making it difficult to achieve real-time and high-precision performance prediction.
[0003] In recent years, with the advancement of artificial intelligence technology, especially the application of neural networks in data analysis and pattern recognition, new possibilities have been provided for real-time analysis of the conductive properties of bonding copper wires. Neural networks can handle complex data relationships and predict unknown data by learning patterns in historical data.
[0004] In the prior art, Zhang Peng's "Exploring the Reliability of Copper Wire Bonding Process in Power Transistor Packaging", article number: 2095-6487 (2019) 04-0206-02, discloses the concept of copper wire bonding, which specifically describes that the reliability of ball solder bonding is mainly affected by the growth state of intermetallic compounds in the solder joint. During the bonding process, a certain amount of intermetallic compounds are formed between the pad and the bond, but once the amount of intermetallic compounds is too large, it will cause an increase in resistance, which in turn leads to an increase in heat at the solder joint position. The temperature increase at the solder joint position will accelerate the growth rate of the intermetallic compounds, which will increase the resistance again. This vicious cycle will lead to the failure of the solder joint, so the intermetallic compounds play a key role in the reliability of copper wire bonding.
[0005] However, existing neural network-based conductivity performance analysis methods often ignore the important role of acoustic emission signals in multi-stress environments, and how to comprehensively use real-time data of mechanical, thermal and electrical stresses for performance evaluation. In addition, existing technologies lack flexibility in threshold setting and dynamic adjustment, making it difficult to adapt to complex and changeable practical application scenarios.
[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention
[0007] The purpose of the present invention is to provide a method and system for analyzing the conductive properties of bonding copper wires based on a neural network to solve the problems raised in the above-mentioned background technology.
[0008] To achieve the above object, the present invention provides the following technical solutions:
[0009] The conductive performance analysis method of bonding copper wire based on neural network includes the following specific steps:
[0010] Step S1: Based on the acoustic emission sensor, real-time monitoring of acoustic emission signal data of the bonding copper wire under all stresses is performed, and the monitoring data is recorded;
[0011] Step S2, applying fast Fourier transform to analyze the spectrum of the acoustic emission signal data, extracting key frequency components, and using this to determine the key stress area of the bonding copper wire, the key stress includes mechanical stress, thermal stress and electrical stress;
[0012] Step S3, collecting the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire in real time, analyzing and processing them, and generating a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2;
[0013] Step S4, collecting the temperature change rate, thermal expansion coefficient and thermal conductivity of the thermal stress area of the bonding copper wire in real time, and analyzing and processing them to generate a second evaluation index, which is used to evaluate and adjust the corresponding characteristics of the thermal stress extracted in step S2;
[0014] Step S5, collecting the electric field strength and dielectric constant change of the bonding copper wire electric stress area in real time, analyzing and processing, and generating a third evaluation index, which is used to evaluate and adjust the corresponding characteristics of the electric stress extracted in step S2;
[0015] Step S6, obtaining the historical data of the conductive performance of the bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, constructing a training data set, and combining the extracted acoustic emission signal features with the known conductive performance data of the copper wire in the training data set, using a deep neural network to construct a real-time performance evaluation model, the input of the model is the signal feature, and the output is the conductive performance prediction value;
[0016] Step S7, obtaining historical data of electrical conductivity, and setting warning thresholds of electrical conductivity of bonding copper wires under different key stresses based on the historical data;
[0017] Step S8, obtaining the first evaluation index, the second evaluation index and the third evaluation index, and performing analysis and processing to construct a threshold adjustment model, which is used to dynamically adjust the warning threshold in step S7;
[0018] Step S9, obtaining the conductivity prediction value output by the real-time performance evaluation model in step S6, and calculating the matching degree between the conductivity prediction value and the warning threshold dynamically adjusted in step S8, to generate the final conductivity evaluation result of the bonding copper wire;
[0019] Step S10, obtaining the final conductive performance evaluation result, determining the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and proposing adjustment suggestions for the control of the intermetallic compounds.
[0020] A bonding copper wire conductivity analysis system based on a neural network, the system is used to execute the bonding copper wire conductivity analysis method based on a neural network, comprising:
[0021] Monitoring and recording module: used to monitor the acoustic emission signal data of the bonding copper wire under all stresses in real time based on the acoustic emission sensor, and record the monitoring data;
[0022] Spectrum analysis module: used to apply fast Fourier transform to analyze the spectrum of acoustic emission signal data, extract key frequency components, and use this to determine the key stress area of the bonding copper wire, which includes mechanical stress, thermal stress and electrical stress;
[0023] The first index generation module is used to collect the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire in real time, analyze and process them, and generate a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2;
[0024] The second index generation module is used to collect the temperature change rate, thermal expansion coefficient and thermal conductivity of the thermal stress area of the bonding copper wire in real time, and analyze and process them to generate a second evaluation index, which is used to evaluate and adjust the corresponding characteristics of the thermal stress extracted in step S2;
[0025] The third index generation module is used to collect the electric field strength and dielectric constant changes of the bonding copper wire electric stress area in real time, and analyze and process them to generate a third evaluation index, which is used to evaluate and adjust the corresponding characteristics of the electric stress extracted in step S2;
[0026] Performance evaluation module: used to obtain the historical data of the conductive performance of the bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, build a training data set, and combine the extracted acoustic emission signal features with the known copper wire conductive performance data in the training data set to build a real-time performance evaluation model using a deep neural network. The model input is the signal feature and the output is the conductive performance prediction value;
[0027] Threshold setting module: used to obtain the historical data of conductivity performance, and based on the historical data, set the warning threshold of the conductivity performance of the bonding copper wire under different key stresses;
[0028] Threshold adjustment module: used to obtain the first evaluation index, the second evaluation index and the third evaluation index, and perform analysis and processing to construct a threshold adjustment model, which is used to dynamically adjust the warning threshold in the threshold setting module;
[0029] Evaluation result generation module: used to obtain the conductivity prediction value output by the real-time performance evaluation model in the performance evaluation module, and calculate the matching degree between the conductivity prediction value and the warning threshold dynamically adjusted by the threshold adjustment module to generate the final conductivity evaluation result of the bonding copper wire;
[0030] Adjustment suggestion module: used to obtain the final conductive performance evaluation results, determine the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and put forward adjustment suggestions for intermetallic compound control.
[0031] Compared with the prior art, the beneficial effects of the present invention are: by real-time monitoring of acoustic emission signals and combining real-time data of mechanical, thermal and electrical stresses, a real-time performance evaluation model is constructed using a deep neural network. The model can not only predict the conductive properties of the bonding copper wire in real time, but also dynamically adjust the warning threshold to adapt to different stress conditions. In addition, the present invention can also propose adjustment suggestions for intermetallic compound control based on the final conductive performance evaluation results, thereby further improving the reliability and performance of the bonding copper wire. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 It is a schematic diagram of the overall method flow of the present invention;
[0033] Figure 2 It is a logic block diagram of the method of the present invention;
[0034] Figure 3 It is a system module block diagram of the present invention. DETAILED DESCRIPTION
[0035] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.
[0036] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly. Example
[0037] See also Figure 1 and Figure 2 , the present invention provides a technical solution:
[0038] A method for analyzing the conductive properties of bonding copper wires based on a neural network, the specific steps comprising:
[0039] Step S1: Based on the acoustic emission sensor, real-time monitoring of acoustic emission signal data of the bonding copper wire under all stresses is performed, and the monitoring data is recorded;
[0040] Step S2, applying fast Fourier transform to analyze the spectrum of the acoustic emission signal data, extracting key frequency components, and using this to determine the key stress area of the bonding copper wire, the key stress includes mechanical stress, thermal stress and electrical stress;
[0041] Step S3, collecting the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire in real time, analyzing and processing them, and generating a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2;
[0042] Step S4, collecting the temperature change rate, thermal expansion coefficient and thermal conductivity of the thermal stress area of the bonding copper wire in real time, and analyzing and processing them to generate a second evaluation index, which is used to evaluate and adjust the corresponding characteristics of the thermal stress extracted in step S2;
[0043] Step S5, collecting the electric field strength and dielectric constant change of the bonding copper wire electric stress area in real time, analyzing and processing, and generating a third evaluation index, which is used to evaluate and adjust the corresponding characteristics of the electric stress extracted in step S2;
[0044] Step S6, obtaining the historical data of the conductive performance of the bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, constructing a training data set, and combining the extracted acoustic emission signal features with the known conductive performance data of the copper wire in the training data set, using a deep neural network to construct a real-time performance evaluation model, the input of the model is the signal feature, and the output is the conductive performance prediction value;
[0045] Step S7, obtaining historical data of electrical conductivity, and setting warning thresholds of electrical conductivity of bonding copper wires under different key stresses based on the historical data;
[0046] Step S8, obtaining the first evaluation index, the second evaluation index and the third evaluation index, and performing analysis and processing to construct a threshold adjustment model, which is used to dynamically adjust the warning threshold in step S7;
[0047] Step S9, obtaining the conductivity prediction value output by the real-time performance evaluation model in step S6, and calculating the matching degree between the conductivity prediction value and the warning threshold dynamically adjusted in step S8, to generate the final conductivity evaluation result of the bonding copper wire;
[0048] Step S10, obtaining the final conductive performance evaluation result, determining the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and proposing adjustment suggestions for the control of the intermetallic compounds. Example
[0049] Further explanation based on the first embodiment, the real-time monitoring of acoustic emission signal data of the bonding copper wire under all stresses based on the acoustic emission sensor specifically includes the following contents:
[0050] Use a highly sensitive acoustic emission sensor to monitor the acoustic emission signals of the bonding copper wire under various stresses in real time;
[0051] The integrity and accuracy of the signal data are ensured by precisely installing the sensor and using a high-speed data acquisition system to record the signal at a sampling frequency of at least 50kHz;
[0052] The specific steps are as follows: first, select a high-sensitivity acoustic emission sensor model PACR15i, which has a sensitivity of 0.01V / μbar and can capture tiny acoustic emission signals;
[0053] Fix the sensor to the predetermined monitoring point of the bonding copper wire through a special fixture to ensure that the contact surface between the sensor and the copper wire is clean and free of oil to reduce signal loss and noise interference;
[0054] Connect the sensor to the NI-9233 data acquisition card, which supports sampling rates up to 50kHz and has 4-channel simultaneous acquisition capabilities;
[0055] Install NI-DAQmx driver and LabVIEW software on the computer, configure the sampling frequency of the data acquisition card to 50kHz, and set the trigger mode to software trigger to ensure that data recording can start immediately when the copper wire is stressed;
[0056] Start the LabVIEW program, start real-time monitoring of the acoustic emission signal, and display the signal strength and waveform through the software interface;
[0057] Record acoustic emission signal data for at least 10 minutes to ensure that sufficient data is captured for subsequent analysis;
[0058] The application of fast Fourier transform to analyze the spectrum of acoustic emission signal data and extract key frequency components specifically includes the following contents:
[0059] Performing spectrum analysis on the acoustic emission signal data collected in step S1 using fast Fourier transform;
[0060] By writing FFT analysis module in LabVIEW environment, the signal is preprocessed and transformed;
[0061] Extract key frequency components and use threshold detection algorithm to determine the frequencies associated with copper wire stress;
[0062] The specific steps are:
[0063] In LabVIEW, first create an FFT analysis module, which contains the following key parts:
[0064] Data preprocessing, using a low-pass filter to remove high-frequency noise and baseline correction to remove the baseline drift of the signal;
[0065] FFT calculation: Perform FFT transformation on the preprocessed signal using the formula: ,in:
[0066] X(f) is the frequency domain signal, which represents the signal amplitude at frequency f;
[0067] It is a time domain signal, indicating the time point The signal value at
[0068] N is the signal length, that is, the total number of sampling points;
[0069] f is the frequency in Hertz (Hz);
[0070] For spectrum analysis, the FFT results are analyzed to extract the frequency components, and the threshold detection algorithm is used to determine the key frequency components;
[0071] The threshold detection algorithm is to calculate the amplitude value of each frequency component in the FFT result; set the threshold to 5% of the signal peak value, that is, , where Threshold is the threshold used to determine whether a frequency component is a critical frequency; Max (Amplitude) is the maximum value of all frequency component amplitude values;
[0072] A threshold comparison is performed on the amplitude value of each frequency component. If the amplitude value is greater than the threshold, the frequency component is considered to be a key frequency component.
