Process optimization method and system based on bonding silver wire ultrasonic welding signal
By analyzing the ultrasonic welding signals of the bonding silver wires and optimizing the process parameters, the inefficiency and low cost caused by the reliance on experience in the traditional bonding silver wire process were solved, thereby improving the reliability of the solder joint quality and increasing production efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN ZHONGBAO NEW MATERIAL TECH CO LTD
- Filing Date
- 2025-12-25
- Publication Date
- 2026-04-21
AI Technical Summary
In existing technologies, the optimization of bonding silver wire processes relies on experience and trial and error, which is inefficient, costly, and makes it difficult to detect microscopic defects or metastable issues during the welding process, resulting in poor process optimization effects.
By extracting ultrasonic welding signals, analyzing energy entropy characteristics, dynamic stiffness coefficient, energy loss rate, and weld joint micro-deformation gradient, and combining multi-objective optimization functions and optimization constraints, the process parameters for bonding silver wires are optimized.
This has improved the reliability and consistency of weld quality, reduced the defect rate of welds, increased production efficiency and reduced costs, and promoted the intelligent upgrading of welding technology.
Smart Images
Figure CN121898516A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a process optimization method and system based on ultrasonic welding signals of bonded silver wires, belonging to the field of data analysis technology. Background Technology
[0002] Silver wire bonding process optimization refers to the systematic improvement and adjustment of the silver wire bonding process used in semiconductor packaging to enhance its performance, reliability, and production efficiency. Silver wire bonding process optimization is a multi-dimensional and systematic project aimed at overcoming the challenges that silver wire bonding may present (such as being more brittle than gold wire, being prone to oxidation, and having electromigration risks), fully leveraging its cost advantages, and making it a stable and reliable option for high-performance and high-reliability semiconductor packaging.
[0003] Traditional methods for optimizing silver wire bonding processes rely heavily on experience and trial and error. This involves setting parameters based on experience, conducting offline testing (such as tensile strength and visual inspection) on welded samples, and repeatedly adjusting the parameters until the requirements are met. This approach is inefficient, costly, and lacks real-time understanding of the internal dynamics of the welding process, making it difficult to detect potential microscopic defects or metastable issues, thus resulting in poor optimization of the silver wire bonding process. Summary of the Invention
[0004] This invention provides a process optimization method and system based on ultrasonic welding signals of bonded silver wires, the main purpose of which is to improve the process optimization effect of bonded silver wires.
[0005] To achieve the above objectives, the present invention provides a process optimization method based on ultrasonic welding signals of bonded silver wires, comprising: The ultrasonic welding signal of the bonded silver wire is extracted, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence; Based on the ultrasonic amplitude signal, the energy entropy characteristics of the bonded silver wire are analyzed, and the dynamic stiffness coefficient of the bonded silver wire is analyzed based on the welding pressure signal. Using the transducer current signal, a nonlinear impedance model of the bonding silver wire is established to analyze the energy loss rate of the bonding silver wire. Based on the solder joint deformation image sequence, the micro-deformation gradient of the solder joint of the bonding silver wire is analyzed. Based on the energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and micro-deformation gradient of the solder joint, the solder joint failure coefficient and optimization constraints of the bonding silver wire are analyzed. By combining the preset multi-objective optimization function, the solder joint failure coefficient, and the optimization constraints, the process parameters of the bonding silver wire are optimized to obtain optimized process parameters, and the process optimization of the bonding silver wire is performed based on the optimized process parameters.
[0006] Optionally, the extraction of the ultrasonic welding signal of the bonded silver wire includes: The signal acquisition sensor configured with the bonded silver wire includes a high-frequency laser vibrometer, a miniature dynamic force sensor, a high-frequency current probe, and a high-speed camera. Construct the sensing topology network of the signal acquisition sensor; Based on the aforementioned sensing topology network, configure the synchronization pulse signal of the signal acquisition sensor; Based on the synchronous pulse signal, the ultrasonic welding signal of the bonded silver wire is extracted, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence.
[0007] Optionally, the step of analyzing the energy entropy characteristics of the bonded silver wire based on the ultrasonic amplitude signal includes: The ultrasonic amplitude signal is subjected to wavelet denoising to obtain a denoised ultrasonic amplitude signal; Calculate the analytical signal of the noise-reduced ultrasonic amplitude signal; Extract the envelope of the analyzed signal; Calculate the energy percentage of the equal-width energy range corresponding to the envelope; Based on the energy ratio, the energy entropy of the noise-reduced ultrasonic amplitude signal is calculated to construct the time-varying entropy curve of the noise-reduced ultrasonic amplitude signal; The energy entropy characteristics of the bonded silver wire are analyzed based on the time-varying entropy curve.