[0073] Use the following formula to label the key frequency components:
[0074] ;
[0075] Among them, KeyFrequency is the set of key frequency components; represents the i-th frequency point; is the frequency component The amplitude value of
[0076] In the LabVIEW interface, create a graphic display area for real-time display of spectrum analysis results and marked key frequency components;
[0077] Display the FFT results in the form of a spectrum graph, with the horizontal axis being frequency and the vertical axis being amplitude;
[0078] On the spectrum graph, use different colors or markers to mark key frequency components;
[0079] Add a data table to list all the marked key frequency components and their corresponding amplitude values;
[0080] Finally, the key frequency components are correlated with the stress state of the copper wire to determine the stress area;
[0081] According to the known copper wire stress-frequency relationship database, the marked key frequency components are matched with the mechanical stress, thermal stress and electrical stress of the copper wire;
[0082] Determine which key frequency components correspond to the specific stress types of the copper wire, and infer the stress distribution of the copper wire based on this. Example
[0083] Further explanation based on the second embodiment, the vibration frequency and magnetic parameters of the mechanical stress area are collected in real time, and analyzed and processed to generate a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2, specifically including the following contents:
[0084] The vibration frequency is marked as ZDf, and the magnetic parameter is marked as CXc;
[0085] Vibration frequency: Use an accelerometer or vibration sensor to monitor the vibration frequency of the copper wire. The change in vibration frequency reflects the dynamic response of the copper wire under mechanical stress.
[0086] Magnetic parameters: If the copper wire material is magnetic, the effect of mechanical stress can be evaluated by measuring its magnetic parameters, such as the change in magnetic permeability;
[0087] The vibration frequency ZDf is obtained as follows:
[0088] 1. Selection of monitoring points: Select the current monitoring point and the vibration monitoring points on its adjacent side;
[0089] 2. Measurement of vibration frequency: Use a vibration sensor to measure the vibration frequencies of the two monitoring points, which are recorded as and ;
[0090] 3. Anomaly Detection: Defining a Normal Range ,if or If any of the above is outside this range, the monitoring point is marked as abnormal.
[0091] 4. Calculation of ZDf: If both monitoring points are normal, calculate the vibration frequency difference between the current monitoring point and the adjacent monitoring point:
[0092] ;
[0093] If any monitoring point is abnormal, ZDf is marked as an abnormal value, as follows:
[0094] ;
[0095] Among them, NaN1 means "not a number" and is used to indicate abnormal conditions;
[0096] The magnetic parameter CXc is obtained as follows:
[0097] 1. Selection of monitoring points: Select the current monitoring point and the magnetic monitoring points on its adjacent side;
[0098] 2. Measurement of magnetic parameters: Use magnetic sensors to measure the magnetic parameters of the two monitoring points. In this embodiment, the magnetic parameters are magnetic permeability, which are recorded as and ;
[0099] 3. Anomaly Detection: Defining a Normal Range ,if or If any of the above is outside this range, the monitoring point is marked as abnormal.
[0100] 4. Calculation of CXc: If both monitoring points are normal, calculate the magnetic parameter difference between the current monitoring point and the adjacent monitoring point:
[0101] ;
[0102] If any monitoring point is abnormal, CXc is marked as an abnormal value, as follows:
[0103] ;
[0104] In the above scenario:
[0105] They are the lower and upper limits of the normal range of vibration frequency;
[0106] They are the lower and upper limits of the normal range of magnetic parameters, respectively;
[0107] NaN2 stands for "not a number" and is used to indicate an abnormal situation;
[0108] The mathematical formula for defining the first evaluation index is as follows:
[0109] ;
[0110] Among them, Index is the first evaluation index, which is used to evaluate the influence of mechanical stress, and its value range is (0,1); when Index is closer to 1, it means that the influence of mechanical stress is more significant; when it is closer to 0, it means that the influence is smaller;
[0111] ZDf is the vibration frequency value, which reflects the dynamic response of the copper wire under mechanical stress; the increase of ZDf indicates that the mechanical stress is enhanced, resulting in an increase in Index;
[0112] CXc is a magnetic parameter value that evaluates the effect of mechanical stress through changes in magnetic permeability; an increase in CXc also indicates an increase in mechanical stress, leading to an increase in Index;
[0113] is the weight coefficient used to adjust ZDf, CXc, and The degree of influence on the Index; these coefficients need to be optimized and adjusted through experimental data;
[0114] When analyzing the influence of mechanical stress, the weight coefficient The distribution of weights is crucial, as they determine the degree of influence of different parameters on the first evaluation index. The following are the weight distribution suggestions under two scenarios:
[0115] Scenario 1: The vibration frequency difference ZDf has a greater impact;
[0116] When the vibration frequency difference ZDf has a significant effect on the mechanical stress, the weight The value range of is 0.6 to 0.7; in this scenario, The value of is set to 0.65, which means that the influence of the vibration frequency difference on Index is dominant;
[0117] In this scenario, =0.65, =0.25, =0.10; such weight distribution makes the change of vibration frequency difference have the greatest impact on Index, while the influence of magnetic parameter difference CXc and frequency component amplitude value Amplitude is relatively small; when ZDf increases significantly, Index will quickly approach 1, indicating that the influence of mechanical stress is significant;
[0118] Scenario 2: The difference in magnetic parameters CXc has a greater impact;
[0119] When the magnetic parameter difference CXc has a significant effect on the mechanical stress, the weight The value range of is 0.6 to 0.7; in this scenario, The value of is set to 0.65, which means that the influence of magnetic parameter difference on Index is dominant;
[0120] In this scenario, =0.15, =0.65, =0.20; such weight distribution makes the change of magnetic parameter difference have the greatest impact on Index, while the vibration frequency difference ZDf and the frequency component amplitude value The effect is relatively small; when CXc increases significantly, the Index will quickly approach 1, indicating that the mechanical stress has a significant effect;
[0121] These weight assignments are based on an assessment of the relative importance of each parameter in the mechanical stress analysis; these weight values are further optimized through verification and adjustment of experimental data to ensure that the Index can accurately reflect the actual impact of mechanical stress;
[0122] is the amplitude value of the jth frequency component, M1 is the total number of frequency components;
[0123] It is the maximum value of all frequency component amplitude values and is used for normalization. ;
[0124] ;
[0125] ;
[0126] ;
[0127] ;
[0128] In the above formula, represents the natural exponential function, which is defined as:
[0129] Where x is an input parameter, which can be any real or complex number; in the current application of the formula, x is given by The calculated value reflects the combined effect of the vibration frequency difference, magnetic parameter difference and frequency component amplitude value;
[0130] n is the index variable for summation, which starts from 0 and increases to infinity. In practical applications, The convergence of does not need to be calculated to infinity, but only to a sufficiently accurate number of digits;
[0131] is each term in the exponential function, where represents the nth power of x, and n! represents the factorial of n, that is, ;
[0132] In the formula of the first evaluation index, exp(x) is used to construct a logistic regression model, where x is a composite parameter input into the exponential function, reflecting the combined effect of multiple physical quantities, such as vibration frequency difference, magnetic parameter difference, and frequency component amplitude value; by adjusting Weight coefficient, evaluation of the effect of the optimization model on mechanical stress. Example
[0133] Further explanation based on the third embodiment, the value range (0,1) of the first evaluation index Index is divided into the following three intervals:
[0134] Interval 1, ; In this range, the mechanical stress has a slight effect on the copper wire, and the Index value indicates a low stress level; at this time, ZDf, CXc and The changes are small and the system is in a light monitoring mode, focusing on small changes to prevent potential stress increases;
[0135] In this range, the value of Index indicates that the mechanical stress has little effect on the copper wire; specifically, when Index is in this range, the value of the vibration frequency difference ZDf is less than 10% of the average frequency change, the value of the magnetic parameter difference CXc is less than 5% of the average permeability change, and the frequency component amplitude value is The average change does not exceed 15%; these quantitative indicators show that the mechanical stress on the copper wire is relatively slight and is not enough to cause significant performance changes;
[0136] In this range, a threshold value is set as Index=0.2 and used as a judgment standard. When the Index is lower than this threshold, the system will start a slight stress monitoring mode, focusing on the slight changes in ZDf and CXc. If the Index is between 0.15 and 0.2, the system will increase the monitoring frequency to once an hour. If the Index is lower than 0.15, it will return to the normal monitoring frequency.
[0137] Set the threshold Index=0.2; when Index<0.2, the monitoring frequency f is once per hour, that is, f=1 time / hour; when When f=2 times / hour, the mathematical expression is:
[0138] ;
[0139] In this range, an increase in ZDf will cause a slight increase in the Index, but the change does not exceed 5%; changes in CXc have a more limited impact on the Index, with a change of no more than 3%; The change of has little impact on the Index. For every 10% increase in amplitude, the Index does not increase by more than 2%. These interaction rules ensure that even if a parameter changes, it will not have a significant impact on the Index, thus maintaining the stability of the system.
[0140] In this interval, the increase in ZDf leads to an increase in Index, and the change range is No more than 5%; the change in CXc leads to the change in Index No more than 3%; The change in the index causes the change in No more than 2%; the mathematical expression is:
[0141] ;
[0142] Interval 2, ; In this range, the mechanical stress has a moderate effect on the copper wire, and the Index value indicates that the stress level is moderate; ZDf, CXc, and The changes begin to be significant, and the system starts the medium stress response mode, increases the monitoring frequency, and records and analyzes the stress change trend;
[0143] When Index is in this range, the effect of mechanical stress on the copper wire is moderate; the value of ZDf is within the range of 10% to 30% of the average frequency change, the value of CXc is within the range of 5% to 20% of the average permeability change, and The average change is between 15% and 30%; these indicators show that the mechanical stress on the copper wire begins to have a certain impact on the performance, but has not yet reached a significant level;
[0144] In this range, the threshold Index=0.5 is set; when the Index exceeds this threshold, the system will start the medium stress response mode, increase the monitoring frequency of ZDf and CXc, and start recording the stress change trend; if the Index is between 0.4 and 0.5, the system will monitor once every half an hour; if the Index exceeds 0.5, the monitoring frequency will be increased to once every 15 minutes;
[0145] Set the threshold Index = 0.5; when Index>0.5, the monitoring frequency f is once every 15 minutes, that is, f=4 times / hour; when When f=2 times / hour, the mathematical expression is:
[0146] ;
[0147] In this interval, an increase in ZDf will cause the Index to rise by 10%; changes in CXc will also increase the impact on the Index by 8%; The change of has a more significant impact on Index. For every 10% increase in amplitude, Index increases by 5%. These interaction rules show that as mechanical stress increases, the influence of each parameter on Index also increases accordingly.
[0148] In this interval, the increase in ZDf leads to an increase in Index, and the change range is No more than 10%; the change in CXc leads to the change in Index Not more than 8%; The change in the index causes the change in No more than 5%; the mathematical expression is:
[0149] ;
[0150] Interval three, ;
[0151] In this range, the mechanical stress has a significant effect on the copper wire, and the Index value is close to 1; ZDf, CXc and When the changes in the values of the conductive materials reach a high level, the system activates the high-stress emergency response mode, immediately conducts a comprehensive inspection and adjusts the production parameters to prevent the conductive performance from being damaged;
[0152] In this range, the value of Index indicates that the mechanical stress has a significant effect on the copper wire; the value of ZDf exceeds 30% of the average frequency change, the value of CXc exceeds 20% of the average permeability change, and The average change is more than 30%; these quantitative indicators show that the mechanical stress on the copper wire has reached a significant level, which has a significant impact on the conductive performance;
[0153] In this range, the threshold Index=0.85 is set; when the Index exceeds this threshold, the system will start the high-stress emergency response mode, immediately conduct a comprehensive inspection of the copper wire, and adjust the production parameters to reduce mechanical stress; if the Index is between 0.8 and 0.85, the system will monitor every 5 minutes; if the Index exceeds 0.85, the emergency shutdown procedure will be immediately executed;
[0154] Set the threshold Index=0.85; when Index>0.85, the system executes the emergency shutdown procedure; when When the monitoring frequency f is once every 5 minutes, that is, f = 12 times / hour; the mathematical expression is:
[0155] ;
[0156] In this range, the increase of ZDf will cause the Index to rise significantly, with a change of 20%; the change of CXc will also significantly increase the impact of the Index, with a change of 15%; The change of has the most significant impact on Index. For every 10% increase in amplitude, Index increases by 10%. These interaction rules show that when mechanical stress increases significantly, the influence of each parameter on Index increases significantly, and urgent measures need to be taken to protect the conductivity of copper wire.
[0157] In this interval, the increase in ZDf leads to an increase in Index, and the change range is No more than 20%; the change in CXc leads to the change in Index Not more than 15%; The change in the index causes the change in No more than 10%; the mathematical expression is:
[0158] . Example
[0159] Further explanation based on the fourth embodiment, the temperature change rate, thermal expansion coefficient and thermal conductivity are marked as ΔT, CTE and kr in sequence, forming the temperature change rate ΔT, thermal expansion coefficient CTE and thermal conductivity kr;
[0160] For the temperature change rate ΔT: it is to monitor the temperature change rate of the copper wire in real time, that is, the temperature change per unit time;
[0161] Acquisition method: Use a temperature sensor (such as a thermocouple or infrared sensor) to continuously monitor the temperature of the copper wire and calculate its rate of change over time; in degrees Celsius per second (°C / s);
[0162] The calculation expression is: , the parameters are explained as follows:
[0163] The time The temperature at , in degrees Celsius (°C);
[0164] are two different time points, in seconds (s);
[0165] ΔT represents the rate of temperature change between these two time points, in degrees Celsius per second (°C / s);
[0166] For the coefficient of thermal expansion CTE: it is the length change rate of the copper wire under thermal stress, reflecting the thermal expansion characteristics of the material;
[0167] Collection method: Measure the length change of the copper wire under temperature change by laser interferometer or displacement sensor; express it as the ratio of length change per degree Celsius (ppm / °C);
[0168] The calculation expression is: , the parameters are explained as follows:
[0169] Temperature The length of the copper wire at the time, in meters (m); Same as above;
[0170] CTE represents the rate of change of length of copper wire under temperature change, and the unit is the proportion of length change per degree Celsius (such as ppm / °C);
[0171] Thermal conductivity kr: The ability of copper wire to conduct heat, which is related to the energy distribution under thermal stress;
[0172] Collection method: Use a thermal conductivity measuring instrument (such as a heat flow meter) to measure the thermal conductivity of the copper wire at different temperatures; the unit is watts per meter Kelvin (W / mK);
[0173] The calculation expression is: , the parameters are explained as follows:
[0174] Q is the amount of heat that passes through the copper wire in time t, in joules (J);
[0175] d is the thickness of the copper wire, in meters (m);
[0176] Am is the cross-sectional area of the copper wire in square meters (m²);
[0177] is the temperature difference between the two ends of the copper wire, in degrees Celsius (°C);
[0178] t is the time of heat transfer in seconds (s);
[0179] kr represents the thermal conductivity of the copper wire, in watts per meter Kelvin (W / mK);
[0180] The second evaluation index is defined as E2, and the calculation formula is:
[0181] ;
[0182] The parameters are explained as follows:
[0183] It is The temperature change rate at each time point, in °C / s;
[0184] is the maximum value of the temperature change rate at all time points, in °C / s;
[0185] It is The coefficient of thermal expansion at each time point, in ppm / °C;
[0186] is the average value of the coefficient of thermal expansion at all time points, in ppm / °C;
[0187] is the standard deviation of the coefficient of thermal expansion, in ppm / °C;
[0188] It is Thermal conductivity at each time point, in W / mK;
[0189] is the minimum value of thermal conductivity at all time points, in W / mK;
[0190] is the total number of time points;
[0191] when When it increases, the first term in the numerator increases, resulting in an increase in the second evaluation index, reflecting the effect of the increase in the temperature change rate on the thermal stress characteristics;
[0192] when Deviation When , the second term in the numerator decreases, resulting in a decrease in the second evaluation index, which reflects the influence of the abnormal change of the thermal expansion coefficient on the thermal stress characteristics;
[0193] when When increases, the third term in the numerator increases, resulting in an increase in the second evaluation index, reflecting the effect of the increase in thermal conductivity on the thermal stress characteristics;
[0194] The value range of the second evaluation index E2 is set to (0, 1); when the index is closer to 1, it means that the thermal stress characteristics are better, and the electrical conductivity of the copper wire is less affected by the thermal stress; when the index is closer to 0, it means that the thermal stress characteristics are worse, and the electrical conductivity of the copper wire is more affected by the thermal stress;
[0195] For the exponential function and natural logarithmic function in the formula, the specific expressions are as follows:
[0196] Exponential function: , where e is the base of natural logarithms and x represents ;
[0197] Natural logarithm function: , which represents the natural logarithm with base e. Example
[0198] Further explanation based on the fifth embodiment, the value range (0,1) of the second evaluation index E2 is divided into three intervals, namely interval I (0, 0.42), interval II [0.42, 0.72), interval III [0.72, 1);
[0199] Interval I (0, 0.42), in interval I, the E2 value is less than 0.42, the thermal stress characteristics of the copper wire are poor, and the conductivity is greatly affected. At this time, ΔT exceeds the average value by 50%, the standard deviation of CTE deviates from the average value by more than 30%, and kr is lower than the average value by 20%. The system needs to strengthen cooling measures to reduce the impact of thermal stress and ensure stable electrical connections;
[0200] In interval I, the value of the second evaluation index E2 is less than 0.42, indicating that the thermal stress characteristics of the copper wire are poor and the electrical conductivity is greatly affected by thermal stress; when the E2 value is in this interval, the temperature change rate ΔT of the copper wire exceeds 50% of the average value, and the thermal expansion coefficient CTE deviates from the average value. More than 30% standard deviation , the thermal conductivity kr is lower than 20% of the average value; in this case, the conductivity of the copper wire is significantly reduced, resulting in unstable electrical connection;
[0201] The judgment criteria are set as follows: when the E2 value is less than 0.42, the thermal stress adjustment mechanism is activated; the threshold is set to 0.25, and when the E2 value is lower than this threshold, the system automatically increases cooling measures to reduce the temperature change rate and thermal expansion coefficient; if the E2 value is between 0.25 and 0.42, the intensity of the cooling measures is adjusted to gradually reduce the impact of thermal stress on the copper wire;
[0202] Interaction rule description: In interval I, the increase of ΔT leads to an increase in E2 value, but the decrease of CTE and kr causes E2 value to decrease; specifically, for every 10% increase of ΔT, E2 value increases by 5%; for every 10% deviation of CTE from the average value, E2 value decreases by 3%; for every 10% decrease of kr, E2 value decreases by 2%; these changes act together on E2 value, affecting the thermal stress characteristics and electrical conductivity of copper wire;
[0203] Mathematical expression of interaction rules:
[0204] ;
[0205] ;
[0206] ;
[0207] Parameter explanation,
[0208] It is the percentage of increase in E2 value for every 10% increase in temperature change rate ΔT;
[0209] It is the percentage by which the E2 value decreases when the coefficient of thermal expansion CTE deviates from the average value by 10%;
[0210] It is the percentage of E2 value decrease for every 10% decrease in thermal conductivity kr;
[0211] They represent the percentage changes of ΔT, CTE, and kr relative to their initial values, respectively;
[0212] In interval I, when ΔT increases, the E2 value will increase, but the decrease in CTE and kr will cause the E2 value to decrease; if ΔT increases by 10%, the E2 value will increase by 5%; if CTE deviates from the average by 10%, the E2 value will decrease by 3%; if kr decreases by 10%, the E2 value will decrease by 2%; other parameters remain unchanged or change slightly to avoid ambiguity in the impact on the E2 value;
[0213] Interval II [0.42, 0.72), E2 value is between 0.42 and 0.72, the thermal stress characteristics of the copper wire are moderate, the conductivity is relatively stable, the system maintains the current cooling measures and adjusts according to the E2 value to optimize the thermal stress state;
[0214] In interval II, the value of the second evaluation index E2 is between 0.42 and 0.72, indicating that the thermal stress characteristics of the copper wire are at a medium level and the electrical conductivity is moderately affected by the thermal stress; the specific quantitative description is: when the E2 value is in this interval, the electrical conductivity of the copper wire is relatively stable;
[0215] The judgment criteria are set as follows: when the E2 value is between 0.42 and 0.72, the current thermal stress management strategy is maintained; the threshold is set to 0.5, when the E2 value is lower than this threshold, the system maintains the current cooling measures unchanged; when the E2 value is higher than this threshold, the system gradually reduces the cooling measures to optimize the thermal stress state of the copper wire;
[0216] Interaction rule description: In interval II, a small change in ΔT has little effect on the E2 value, but changes in CTE and kr will significantly affect the E2 value; specifically, for every 5% change in ΔT, the E2 value changes by 1%; for every 5% change in CTE, the E2 value changes by 2%; for every 5% change in kr, the E2 value changes by 1.5%; these changes work together on the E2 value to maintain the thermal stress characteristics of the copper wire and the stability of the conductive properties;
[0217] Mathematical expression of interaction rules:
[0218] ;
[0219] ;
[0220] ;
[0221] Parameter explanation:
[0222] It is the percentage change of E2 value for every 5% change of temperature change rate ΔT;
[0223] It is the percentage change of E2 value for every 5% change of coefficient of thermal expansion CTE;
[0224] It is the percentage change of E2 value for every 5% change of thermal conductivity kr;
[0225] They represent the percentage changes of ΔT, CTE, and kr relative to their initial values, respectively;
[0226] In interval II, a small change in ΔT has little effect on the E2 value, but changes in CTE and kr will significantly affect the E2 value; if ΔT changes by 5%, the E2 value will change by 1%; if CTE changes by 5%, the E2 value will change by 2%; if kr changes by 5%, the E2 value will change by 1.5%; other parameters remain unchanged or change slightly to avoid ambiguity in the impact on the E2 value;
[0227] Interval III [0.72, 1), in interval III, the E2 value is greater than 0.72, the thermal stress characteristics of the copper wire are good, the conductivity is slightly affected, ΔT is 20% lower than the average value, kr is 10% higher than the average value, and the system gradually reduces the cooling measures until it stops to maintain the best thermal stress state of the copper wire;
[0228] In interval III, the value of the second evaluation index E2 is greater than 0.72, indicating that the thermal stress characteristics of the copper wire are good and the electrical conductivity is less affected by thermal stress; the specific quantitative description is: when the E2 value is in this interval, the temperature change rate ΔT of the copper wire is lower than 20% of the average value, and the thermal conductivity kr is higher than 10% of the average value; in this case, the electrical conductivity of the copper wire is good and the electrical connection is stable;
[0229] The judgment criteria are set as follows: when the E2 value is greater than 0.72, reduce the thermal stress management measures; the threshold is set to 0.75, when the E2 value is higher than this threshold, the system gradually reduces the cooling measures until it stops completely; when the E2 value is between 0.72 and 0.75, the system maintains the current cooling measures to maintain the thermal stress state of the copper wire;
[0230] Interaction rule description: In interval III, the decrease of ΔT leads to a decrease in E2 value, but the increase of CTE and kr causes the E2 value to increase; specifically, for every 5% decrease in ΔT, the E2 value decreases by 1%; the E2 value increases by 1.5%; for every 5% increase in kr, the E2 value increases by 2%; these changes act together on the E2 value to optimize the thermal stress characteristics and electrical conductivity of the copper wire;
[0231] Mathematical expression of interaction rules:
[0232] ;
[0233] Parameter explanation:
[0234] It is the percentage by which the E2 value decreases when the temperature change rate ΔT decreases by 5%;
[0235] It is the percentage of increase in E2 value for every 5% decrease in the average value of the coefficient of thermal expansion CTE;
[0236] It is the percentage of increase in E2 value for every 5% increase in thermal conductivity kr;
[0237] Respectively represent the percentage change of ΔT, CTE, and kr relative to their initial values. Example
[0238] Further explanation based on Example 6, the electric field strength and dielectric constant change are marked as Eq and Δε respectively, forming the electric field strength Eq and dielectric constant change Δε;
[0239] For the electric field strength Eq, the electric field strength refers to the electric stress intensity of the copper wire, with volts per meter (V / m) as the unit. This parameter reflects the stress state of the copper wire in the electric field;
[0240] Collection method: Use electric field sensors or calculation methods (based on voltage and distance) to measure or estimate the electric field strength of the copper wire;
[0241] The mathematical expression of Eq is:
[0242] ;
[0243] Parameter explanation:
[0244] V is the voltage difference between the two ends of the copper wire. The voltage difference between the two ends of the copper wire is measured by a voltage sensor and the unit is volt (V).