[0008] Optionally, extracting the envelope of the parsed signal includes: The instantaneous amplitude of the analytic signal is calculated using the following formula: ; in, Indicates analytic signal Instantaneous amplitude at time, Indicates an analytical signal. This indicates that the analytical signal corresponds to the noise-reduced ultrasonic amplitude signal. Indicates the noise-reducing ultrasonic amplitude signal Hilbert transform, Represents the imaginary unit. This indicates the modulo operation of a complex number; Based on the instantaneous amplitude, the instantaneous amplitude of the analytical signal is extracted.
[0009] Optionally, analyzing the dynamic stiffness coefficient of the bonding silver wire based on the welding pressure signal includes: The welding pressure signal is baseline-corrected to obtain a corrected welding pressure signal. Identify the key feature points of the corrective welding pressure signal; Based on the aforementioned key feature points, a pressure-displacement relationship model for the bonded silver wire is established. Based on the pressure-displacement relationship model, the frequency domain stiffness coefficient and time domain stiffness coefficient of the bonded silver wire are extracted. The dynamic stiffness coefficient of the bonded silver wire is determined by combining the frequency domain stiffness coefficient and the time domain stiffness coefficient.
[0010] Optionally, establishing a nonlinear impedance model of the bonded silver wire using the transducer current signal includes: Calculate the instantaneous power and resonant frequency offset of the transducer current signal; Based on the instantaneous power and resonant frequency offset, an equivalent circuit for the bonding silver wire is established. Identify the impedance parameters of the equivalent circuit; Based on the impedance parameters, a nonlinear impedance model of the bonded silver wire is established.
[0011] Optionally, the analysis of the energy loss rate of the bonded silver wire includes: Based on the nonlinear impedance model of the bonded silver wire, the dynamic impedance of the bonding interface of the bonded silver wire is calculated. Based on the dynamic impedance of the bonding interface, the energy loss rate of the bonding silver wire is calculated using the following formula: ; in, Indicates bonding silver wire Energy loss rate at any given time This represents the linear weighting coefficient for loss. Indicates the dynamic impedance of the bonding interface. This represents the secondary loss coefficient. This indicates the resonant frequency shift of the bonded silver wire. This represents the loss coefficient due to the rate of change of frequency. This represents the rate of change of the absolute value of the frequency offset. This represents a differential operator.
[0012] Optionally, the step of analyzing the microscopic deformation gradient of the bonding silver wires based on the solder joint deformation image sequence includes: The weld joint deformation image sequence is filtered to obtain a filtered weld joint deformation image; Locate the solder joint region in the filtered solder joint deformation image; Calculate the displacement field of the solder joint region to calculate the Jacobian matrix of the bonding silver wire; The microscopic deformation gradient of the bonding silver wire joints is analyzed based on the Jacobian matrix.
[0013] Optionally, the step of combining the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient, and the micro-deformation gradient of the solder joint to analyze the solder joint failure coefficient and optimization constraints of the bonding silver wire includes: The energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and weld joint micro-deformation gradient are analyzed to determine the weld joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and weld joint deformation failure coefficient. The failure coefficient of the bonding silver wire is analyzed by combining the energy entropy solder joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and solder joint deformation failure coefficient. Based on the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient, and the micro-deformation gradient of the solder joint, the single-feature constraints and coupling constraints of the bonding silver wire are determined. Based on the single feature constraint and coupling constraint, the optimal constraint conditions for the bonded silver wire are determined.
[0014] To address the aforementioned problems, the present invention also provides a process optimization system based on ultrasonic welding signals of bonded silver wires, the system comprising: The welding signal acquisition module is used to extract the ultrasonic welding signal of the bonded silver wire, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence; The dynamic stiffness analysis module is used to analyze the energy entropy characteristics of the bonded silver wire based on the ultrasonic amplitude signal, and to analyze the dynamic stiffness coefficient of the bonded silver wire based on the welding pressure signal. The deformation gradient analysis module is used to establish a nonlinear impedance model of the bonding silver wire using the transducer current signal, so as to analyze the energy loss rate of the bonding silver wire and analyze the micro-deformation gradient of the bonding silver wire based on the solder joint deformation image sequence. The solder joint failure analysis module is used to analyze the solder joint failure coefficient and optimization constraints of the bonded silver wire by combining the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient and the solder joint micro-deformation gradient. The bonding process optimization module is used to optimize the process parameters of the bonding silver wire by combining a preset multi-objective optimization function, the solder joint failure coefficient and the optimization constraints, to obtain optimized process parameters, and to perform process optimization of the bonding silver wire based on the optimized process parameters.