[0245] is the calculated length of the copper wire, obtained by measuring the actual length of the copper wire, in meters (m);
[0246] For the dielectric constant change Δε, the dielectric constant change refers to the percentage change of the dielectric constant of the copper wire material under the action of electrical stress. This parameter reflects the change of the polarization characteristics of the material under electrical stress;
[0247] Collection method: Use dielectric constant measuring equipment to monitor the dielectric constant of the copper wire material in real time and calculate its rate of change relative to the initial value;
[0248] Mathematical expression of dielectric constant change Δε:
[0249] ;
[0250] Parameter explanation:
[0251] It is the real-time dielectric constant of the copper wire material under the action of electrical stress. The dielectric constant of the copper wire material is monitored in real time by the dielectric constant measuring equipment.
[0252] It is the initial dielectric constant of the copper wire material, and is measured when there is no electrical stress.
[0253] The third evaluation index is defined as E3, and the calculation formula is as follows:
[0254] ;
[0255] Parameter explanation:
[0256] E3 is the third evaluation index, with a value range of (0,1), which represents the evaluation of the influence of electrical stress on the conductive properties of copper wire;
[0257] Eq is the electric field strength, in volts per meter (V / m), reflecting the stress state of the copper wire in the electric field;
[0258] Δε is the percentage change in dielectric constant, reflecting the change in polarization characteristics of the material under electrical stress;
[0259] c1 is the adjustment coefficient, which is used to adjust the influence of electric field intensity on the evaluation index;
[0260] c2 is the adjustment coefficient, which is used to adjust the influence of the change of dielectric constant on the evaluation index;
[0261] When the electric field strength has a greater influence, the weight distribution is as follows:
[0262] When the electric field intensity Eq has a crucial effect on the conductivity of the copper wire and the relative dielectric constant change Δε has a greater impact, the value range of c1 is 0.7 to 0.8, and the value range of c2 is 0.2 to 0.3;
[0263] In the scenario where the electric field strength has a greater impact, assuming that c1 is 0.75 and c2 is 0.25, the change in electric field strength will dominate the impact on the third evaluation index E3, and the impact of the change in dielectric constant will be relatively small; specifically, a slight change in electric field strength will lead to a significant change in E3, while the change in dielectric constant has a relatively gentle impact on E3. This weight distribution reflects the sensitivity of the copper wire's conductive performance when the electric field strength is the main influencing factor;
[0264] When the dielectric constant change has a greater impact, the weight distribution is as follows:
[0265] When the dielectric constant change Δε is crucial to the conductivity of the copper wire and the relative electric field strength Eq has a greater impact, the value range of c1 is 0.3 to 0.4, and the value range of c2 is 0.6 to 0.7;
[0266] Interaction description: In the scenario where the change in dielectric constant has a greater impact, assuming that c1 is 0.35 and c2 is 0.65; at this time, the change in dielectric constant will dominate the impact on the third evaluation index E3, and the impact of electric field strength will be relatively small; specifically, a small change in dielectric constant will lead to a significant change in E3, while the change in electric field strength has a relatively gentle impact on E3. This weight distribution reflects the sensitivity of copper wire conductivity when the change in dielectric constant is the main influencing factor;
[0267] A positive constant term used to ensure that the value of E3 will not be 0;
[0268] When E3 is closer to 1, it means that the electrical stress has a greater impact on the conductivity of the copper wire, and the conductivity of the copper wire is worse;
[0269] When E3 is closer to 0, it means that the influence of electrical stress on the conductivity of copper wire is smaller, and the conductivity of copper wire is better;
[0270] Normalization function:
[0271] ;
[0272] Parameter explanation:
[0273] is any value, indicating the value that needs to be exponentially calculated;
[0274] and The minimum and maximum values in the set of values, respectively. Example
[0275] Further description is given on the basis of the seventh embodiment, the value range (0,1) of the third evaluation index E3 is divided into the following intervals, and the evaluation adjustment description is performed for each interval;
[0276] Interval 1 (0.00, 0.15]; within this interval, the electrical stress affects the conductivity of the copper wire by less than 2%, which can be ignored, the electric field strength is 25% lower than the average value, and the dielectric constant changes by less than 3%; it is suitable for precision electronic equipment, which needs to maintain small changes in electric field strength and dielectric constant to maintain optimal performance;
[0277] When the third evaluation index E3 is in the range of 0.00 to 0.15, it indicates that the influence of electrical stress on the electrical conductivity of the copper wire is extremely small, and the electrical conductivity of the copper wire is close to the ideal state; specifically, the E3 value in this range indicates that the electric field strength Eq is lower than 25% of the average value, and the dielectric constant change Δε is less than 3%; in this case, the electrical conductivity of the copper wire is almost unaffected by electrical stress, and is suitable for precision electronic equipment with extremely high requirements for electrical conductivity;
[0278] The judgment criteria and rules are:
[0279] In this range, the threshold is set to 0.10 to judge the degree of optimization of the copper wire's conductive properties. When the E3 value is lower than 0.10, the copper wire's conductive properties are considered to be optimized. At this time, the electric field strength should be kept below 20% of the average value, and the dielectric constant change should be controlled within 2%. If the E3 value is higher than 0.10, the electric field strength and dielectric constant need to be further adjusted to reduce the impact of electrical stress on conductive properties.
[0280] The interaction rules are described as:
[0281] In this range, the increase of electric field strength Eq will lead to a slight increase of E3 value, and the increase is 1% for every 10% increase of Eq; at the same time, the increase of dielectric constant change Δε will also lead to an increase of E3 value, but the effect is smaller, that is, every 5% increase of Δε, the E3 value increases by 0.5%; this interaction rule shows that although the influence of electric stress is very small, it is still necessary to pay attention to the slight changes of electric field strength and dielectric constant to maintain the best conductivity of copper wire;
[0282] Mathematical expression of interaction rules: , ;
[0283] Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3;
[0284] When Eq increases by 10%, E3 increases by 1%; when Δε increases by 5%, E3 increases by 0.5%; other parameters remain unchanged to avoid ambiguity in the impact on E3;
[0285] Interval 2 (0.15, 0.35]; in this interval, the electrical conductivity of the copper wire begins to be affected by electrical stress, the electric field strength is between 25% and 45% of the average value, and the dielectric constant changes by 3% to 7%; the electric field strength and dielectric constant need to be stabilized to meet most application requirements;
[0286] When E3 is between 0.15 and 0.35, the influence of electrical stress on the electrical conductivity of copper wire begins to appear, but it is still at a low level; the E3 value in this range indicates that the electric field strength Eq is between 25% and 45% of the average value, and the dielectric constant change Δε is between 3% and 7%; in this case, the electrical conductivity of copper wire begins to be affected by electrical stress to a certain extent, but it can still meet the needs of most applications;
[0287] The judgment criteria and rules are as follows: within this range, the threshold is set to 0.25 to evaluate the stability of the copper wire's conductive properties; when the E3 value is lower than 0.25, the copper wire's conductive properties are considered stable, at which time the electric field strength should be kept below 35% of the average value, and the dielectric constant change should be controlled within 5%; if the E3 value is higher than 0.25, the electric field strength and dielectric constant need to be adjusted to reduce the impact of electrical stress on conductive properties;
[0288] The interaction rules are described as:
[0289] In this range, the increase of electric field strength Eq will lead to an increase of E3 value, and the increase is 2% for every 10% increase of Eq; at the same time, the increase of dielectric constant change Δε will also lead to an increase of E3 value, and the increase is 1.5% for every 5% increase of Δε; this interaction rule shows that the increase of electric field strength and dielectric constant will affect the conductivity of copper wire, and these parameters need to be closely monitored and adjusted;
[0290] Mathematical expression of interaction rules: , ;
[0291] Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3;
[0292] Description of interaction content: When Eq increases by 10%, E3 increases by 2%; when Δε increases by 5%, E3 increases by 1.5%; other parameters remain unchanged to avoid ambiguity in the impact on E3;
[0293] Interval 3 (0.35, 0.60]; the conductivity of the copper wire is affected by moderate electrical stress, the electric field strength is between 45% and 65% of the average value, and the dielectric constant changes by 7% to 12%; the electric field strength and dielectric constant need to be finely adjusted to maintain stable conductivity;
[0294] When the third evaluation index E3 is in the range of 0.35 to 0.60, the influence of electrical stress on the electrical conductivity of the copper wire increases significantly; the E3 value in this range indicates that the electric field strength Eq is between 45% and 65% of the average value, and the dielectric constant change Δε is between 7% and 12%; in this case, the electrical conductivity of the copper wire is affected by moderate electrical stress, and the electric field strength and dielectric constant need to be finely adjusted to maintain the electrical conductivity;
[0295] The judgment criteria and rules are as follows: within this range, the threshold is set to 0.48 to evaluate the equilibrium state of the copper wire's conductive properties; when the E3 value is lower than 0.48, the copper wire's conductive properties are considered to be low, and the electric field strength should be kept below 55% of the average value, and the dielectric constant change should be controlled within 10%; if the E3 value is higher than 0.48, it indicates that the electric stress has a greater impact, and the adjustment frequency of the electric field strength and dielectric constant needs to be increased;
[0296] The interaction rule is described as follows: In this interval, the increase in electric field strength Eq will lead to an increase in E3 value, and the increase is 3% for every 10% increase in Eq; at the same time, the increase in dielectric constant change Δε will also lead to an increase in E3 value, and the increase is 2% for every 5% increase in Δε; this interaction rule shows that the increase in electric field strength and dielectric constant will significantly affect the conductivity of copper wire, and these parameters need to be closely monitored and adjusted to maintain the stability of conductivity;
[0297] Mathematical expression of interaction rules: , ;
[0298] Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3;
[0299] When Eq increases by 10%, E3 increases by 3%; when Δε increases by 5%, E3 increases by 2%; other parameters remain unchanged to avoid ambiguity in the impact on E3;
[0300] Interval 4 (0.60, 1.00); in this interval, the electrical conductivity of the copper wire is significantly affected by electrical stress, the electric field strength is between 65% and 85% of the average value, and the dielectric constant changes by 12% to 18%. The electric field strength and dielectric constant need to be adjusted frequently to reduce the loss of electrical conductivity.
[0301] When the third evaluation index E3 is between 0.6 and 1, the electric stress has a greater impact on the electrical conductivity of the copper wire; the E3 value in this range indicates that the electric field strength Eq is between 65% and 85% of the average value, and the dielectric constant change Δε is between 12% and 18%; in this case, the electrical conductivity of the copper wire is significantly affected by the electric stress, and the electric field strength and dielectric constant need to be frequently adjusted to reduce the loss of electrical conductivity;
[0302] The judgment criteria and rules are as follows: within this range, the threshold is set at 0.73 to evaluate the degree of damage to the copper wire's conductivity; when the E3 value is lower than 0.73, the copper wire's conductivity is considered acceptable, and the electric field strength should be kept below 75% of the average value, and the dielectric constant change should be controlled within 15%; if the E3 value is higher than 0.73, it indicates that the electric stress is too large, and the electric field strength and dielectric constant need to be adjusted immediately to avoid further deterioration of the conductivity;
[0303] The interaction rule is described as follows: In this interval, the increase in electric field strength Eq will lead to an increase in E3 value, and the increase is 4% for every 10% increase in Eq; at the same time, the increase in dielectric constant change Δε will also lead to an increase in E3 value, and the increase is 3% for every 5% increase in Δε; this interaction rule shows that the increase in electric field strength and dielectric constant will have a significant impact on the conductive properties of copper wire, and immediate measures need to be taken to reduce the negative impact of electrical stress;
[0304] Mathematical expression of interaction rules: , ;
[0305] Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3;
[0306] When Eq increases by 10%, E3 increases by 4%; when Δε increases by 5%, E3 increases by 3%; other parameters remain unchanged to avoid ambiguity in the impact on E3. Example
[0307] Further illustrating on the basis of Example 8, the training data set of the deep neural network model includes historical acoustic emission signal data and corresponding bonding copper wire conductive performance data;
[0308] The real-time performance evaluation model is constructed, the input of which is the signal feature and the output is the predicted value of the conductive performance, and specifically includes the following contents:
[0309] Data acquisition and preprocessing: using high-precision sensors to collect the conductivity data and acoustic emission signal data of bonding copper wires under various stress conditions, and cleaning the data to remove noise and inconsistency to ensure data quality;
[0310] 1.1. Data collection: Use high-precision sensors to collect historical data on the conductivity of bonding copper wires under different mechanical stress, thermal stress, and electrical stress conditions, as well as the corresponding acoustic emission signal data;
[0311] 1.2. Data preprocessing: Clean the collected data, remove noise and inconsistent data, and ensure data quality; use the formula Perform data cleaning, including is the cleaned data. is the original data, Noise is the noise data;
[0312] Feature extraction: extract key features from the preprocessed acoustic emission signal data. These key features will be used for subsequent model training;
[0313] The specific feature extraction is:
[0314] Acoustic emission signal feature extraction: Extract features from the preprocessed acoustic emission signal data, such as the frequency, amplitude, and duration of the signal; use the formula , where Features is the extracted feature vector, Extract is the feature extraction function, is the acoustic emission signal data;
[0315] Construct a training data set, combine the extracted features with the corresponding conductive performance data, and form a training data set for training the deep neural network model to ensure that the model training has sufficient data support;
[0316] The data set construction is as follows: the extracted features are combined with the corresponding bonding copper wire conductivity data to construct a training data set; the formula is used ,in is the training dataset, It is The characteristics of the samples, is the corresponding conductivity performance data;
[0317] Model construction and training, a deep neural network is selected as the prediction model and trained using the constructed training data set to build a model that can predict the conductive properties of bonding copper wire;
[0318] The model selection is as follows: select a deep neural network (DNN) as the prediction model, train the DNN model using the constructed training data set; use the formula ,in is the trained DNN model, and Train is the training function;
[0319] Real-time performance evaluation: input the real-time collected acoustic emission signal characteristics into the trained model, and the model outputs the real-time conductive performance prediction value to achieve real-time monitoring and evaluation of the bonding copper wire performance;
[0320] Specifically, the acoustic emission signal features collected in real time are input into the trained DNN model; the model outputs the real-time predicted conductive properties of the bonding copper wire; the formula is used ,in is the predicted conductivity property, It is real-time feature data;
[0321] Model validation and optimization: Use independent test data sets to validate the model’s prediction accuracy, and adjust model parameters based on the validation results to optimize model performance and ensure the model’s prediction accuracy and reliability;
[0322] Specifically, use an independent test data set to verify the prediction accuracy of the model; adjust the model parameters based on the verification results to optimize the model performance; use the formula ,in is the optimized model, Optimize is the optimization function, is the test dataset. Example
[0323] Further explanation is given on the basis of the ninth embodiment, the acquisition of the conductivity historical data and setting the warning threshold of the conductivity of the bonding copper wire under different key stresses based on the historical data specifically includes the following contents:
[0324] 1. Obtain historical data on electrical conductivity:
[0325] 1.1. Data collection: Systematically collect historical data on the conductivity of bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, including but not limited to current, voltage, temperature and stress level;
[0326] 1.2. Data verification: Verify the collected historical data to ensure the accuracy and reliability of the data; use formula ,in is the verified data, Data is the original data, is the estimated error rate;
[0327] 1. Threshold calculation
[0328] 1.1. Basic threshold calculation: First, calculate the average conductivity based on historical data and standard deviation ; Use the following formula to calculate the basic warning threshold :
[0329] ;
[0330] Among them, k3 is a constant used to adjust the sensitivity of the threshold;
[0331] 2. Threshold optimization
[0332] 2.1. Sensitivity analysis: In order to optimize the threshold, it is necessary to analyze the impact of different k3 values on the warning threshold. Specifically, by simulating the changes in conductivity under different stress conditions, the sensitivity of the k3 value to the threshold is evaluated;
[0333] 2.2. Optimization formula: According to the sensitivity analysis results, adjust the k3 value to optimize the warning threshold, and use the following formula for optimization;
[0334] ;
[0335] in, is an adjustment factor used to fine-tune the threshold; It is the sensitivity factor calculated based on the k3 value, reflecting the influence of k3 on the threshold change. To optimize the threshold;
[0336] 2.3. Sensitivity factor calculation: Sensitivity factor Calculated by the following formula:
[0337] ;
[0338] in, is the change in threshold, is the change in k3;
[0339] By introducing sensitivity analysis and optimization formulas, the warning threshold is adjusted dynamically to ensure that the threshold can respond to changes in conductive properties in a timely manner and will not cause false alarms due to oversensitivity;
[0340] By accurately calculating and dynamically adjusting the warning threshold, real-time monitoring and early warning of the conductive properties of the bonding copper wire are achieved.
[0341] Embodiment eleven:
[0342] Further explanation based on the tenth embodiment, the first evaluation index, the second evaluation index and the third evaluation index are obtained, analyzed and processed, and a threshold adjustment model is constructed. The model is used to dynamically adjust the warning threshold in step S7, and specifically includes the following contents:
[0343] The threshold adjustment model formula is:
[0344] ;
[0345] Parameter explanation,
[0346] ATd is the adjusted warning threshold, which is used to dynamically adjust the conductive performance warning threshold of the bonding copper wire;
[0347] is the basic warning threshold, calculated from the average conductivity and standard deviation;
[0348] is the adjustment factor used to fine-tune the sensitivity of the threshold;
[0349] is the sensitivity factor, reflecting the influence of k3 on the threshold change;
[0350] is an adjustment coefficient, which is used to adjust the influence of the first evaluation index, the second evaluation index and the third evaluation index on the threshold adjustment;
[0351] Index is the first evaluation index, reflecting the influence of mechanical stress;
[0352] E2 is the second evaluation index, reflecting the influence of thermal stress;
[0353] E3 is the third evaluation index, reflecting the influence of electrical stress;
[0354] Set the range of ATd to ; When ATd is closer to 0, it means that the current stress state has less influence on the conductivity; when ATd is closer to When , it means that the stress state has a more significant impact on the conductivity, and corresponding measures need to be taken;
[0355] ATd value range Divide into the following intervals, and give corresponding descriptions for each interval;
[0356] Interval 1 ; In this range, the ATd value is less than half of the basic warning threshold, indicating that the bonding copper wire is less affected by stress; Specifically, when Index is lower than 20% of the average value, E2 is lower than 15% of the average value, and E3 is lower than 10% of the average value, ATd will remain at a low level; at this time, the system should take slight monitoring measures and automatically check the conductivity once an hour to ensure stable performance;
[0357] The judgment criteria and rules are as follows:
[0358] In this range, set the ATd middle threshold to ; When ATd is lower than this value, the system enters low-risk monitoring mode and increases the frequency of data collection to once every 30 minutes; if ATd exceeds this value but is still lower than the upper limit of the interval, the early warning system is activated to prompt the operator to perform manual inspection;
[0359] The interaction rules are described as follows:
[0360] In this range, an increase in Index will lead to an increase of 5% in ATd, while an increase in E2 and E3 will lead to an increase of 3% and 2% in ATd, respectively. If Index decreases, ATd will decrease by 7%, while a decrease in E2 and E3 will lead to a decrease of 4% and 3% in ATd, respectively. These changes ensure that the warning threshold can be adjusted in time even under slight stress changes to maintain the sensitivity and accuracy of the system.
[0361] ;
[0362] Parameter explanation, is the change in the adjusted warning threshold; It is the change of Index, reflecting the change of mechanical stress; is the change in E2, reflecting the change in thermal stress; is the change of E3, reflecting the change of electrical stress;
[0363] when , hour,
[0364] ; At this time, if ATd exceeds the middle threshold of the current interval, the system will enter the early warning mode;
[0365] Interval 2 ; When ATd is in this range, the stress on the bonding copper wire is moderate; specifically, Index is between 20% and 50% of the average value, E2 is between 15% and 40% of the average value, and E3 is between 10% and 30% of the average value; at this time, the system should strengthen monitoring, automatically check the conductivity every 15 minutes, and be prepared to deal with performance degradation;
[0366] The judgment criteria and rules are as follows:
[0367] In this interval, the middle threshold is set to ; When ATd is lower than this value, the system maintains normal monitoring mode, but increases the depth of data analysis; if ATd exceeds this value, the automatic adjustment mechanism is activated to dynamically adjust the warning threshold according to real-time data;
[0368] The interaction rules are described as follows:
[0369] In this range, an increase in Index will lead to a 10% increase in ATd, while an increase in E2 and E3 will lead to a 7% and 5% increase in ATd respectively; if Index decreases, ATd will decrease by 12%, while a decrease in E2 and E3 will lead to a 8% and 6% decrease in ATd respectively; these changes ensure that the system can effectively adjust the warning threshold and maintain stable performance under moderate stress changes;
[0370] ;
[0371] Parameter explanation, is the change in the adjusted warning threshold; It is the change of Index, reflecting the change of mechanical stress; is the change in E2, reflecting the change in thermal stress; is the change of E3, reflecting the change of electrical stress;
[0372] when , , hour,
[0373]
[0374] At this time, if ATd exceeds the middle threshold of the interval, the system will start the automatic adjustment mechanism;
[0375] Interval 3 ; When ATd is in this range, the bonding copper wire is greatly affected by stress; specifically, Index exceeds 50% of the average value, E2 exceeds 40% of the average value, and E3 exceeds 30% of the average value; at this time, the system should immediately start the emergency response procedure, automatically check the conductivity every 5 minutes, and be prepared for manual intervention to prevent performance collapse;
[0376] The judgment criteria and rules are as follows:
[0377] In this interval, the middle threshold is set to ; When ATd is lower than this value, the system enters high-risk monitoring mode and activates all available resources for data analysis; if ATd exceeds this value, the highest-level management personnel will be notified immediately and emergency repairs or component replacement will be prepared;
[0378] The interaction rules are described as follows:
[0379] In this range, an increase in Index will cause ATd to increase by 15%, while increases in E2 and E3 will cause ATd to increase by 10% and 8% respectively; if Index decreases, ATd will decrease by 20%, while decreases in E2 and E3 will cause ATd to decrease by 12% and 10% respectively; these changes ensure that under large stress changes, the system can quickly and effectively adjust the warning threshold to protect the equipment from damage;
[0380] ;
[0381] Parameter explanation, is the change in the adjusted warning threshold; It is the change of Index, reflecting the change of mechanical stress; is the change in E2, reflecting the change in thermal stress; is the change of E3, reflecting the change of electrical stress;
[0382] when hour,
[0383] ;
[0384] At this point, if ATd exceeds the middle threshold of the interval, the system will immediately notify the highest-level management and prepare for emergency repairs or component replacement;
[0385] The division and description of these intervals ensure that the threshold adjustment model can dynamically adjust the warning threshold according to different stress levels, thereby achieving effective monitoring and protection of the conductive properties of the bonding copper wire.
[0386] Embodiment 12:
[0387] Further explanation is given on the basis of Example 11. In this example, five different types of bonding copper wires are selected as test objects. Each copper wire is tested under different mechanical, thermal and electrical stress conditions. The purpose of the test is to verify the effectiveness and accuracy of the threshold adjustment model in practical applications. First, the historical conductive performance data of each copper wire is collected, including the performance under different stress conditions. Based on these data, the basic warning threshold of each copper wire is calculated. ;
[0388] Next, a series of experiments were designed to simulate different levels of mechanical stress corresponding to the first evaluation index Index, thermal stress corresponding to the second evaluation index E2, and electrical stress corresponding to the third evaluation index E3. In each experiment, the actual value of the stress was recorded, and the threshold adjustment model was used to calculate the adjusted warning threshold ATd. Special attention was paid to the performance of the model in different stress ranges, that is, in the low stress range , medium stress range and high stress range Response;
[0389] During the experiment, high-precision sensors were used to monitor the conductivity of the copper wire and adjust the stress level in real time to simulate changes in the actual working environment. An automatic alarm system was also set up. When ATd exceeds the preset intermediate threshold, the system will automatically sound an alarm and take corresponding measures according to different monitoring modes (low risk, warning, automatic adjustment, high risk, emergency);
[0390] This series of experiments aims to demonstrate how the threshold adjustment model can dynamically adjust the warning threshold according to real-time data to ensure the stability and reliability of the bonding copper wire under various stress conditions. See the table below;
[0391] Table 1
[0392] This table shows the experimental data of five different copper wires under different stress conditions. Each row represents a specific test object and lists its corresponding first evaluation index Index, second evaluation index E2, third evaluation index E3, basic warning threshold , the adjusted warning threshold ATd and the corresponding monitoring mode. These data clearly show how the threshold adjustment model dynamically adjusts the warning threshold according to the changes in stress levels and takes corresponding monitoring measures according to the adjusted threshold;
[0393] By comparing the data in the table, we can clearly see the creativity and novelty of the threshold adjustment model proposed in the present invention in dynamically adjusting the warning threshold;
[0394] First, the model can adjust the warning threshold ATd in real time according to different stress levels Index, E2, and E3. This dynamic adjustment mechanism ensures that the monitoring system can quickly respond to various stress changes, thereby protecting the conductive properties of the bonding copper wire;
[0395] 1. Dynamically adjust the warning threshold: As can be seen from the table, as Index, E2, and E3 increase, ATd also increases, which shows that the model can adjust the warning threshold according to the actual stress situation, ensuring that the warning level is raised in time when the stress increases to prevent performance degradation;
[0396] 2. Interval division and response: The data in the table shows the changes in monitoring modes under different stress intervals. For example, copper wire A is in the low stress interval and its monitoring mode is "warning", while copper wires C, D, and E are in the high stress interval and their monitoring mode is "emergency". This mechanism of adjusting the monitoring mode according to the stress level enables the system to take the most appropriate response measures for different degrees of stress changes.
[0397] Analysis of numerical interaction changes:
[0398] 1. Impact of Index increase: When Index increases, ATd increases. For example, when Index of copper wire B increases from 10% to 25%, ATd increases from 105.61 to 135.55. This indicates that the increase in mechanical stress leads to an increase in the warning threshold, thus triggering stricter monitoring measures.
[0399] 2. Influence of E2 and E3: Similarly, the increase of E2 and E3 will also lead to the increase of ATd. For example, the E2 and E3 of copper wire C increase from 5% and 3% to 30% and 20% respectively, and ATd increases from 105.61 to 177.70; this shows that the increase of thermal stress and electrical stress will also affect the setting of the warning threshold;
[0400] 1. Low stress interval: In this interval, the increase of ATd is small. For example, the ATd of copper wire A increases by 5.61%. The system response is "warning", indicating that under slight stress changes, the system can maintain a lower monitoring frequency but still remain alert;
[0401] 2. Medium stress interval: In this interval, the increase of ATd is more significant. For example, the ATd of copper wire B increases by 12.96%. The system starts the "automatic adjustment" mode, indicating that the system can automatically adjust the monitoring strategy according to the medium stress changes;
[0402] 3. High stress interval: In this interval, the increase of ATd is the largest. For example, the ATd of copper wire C increases by 18.47%. The system enters the "emergency" mode, indicating that under high stress conditions, the system can quickly increase the monitoring level and prepare to take emergency measures;
[0403] Through these specific numerical analyses, it can be seen that the threshold adjustment model proposed in the present invention can not only dynamically adjust the warning threshold according to stress changes, but also adopt corresponding monitoring modes according to different stress intervals, thereby effectively protecting the conductive properties of the bonding copper wire; this dynamic adjustment mechanism and interval response strategy are the core innovations of the present invention, which significantly improve the flexibility and response efficiency of the monitoring system.
[0404] Embodiment 13:
[0405] Further explanation based on the eleventh embodiment, the conductivity prediction value output by the real-time performance evaluation model in step S6 is obtained, and the conductivity prediction value is matched with the warning threshold dynamically adjusted in step S8 to generate the final conductivity evaluation result of the bonding copper wire; specifically including the following contents:
[0406] Obtaining the predicted value of the conductive performance of the real-time performance evaluation model;
[0407] The deep neural network model constructed in step S6 receives the real-time signal characteristics as input and outputs a predicted value of the conductive performance; this predicted value is calculated based on the current mechanical stress, thermal stress and electrical stress conditions, and reflects the expected conductive performance of the bonding copper wire under the current stress state;
[0408] Get the dynamically adjusted warning threshold;
[0409] According to the threshold adjustment model in step S8, a dynamically adjusted warning threshold is obtained; this threshold is calculated based on historical data and current evaluation indexes (first evaluation index, second evaluation index and third evaluation index), and is used to reflect the threshold of monitoring and response measures that should be taken under current stress conditions;
[0410] Calculate the matching degree between the predicted value of conductivity and the warning threshold;
[0411] The conductivity prediction value is calculated using the following formula The matching degree PPd with the warning threshold ATd:
[0412] ;
[0413] Parameter explanation, is the predicted value of the conductive performance output by the real-time performance evaluation model; ATd is the dynamically adjusted warning threshold; PPd is the matching degree, which indicates the closeness between the predicted value and the threshold, expressed as a percentage;
[0414] Based on the matching calculation results, the final conductivity evaluation results are generated; if the matching degree is lower than the preset threshold, it is considered that the conductivity prediction value matches the warning threshold and the system is in a normal state; if the matching degree is higher than the preset threshold, corresponding monitoring and maintenance measures need to be taken according to the specific value of the matching degree.
[0415] Embodiment 14:
[0416] Further explanation is given on the basis of Example 13, the obtaining of the final conductivity evaluation result, determining the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and proposing adjustment suggestions for the control of the intermetallic compounds specifically include the following contents:
[0417] Analyze the influence of thermal stress on the electrical conductivity evaluation results;
[0418] According to the final conductivity evaluation result generated in step S9, the heat increase at the solder joint position is analyzed; the solder joint temperature change rate is calculated using the following formula :
[0419] ;
[0420] Parameter explanation, is the current solder point temperature; is the base temperature of the solder joint, usually the temperature under normal working conditions; is the rate of change of solder joint temperature, expressed as a percentage;
[0421] Determine the amount of intermetallic compounds;
[0422] Based on the solder joint temperature change rate, combined with historical data and experimental results, the amount of intermetallic compounds formed between the pad and the bond is determined; if the temperature change rate exceeds a preset threshold, it indicates that the growth rate of the intermetallic compounds is accelerated;
[0423] Propose adjustments for intermetallic compound control;
[0424] Based on the determination of the number of IMCs, specific bonding process parameter adjustment recommendations are made; these recommendations include adjusting the temperature, pressure, and time parameters during the bonding process to control the growth of IMCs and avoid excessive IMCs;
[0425] Implement bonding process optimization;
[0426] Implementing optimization of the bonding process based on the adjustment recommendations made in step S10.3; this involves adjusting the parameters of the bonding equipment in real time to ensure that the growth of the intermetallic compound is effectively controlled;
[0427] Setting up and implementing early warning systems;
[0428] Based on the evaluation results of the conductive performance, set the warning threshold; once the resistance or temperature exceeds the preset threshold, an alarm is immediately issued so that the bonding process can be adjusted in time or other measures can be taken; the implementation of the early warning system should include real-time monitoring and rapid response mechanisms;
[0429] The innovation of this technical solution is to combine the evaluation of conductive properties with the control of intermetallic compounds, and accurately control the growth of intermetallic compounds by real-time monitoring and analysis of solder joint temperature changes; this method not only improves the stability and reliability of the bonding process, but also can dynamically adjust the process parameters according to real-time data, ensuring the performance optimization of the bonding copper wire under various working conditions;
[0430] Analyze the influence of thermal stress on the electrical conductivity evaluation results;
[0431] According to the final conductivity evaluation result generated in step S9, the heat increase at the solder joint position is analyzed; the solder joint temperature change rate is calculated using the following formula :
[0432] ;
[0433] Parameter explanation,
[0434] is the current solder point temperature;
[0435] is the base temperature of the solder joint, usually the temperature under normal working conditions;
[0436] is the rate of change of solder joint temperature, expressed as a percentage;
[0437] Determine the amount of intermetallic compounds;
[0438] Based on the solder joint temperature change rate, combined with historical data and experimental results, the amount of intermetallic compounds formed between the pad and the bond is determined; if the temperature change rate exceeds a preset threshold, it indicates that the growth rate of the intermetallic compounds is accelerated;
[0439] Propose adjustments for intermetallic compound control;
[0440] Based on the determination of the number of IMCs, specific bonding process parameter adjustment recommendations are made; these recommendations include adjusting the temperature, pressure, and time parameters during the bonding process to control the growth of IMCs and avoid excessive IMCs;
[0441] Implement bonding process optimization;
[0442] Implementing optimization of the bonding process; this involves adjusting the parameters of the bonding equipment in real time to ensure that the growth of the intermetallic compounds is effectively controlled;
[0443] Setting up and implementing early warning systems;
[0444] Based on the evaluation results of the conductive performance, set the warning threshold; once the resistance or temperature exceeds the preset threshold, an alarm is immediately issued so that the bonding process can be adjusted in time or other measures can be taken; the implementation of the early warning system should include real-time monitoring and rapid response mechanisms;
[0445] The innovation of this technical solution is to combine the evaluation of conductive properties with the control of intermetallic compounds, and accurately control the growth of intermetallic compounds by real-time monitoring and analysis of solder joint temperature changes; this method not only improves the stability and reliability of the bonding process, but also can dynamically adjust the process parameters according to real-time data, ensuring the performance optimization of the bonding copper wire under various working conditions;
[0446] The quantitative implementation example is described in detail as follows:
[0447] Analyze the influence of thermal stress on the electrical conductivity evaluation results;
[0448] Example: Assume that in step S9, the conductivity evaluation result shows that the solder point temperature rises from 25°C under normal working conditions to 30°C; the temperature change rate is calculated using the above formula:
[0449] ;
[0450] Output: The solder point temperature change rate is 20%;
[0451] Determine the amount of intermetallic compounds;
[0452] Example: According to historical data, when the solder joint temperature change rate exceeds 15%, the amount of intermetallic compounds begins to increase significantly;
[0453] Output: It is determined that the amount of intermetallic compounds is increasing;
[0454] Propose adjustments for intermetallic compound control;
[0455] Based on a temperature change rate of 20%, the following adjustment suggestions are made:
[0456] Reduce the bonding temperature from 200°C to 180°C;
[0457] Reduce bonding pressure from 10N to 8N;
[0458] Shortened bonding time from 10 seconds to 8 seconds;
[0459] Output: Specific bonding process parameter adjustment suggestions;
[0460] Implement bonding process optimization;
[0461] Adjust bonding equipment parameters in real time based on adjustment suggestions:
[0462] Set the bonding temperature to 180°C;
[0463] Set the bonding pressure to 8N;
[0464] Set the bonding time to 8 seconds;
[0465] Output: Bonding process parameters are optimized;
[0466] Setting up and implementing early warning systems;
[0467] To set the warning threshold:
[0468] When the soldering point temperature exceeds 32°C, an over-temperature alarm is issued;
[0469] When the resistance value exceeds the normal value by 10%, an abnormal resistance alarm is issued;
[0470] Once the monitoring system detects any of the above conditions, an alarm will be immediately issued through the system, and the bonding process parameters will be automatically adjusted or the operator will be notified to take measures;
[0471] Output: The early warning system is set up and starts real-time monitoring;
[0472] Adjustment recommendations for intermetallic control;
[0473] Temperature control: Lowering the bonding temperature can reduce the rate of formation of intermetallic compounds. It is recommended to reduce the bonding temperature from 200°C to 180°C;
[0474] Pressure control: Reduce bonding pressure to reduce the pressure-induced factors of intermetallic compound formation. It is recommended to reduce the bonding pressure from 10N to 8N;
[0475] Time control: shorten the bonding time to reduce the time window for the formation of intermetallic compounds. It is recommended to shorten the bonding time from 10 seconds to 8 seconds;
[0476] Through these specific quantitative measures and adjustment suggestions, the growth of intermetallic compounds can be effectively controlled, the bonding process can be optimized, and the conductivity and reliability of the bonding copper wire can be ensured; these measures not only improve the precise control of the process, but also enhance the automation and intelligence level of the production process;
[0477] Embodiment 15:
[0478] See also Figure 3 , a bonding copper wire conductive performance analysis system based on a neural network, the system is used to execute the bonding copper wire conductive performance analysis method based on a neural network, comprising:
[0479] Monitoring and recording module: used to monitor the acoustic emission signal data of the bonding copper wire under all stresses in real time based on the acoustic emission sensor, and record the monitoring data;
[0480] Spectrum analysis module: used to apply fast Fourier transform to analyze the spectrum of acoustic emission signal data, extract key frequency components, and use this to determine the key stress area of the bonding copper wire, which includes mechanical stress, thermal stress and electrical stress;
[0481] The first index generation module is used to collect the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire in real time, analyze and process them, and generate a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2;
[0482] The second index generation module is used to collect the temperature change rate, thermal expansion coefficient and thermal conductivity of the thermal stress area of the bonding copper wire in real time, and analyze and process them to generate a second evaluation index, which is used to evaluate and adjust the corresponding characteristics of the thermal stress extracted in step S2;
[0483] The third index generation module is used to collect the electric field strength and dielectric constant changes of the bonding copper wire electric stress area in real time, and analyze and process them to generate a third evaluation index, which is used to evaluate and adjust the corresponding characteristics of the electric stress extracted in step S2;
[0484] Performance evaluation module: used to obtain the historical data of the conductive performance of the bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, build a training data set, and combine the extracted acoustic emission signal features with the known copper wire conductive performance data in the training data set to build a real-time performance evaluation model using a deep neural network. The model input is the signal feature and the output is the conductive performance prediction value;
[0485] Threshold setting module: used to obtain the historical data of conductivity performance, and based on the historical data, set the warning threshold of the conductivity performance of the bonding copper wire under different key stresses;
[0486] Threshold adjustment module: used to obtain the first evaluation index, the second evaluation index and the third evaluation index, and perform analysis and processing to construct a threshold adjustment model, which is used to dynamically adjust the warning threshold in the threshold setting module;
[0487] Evaluation result generation module: used to obtain the conductivity prediction value output by the real-time performance evaluation model in the performance evaluation module, and calculate the matching degree between the conductivity prediction value and the warning threshold dynamically adjusted by the threshold adjustment module to generate the final conductivity evaluation result of the bonding copper wire;
[0488] Adjustment suggestion module: used to obtain the final conductive performance evaluation results, determine the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and put forward adjustment suggestions for intermetallic compound control.
[0489] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.
[0490] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.
[0491] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0492] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.
Claims
1. A method for analyzing the conductive properties of bonding copper wire based on a neural network, characterized in that: The specific steps include: Step S1: Based on the acoustic emission sensor, real-time monitoring of acoustic emission signal data of the bonding copper wire under all stresses is performed, and the monitoring data is recorded; Step S2, applying fast Fourier transform to analyze the spectrum of the acoustic emission signal data, extracting key frequency components, and using this to determine the key stress area of the bonding copper wire, the key stress includes mechanical stress, thermal stress and electrical stress; Step S3, collecting the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire in real time, analyzing and processing them, and generating a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2; Step S4, collecting the temperature change rate, thermal expansion coefficient and thermal conductivity of the thermal stress area of the bonding copper wire in real time, and analyzing and processing them to generate a second evaluation index, which is used to evaluate and adjust the corresponding characteristics of the thermal stress extracted in step S2; Step S5, collecting the electric field strength and dielectric constant change of the bonding copper wire electric stress area in real time, analyzing and processing, and generating a third evaluation index, which is used to evaluate and adjust the corresponding characteristics of the electric stress extracted in step S2; Step S6, obtaining the historical data of the conductive performance of the bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, constructing a training data set, and combining the extracted acoustic emission signal features with the known conductive performance data of the copper wire in the training data set, using a deep neural network to construct a real-time performance evaluation model, the input of the model is the signal feature, and the output is the conductive performance prediction value; Step S7, obtaining historical data of electrical conductivity, and setting warning thresholds of electrical conductivity of bonding copper wires under different key stresses based on the historical data; Step S8, obtaining the first evaluation index, the second evaluation index and the third evaluation index, and performing analysis and processing to construct a threshold adjustment model, which is used to dynamically adjust the warning threshold in step S7; Step S9, obtaining the conductivity prediction value output by the real-time performance evaluation model in step S6, and calculating the matching degree between the conductivity prediction value and the warning threshold dynamically adjusted in step S8, to generate the final conductivity evaluation result of the bonding copper wire; Step S10, obtaining the final conductive performance evaluation result, determining the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and proposing adjustment suggestions for the control of the intermetallic compounds.
2. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 1, characterized in that: The application of fast Fourier transform to analyze the spectrum of acoustic emission signal data and extract key frequency components specifically includes the following contents: Performing spectrum analysis on the acoustic emission signal data collected in step S1 using fast Fourier transform; By writing FFT analysis module in LabVIEW environment, the signal is preprocessed and transformed; Extract key frequency components and use threshold detection algorithm to determine the frequencies associated with copper wire stress; The threshold detection algorithm is specifically to calculate the amplitude value of each frequency component in the FFT result; set the threshold to 5% of the signal peak value, that is, , where Threshold is the threshold used to determine whether a frequency component is a critical frequency; Max (Amplitude) is the maximum value of all frequency component amplitude values; The amplitude value of each frequency component is compared with a threshold value. If the amplitude value is greater than the threshold value, the frequency component is considered to be a key frequency component.
3. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 2, characterized in that: The real-time collection of the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire and the analysis and processing are performed to generate a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2, and specifically includes the following contents: The vibration frequency is marked as ZDf, and the magnetic parameter is marked as CXc; The mathematical formula for defining the first evaluation index is as follows: ; Among them, Index is the first evaluation index, which is used to evaluate the influence of mechanical stress, and its value range is set to (0,1); when Index is closer to 1, it means that the influence of mechanical stress is more significant; when it is closer to 0, it means that the influence is smaller; ZDf is the vibration frequency value, which reflects the dynamic response of the copper wire under mechanical stress; CXc is the magnetic parameter value, which evaluates the influence of mechanical stress through the change of magnetic permeability; is a positive weight coefficient; is the amplitude value of the jth frequency component, M1 is the total number of frequency components; It is the maximum value of all frequency component amplitudes.
4. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 3, characterized in that: The value range (0,1) of the first evaluation index Index is divided into the following three intervals: Interval 1, ; In this range, the mechanical stress has a slight effect on the copper wire, and the Index value indicates a low stress level; at this time, ZDf, CXc and The changes are small and the system is in slight monitoring mode; Set the interval 1 threshold to Index=0.2; when Index<0.2, the monitoring frequency f is once per hour, that is, f=1 time / hour; when When f=2 times / hour, the mathematical expression is: ; In this interval, the increase in ZDf leads to an increase in the Index, and the change in the Index is No more than 5%; the change in CXc leads to the change in Index No more than 3%; The change in the index causes the change in No more than 2%; the mathematical expression is: ; Interval 2, ; In this range, the mechanical stress has a moderate effect on the copper wire, and the Index value indicates that the stress level is moderate; ZDf, CXc, and The changes begin to be significant, and the system starts the medium stress response mode, increases the monitoring frequency, and records and analyzes the stress change trend; Set the threshold Index = 0.5; when Index>0.5, the monitoring frequency f is once every 15 minutes, that is, f = 4 times / hour; when When f=2 times / hour, the mathematical expression is: ; In this interval, the increase in ZDf leads to an increase in Index, and the change range is No more than 10%; the change in CXc leads to the change in Index Not more than 8%; The change in the index causes the change in No more than 5%; the mathematical expression is: ; Interval three, ; In this range, the mechanical stress has a significant effect on the copper wire, and the Index value is close to 1; ZDf, CXc and When the changes in the values of the conductive materials reach a high level, the system activates the high-stress emergency response mode, immediately conducts a comprehensive inspection and adjusts the production parameters to prevent the conductive performance from being damaged; Set the threshold Index=0.85; when Index>0.85, the system executes the emergency shutdown procedure; when When the monitoring frequency f is once every 5 minutes, that is, f = 12 times / hour; the mathematical expression is: ; In this interval, the increase in ZDf leads to an increase in Index, and the change range is No more than 20%; the change in CXc leads to the change in Index Not more than 15%; The change in the index causes the change in Not more than 10%; The mathematical expression is: 。 5. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 4, characterized in that: The temperature change rate, thermal expansion coefficient and thermal conductivity are marked as ΔT, CTE and kr in sequence, forming the temperature change rate ΔT, thermal expansion coefficient CTE and thermal conductivity kr; The second evaluation index is defined as E2, and the calculation formula is: ; The parameters are explained as follows: It is The temperature change rate at each time point; is the maximum value of the temperature change rate at all time points; It is The coefficient of thermal expansion at a time point; is the average value of the coefficient of thermal expansion at all time points; is the standard deviation of the coefficient of thermal expansion; It is Thermal conductivity at each time point; is the minimum value of thermal conductivity at all time points; is the total number of time points; the value range of the second evaluation index E2 is set to (0,1); when the index is closer to 1, it means that the thermal stress characteristics are better, and the electrical conductivity of the copper wire is less affected by the thermal stress; when the index is closer to 0, it means that the thermal stress characteristics are worse, and the electrical conductivity of the copper wire is more affected by the thermal stress.
6. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 5, characterized in that: The value range (0,1) of the second evaluation index E2 is divided into three intervals, namely interval I (0, 0.42), interval II [0.42, 0.72), and interval III [0.72, 1); Interval I (0, 0.42), in interval I, the E2 value is less than 0.42, the thermal stress characteristics of the copper wire are poor, and the conductivity is greatly affected. At this time, ΔT exceeds the average value by 50%, the standard deviation of CTE deviates from the average value by more than 30%, and kr is lower than the average value by 20%. The system needs to strengthen cooling measures to reduce the impact of thermal stress and ensure stable electrical connections; When the E2 value is less than 0.42, the thermal stress adjustment mechanism is activated; the threshold is set to 0.25, and when the E2 value is lower than this threshold, the system automatically increases cooling measures to reduce the temperature change rate and thermal expansion coefficient; if the E2 value is between 0.25 and 0.42, the intensity of the cooling measures is adjusted to gradually reduce the impact of 5% thermal stress on the copper wire; Mathematical expression of interaction rules: ; ; ; Parameter explanation, It is the percentage of increase in E2 value for every 10% increase in temperature change rate ΔT; It is the percentage by which the E2 value decreases when the coefficient of thermal expansion CTE deviates from the average value by 10%; It is the percentage of decrease in E2 value for every 10% decrease in thermal conductivity kr; They represent the percentage changes of ΔT, CTE, and kr relative to their initial values, respectively; In interval I, when ΔT increases, the E2 value increases, but the decrease of CTE and kr causes the E2 value to decrease; Interval II [0.42, 0.72), E2 value is between 0.42 and 0.72, the thermal stress characteristics of the copper wire are moderate, the conductivity is relatively stable, the system maintains the current cooling measures and adjusts according to the E2 value to optimize the thermal stress state; When the E2 value is between 0.42 and 0.72, the current thermal stress management strategy is maintained; the threshold is set to 0.5, and when the E2 value is lower than this threshold, the system maintains the current cooling measures unchanged; when the E2 value is higher than this threshold, the system gradually reduces the cooling measures by 10% to optimize the thermal stress state of the copper wire; Mathematical expression of interaction rules: ; ; ; Parameter explanation: It is the percentage change of E2 value for every 5% change of temperature change rate ΔT; It is the percentage change of E2 value for every 5% change of coefficient of thermal expansion CTE; It is the percentage change of E2 value for every 5% change of thermal conductivity kr; They represent the percentage changes of ΔT, CTE, and kr relative to their initial values, respectively; In interval II, a small change in ΔT has little effect on the E2 value, but changes in CTE and kr will significantly affect the E2 value; Interval III [0.72, 1), in interval III, the E2 value is greater than 0.72, the thermal stress characteristics of the copper wire are good, the conductivity is slightly affected, and the system gradually reduces the cooling measures by 12% until it stops to maintain the optimal thermal stress state of the copper wire; When the E2 value is greater than 0.72, the thermal stress management measures are reduced; the threshold is set to 0.75, and when the E2 value is higher than this threshold, the system gradually reduces the cooling measures by 14% until it stops completely; when the E2 value is between 0.72 and 0.75, the system maintains the current cooling measures to maintain the thermal stress state of the copper wire; Mathematical expression of interaction rules: ; Parameter explanation: It is the percentage of decrease of E2 value when the temperature change rate ΔT decreases by 5%; It is the percentage of increase in E2 value for every 5% decrease in the average value of the coefficient of thermal expansion CTE; It is the percentage of increase in E2 value for every 5% increase in thermal conductivity kr; They represent the percentage changes of ΔT, CTE, and kr relative to their initial values, respectively; In interval III, when ΔT decreases, the E2 value decreases, but the increase in CTE and kr causes the E2 value to increase.
7. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 6, characterized in that: The electric field strength and the dielectric constant change are marked as Eq and Δε respectively, forming the electric field strength Eq and the dielectric constant change Δε; The third evaluation index is defined as E3, and the calculation formula is as follows: ; Parameter explanation: E3 is the third evaluation index, with a value range of (0,1), which represents the evaluation of the influence of electrical stress on the conductive properties of copper wire; Eq is the electric field intensity, reflecting the stress state of the copper wire in the electric field; Δε is the percentage change in dielectric constant, reflecting the change in polarization characteristics of the material under electrical stress; c1 is the adjustment coefficient, which is used to adjust the influence of electric field intensity on the evaluation index; c2 is the adjustment coefficient, which is used to adjust the influence of the change of dielectric constant on the evaluation index; A positive constant term used to ensure that the value of E3 will not be 0; When E3 is closer to 1, it means that the electrical stress has a greater impact on the conductivity of the copper wire, and the conductivity of the copper wire is worse; When E3 is closer to 0, it means that the influence of electrical stress on the conductivity of the copper wire is smaller, and the conductivity of the copper wire is better.
8. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 7, characterized in that: The third evaluation index E3 value range (0,1) is divided into the following intervals, and the evaluation adjustment description is given for each interval; Interval 1 (0.00, 0.15]; within this interval, the electrical stress affects the copper wire conductivity by less than 2%, the electric field strength is 25% lower than the average value, and the dielectric constant changes by less than 3%; the electric field strength and dielectric constant must be kept within a small change of 2% to maintain the best performance; In this range, the threshold is set to 0.10 to determine the degree of optimization of the copper wire's conductive properties. When the E3 value is lower than 0.10, the electric field strength is kept below 20% of the average value, and the dielectric constant change is controlled within 2%. If the E3 value is higher than 0.10, the electric field strength and dielectric constant need to be further adjusted by 3% to reduce the impact of electrical stress on conductive properties. Mathematical expression of interaction rules: , ; Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3; When Eq increases by 10%, E3 increases by 1%; when Δε increases by 5%, E3 increases by 0.5%; other parameters remain unchanged; Interval 2 (0.15, 0.35]; in this interval, the electrical conductivity of the copper wire begins to be affected by the electrical stress, the electric field strength is in the range of 25% to 45% of the average value, and the dielectric constant changes in the range of 3% to 7%; In this range, the threshold is set to 0.25 to evaluate the stability of the copper wire's conductive properties. When the E3 value is lower than 0.25, the copper wire's conductive properties are stable, and the dielectric constant change is controlled within 5%. If the E3 value is higher than 0.25, the electric field strength needs to be adjusted by 4% and the dielectric constant by 3% to reduce the impact of electrical stress on conductive properties. Mathematical expression of interaction rules: , ; Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3; When Eq increases by 10%, E3 increases by 2%; when Δε increases by 5%, E3 increases by 1.5%; other parameters remain unchanged; Interval 3 (0.35, 0.60]; the conductivity of the copper wire is affected by moderate electrical stress, the electric field strength is in the range of 45% to 65% of the average value, and the dielectric constant changes in the range of 7% to 12%; 2% fine adjustment of the electric field strength and dielectric constant is required to maintain stable conductivity; In this range, the threshold is set to 0.48 to evaluate the balance state of the copper wire's conductivity. When the E3 value is lower than 0.48, the copper wire's conductivity is low. At this time, the electric field strength is kept below 55% of the average value, and the dielectric constant change is controlled within 10%. If the E3 value is higher than 0.48, it indicates that the electric stress has a greater impact, and the adjustment frequency of the electric field strength and dielectric constant needs to be increased by 8%. Mathematical expression of interaction rules: , ; Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3; When Eq increases by 10%, E3 increases by 3%; when Δε increases by 5%, E3 increases by 2%; other parameters remain unchanged. Interval 4 (0.60, 1.00); in this interval, the electrical conductivity of the copper wire is significantly affected by electrical stress, the electric field strength is in the range of 65% to 85% of the average value, and the dielectric constant changes in the range of 12% to 18%. It is necessary to increase the adjustment frequency of the electric field strength and dielectric constant by 10% to reduce the loss of electrical conductivity; In this range, the threshold is set to 0.73 to evaluate the degree of damage to the copper wire's conductive properties. When the E3 value is lower than 0.73, the electric field strength is kept below 75% of the average value, and the dielectric constant change is controlled within 15%. If the E3 value is higher than 0.73, it indicates that the electric stress is too large, and the electric field strength and dielectric constant need to be adjusted immediately to avoid further deterioration of the conductive properties. Mathematical expression of interaction rules: , ; Parameter explanation: It indicates the effect of every 10% increase in electric field strength Eq on E3, It indicates the effect of every 5% increase in dielectric constant change Δε on E3; When Eq increases by 10%, E3 increases by 4%; when Δε increases by 5%, E3 increases by 3%; other parameters remain unchanged; The training data set of the deep neural network model includes historical acoustic emission signal data and corresponding bonding copper wire conductive performance data; The obtaining of the historical data of the electrical conductivity and setting the warning threshold of the electrical conductivity of the bonding copper wire under different key stresses based on the historical data specifically includes the following contents: Collect historical data on the conductivity of copper bonding wires under different mechanical, thermal and electrical stress conditions. Verify the collected historical data to ensure the accuracy and reliability of the data; Calculate average conductivity based on historical data and standard deviation ; The basic warning threshold is calculated using the following formula : ; Among them, k3 is a constant used to adjust the sensitivity of the threshold.
9. The method for analyzing the conductive properties of bonding copper wires based on a neural network according to claim 8, characterized in that: The first evaluation index, the second evaluation index and the third evaluation index are obtained, analyzed and processed, and a threshold adjustment model is constructed. The model is used to dynamically adjust the warning threshold in step S7, and specifically includes the following contents: The threshold adjustment model formula is: ; Parameter explanation: ATd is the adjusted warning threshold, which is used to dynamically adjust the conductive performance warning threshold of the bonding copper wire; is the basic warning threshold, calculated from the average conductivity and standard deviation; is the adjustment factor used to fine-tune the sensitivity of the threshold; is the sensitivity factor, reflecting the influence of k3 on the threshold change; is an adjustment coefficient, which is used to adjust the influence of the first evaluation index, the second evaluation index and the third evaluation index on the threshold adjustment; Index is the first evaluation index, reflecting the influence of mechanical stress; E2 is the second evaluation index, reflecting the influence of thermal stress; E3 is the third evaluation index, reflecting the influence of electrical stress; Set the range of ATd to ; When ATd is closer to 0, it means that the current stress state has less influence on the conductivity; As ATd approaches When , it means that the stress state has a more significant impact on the conductivity, and corresponding measures need to be taken; ATd value range Divide into the following intervals, and give corresponding descriptions for each interval; Interval 1 ; Set the ATd middle threshold to ; When ATd is lower than this value, the system enters low-risk monitoring mode and increases the frequency of data collection to once every 30 minutes; if ATd exceeds this value but is still lower than the upper limit of the interval, the early warning system is activated to prompt the operator to perform manual inspection; In this range, an increase in Index will lead to an increase of 5% in ATd, while an increase in E2 and E3 will lead to an increase of 3% and 2% in ATd, respectively. If Index decreases, ATd will decrease by 7%, while a decrease in E2 and E3 will lead to a decrease of 4% and 3% in ATd, respectively. ; Parameter explanation, is the change in the adjusted warning threshold; is the change in Index; is the change in E2; is the change in E3; Interval 2 ; In this interval, set the middle threshold to ; When ATd is lower than this value, the system maintains normal monitoring mode, but increases the data analysis depth by 5%; In this range, an increase in Index will lead to a 10% increase in ATd, while an increase in E2 and E3 will lead to an increase in ATd of 7% and 5%, respectively. If Index decreases, ATd will decrease by 12%, while a decrease in E2 and E3 will lead to a decrease in ATd of 8% and 6%, respectively. ; Parameter explanation, is the change in the adjusted warning threshold; It is the change of Index, reflecting the change of mechanical stress; is the change in E2, reflecting the change in thermal stress; is the change of E3, reflecting the change of electrical stress; Interval 3 ; In this interval, set the middle threshold to ; When ATd is lower than this value, the system enters high-risk monitoring mode and activates all available resources for data analysis; If ATd exceeds this value, immediately notify the highest level of management and prepare for emergency repairs or component replacement; In this range, an increase in Index will lead to an increase in ATd of 15%, while increases in E2 and E3 will lead to increases in ATd of 10% and 8%, respectively. If Index decreases, ATd will decrease by 20%, while decreases in E2 and E3 will lead to decreases in ATd of 12% and 10%, respectively. ; Parameter explanation, is the change in the adjusted warning threshold; It is the change of Index, reflecting the change of mechanical stress; is the change in E2, reflecting the change in thermal stress; It is the change of E3, reflecting the change of electrical stress.
10. A bonding copper wire conductivity analysis system based on neural network, characterized in that: The system is used to execute the method for analyzing the conductive properties of bonding copper wires based on a neural network according to any one of claims 1 to 9, comprising: Monitoring and recording module: used to monitor the acoustic emission signal data of the bonding copper wire under all stresses in real time based on the acoustic emission sensor, and record the monitoring data; Spectrum analysis module: used to apply fast Fourier transform to analyze the spectrum of acoustic emission signal data, extract key frequency components, and use this to determine the key stress area of the bonding copper wire, which includes mechanical stress, thermal stress and electrical stress; The first index generation module is used to collect the vibration frequency and magnetic parameters of the mechanical stress area of the bonding copper wire in real time, analyze and process them, and generate a first evaluation index, which is used to evaluate and adjust the corresponding characteristics of the mechanical stress extracted in step S2; The second index generation module is used to collect the temperature change rate, thermal expansion coefficient and thermal conductivity of the thermal stress area of the bonding copper wire in real time, and analyze and process them to generate a second evaluation index, which is used to evaluate and adjust the corresponding characteristics of the thermal stress extracted in step S2; The third index generation module is used to collect the electric field strength and dielectric constant changes of the bonding copper wire electric stress area in real time, and analyze and process them to generate a third evaluation index, which is used to evaluate and adjust the corresponding characteristics of the electric stress extracted in step S2; Performance evaluation module: used to obtain the historical data of the conductive performance of the bonding copper wire under different mechanical stress, thermal stress and electrical stress conditions, build a training data set, and combine the extracted acoustic emission signal features with the known copper wire conductive performance data in the training data set to build a real-time performance evaluation model using a deep neural network. The model input is the signal feature and the output is the conductive performance prediction value; Threshold setting module: used to obtain the historical data of conductivity performance, and based on the historical data, set the warning threshold of the conductivity performance of the bonding copper wire under different key stresses; Threshold adjustment module: used to obtain the first evaluation index, the second evaluation index and the third evaluation index, and perform analysis and processing to construct a threshold adjustment model, which is used to dynamically adjust the warning threshold in the threshold setting module; Evaluation result generation module: used to obtain the conductivity prediction value output by the real-time performance evaluation model in the performance evaluation module, and calculate the matching degree between the conductivity prediction value and the warning threshold dynamically adjusted by the threshold adjustment module to generate the final conductivity evaluation result of the bonding copper wire; Adjustment suggestion module: used to obtain the final conductive performance evaluation results, determine the amount of intermetallic compounds formed between the pad and the bond during the bonding process, and put forward adjustment suggestions for intermetallic compound control.