[0015] First, this method significantly improves the reliability and consistency of weld quality. Traditional welding processes often rely on experience or static parameter settings, making it difficult to cope with uncertainties caused by material differences and environmental fluctuations. This method, however, monitors ultrasonic amplitude, welding pressure, transducer current, and weld deformation in real time, and analyzes weld failure risks using advanced indicators such as energy entropy, energy loss rate, dynamic stiffness coefficient, and micro-deformation gradient. It can accurately identify potential factors that may lead to weld failure, making process optimization no longer a "black box" operation but a precise control based on scientific data. This significantly reduces the weld defect rate and ensures stable and reliable product performance. Second, this method achieves dynamic optimization and adaptive control of welding process parameters. By establishing a nonlinear impedance model of the transducer current signal to analyze energy loss, and combining other characteristics to analyze the failure coefficient, the system can automatically calculate the optimal combination of process parameters based on a preset multi-objective optimization function (which may simultaneously consider quality, efficiency, cost, etc.) and optimization constraints. The welding process can be dynamically adjusted based on real-time signal feedback. For example, when abnormal energy loss or excessive deformation gradient is detected, the welding power or pressure is automatically fine-tuned, avoiding over-welding or under-welding problems that may occur under traditional fixed parameters. This achieves more refined process control. Furthermore, this method improves production efficiency and reduces manufacturing costs. Intelligent process optimization reduces trial-and-error costs and rework and scrap losses caused by poor welding. Optimized process parameters can usually shorten welding time or reduce energy consumption while ensuring quality, thereby increasing production cycle time and saving energy. In addition, comprehensive monitoring and data analysis of the welding process provides valuable data support for equipment maintenance and process improvement, helping to achieve preventive maintenance and continuous improvement. Finally, this method promotes the intelligent upgrading of welding technology. It combines advanced signal processing, machine learning, and multi-objective optimization technologies with traditional welding processes, not only improving the intelligence level of individual processes but also laying the foundation for the digital and intelligent transformation of the entire manufacturing process. Therefore, this invention can improve the process optimization effect of bonding silver wires. Attached Figure Description
[0016] Figure 1 This is a flowchart illustrating a process optimization method based on ultrasonic welding signals of bonded silver wires according to an embodiment of the present invention. Figure 2 A schematic diagram of stiffness coefficient analysis for implementing the process optimization method based on ultrasonic welding signal of bonded silver wire according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a module for implementing the process optimization method based on ultrasonic welding signals of bonded silver wires, according to an embodiment of the present invention.
[0017] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0019] This application provides a process optimization method based on ultrasonic welding signals of bonded silver wires. The execution subject of this process optimization method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the process optimization method based on ultrasonic welding signals of bonded silver wires can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.
[0020] Example 1: Reference Figure 1 The diagram shown is a flowchart illustrating a process optimization method based on ultrasonic welding signals of bonded silver wires according to an embodiment of the present invention. In this embodiment, the process optimization method based on ultrasonic welding signals of bonded silver wires includes: S1. Extract the ultrasonic welding signal of the bonded silver wire, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence.
[0021] This invention extracts ultrasonic welding signals from bonded silver wires to obtain highly reliable multimodal welding process signals, laying a data foundation for subsequent feature extraction and process optimization.
[0022] Specifically, the extraction of ultrasonic welding signals from the bonded silver wires includes: The signal acquisition sensor configured with the bonded silver wire includes a high-frequency laser vibrometer, a miniature dynamic force sensor, a high-frequency current probe, and a high-speed camera. Construct the sensing topology network of the signal acquisition sensor; Based on the aforementioned sensing topology network, configure the synchronization pulse signal of the signal acquisition sensor; Based on the synchronous pulse signal, the ultrasonic welding signal of the bonded silver wire is extracted, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence.
[0023] The high-frequency laser vibrometer refers to a non-contact optical sensor that uses the laser Doppler effect to accurately measure the vibration velocity, displacement, or acceleration of an object's surface. The miniature dynamic force sensor refers to a sensor capable of accurately measuring dynamically changing forces. The high-frequency current probe refers to a sensor used for non-invasive measurement of high-frequency current in wires. The high-speed camera refers to a camera with an extremely high frame rate (typically far exceeding the 25 / 30 of ordinary video). The image acquisition device (fps, reaching thousands or even tens of thousands of frames per second), the sensing topology network refers to the overall architecture of various sensors, data acquisition devices, trigger sources and their physical and logical connections, the synchronization pulse signal refers to a precise, time-stamped digital signal (usually a pulse or level change) used to trigger all sensors and data acquisition systems to start data acquisition at the same time or according to a predetermined relationship, the ultrasonic amplitude signal refers to the high-frequency waveform extracted from the laser vibrometer data acquired by DAQ, the welding pressure signal refers to the dynamic pressure curve extracted from the data of the miniature dynamic force sensor, the transducer current signal refers to the high-frequency AC signal extracted from the high-frequency current probe data, and the weld joint deformation image sequence refers to the image frames acquired by the high-speed camera and arranged in chronological order, which intuitively shows the dynamic changes of the weld joint area such as deformation, melting, and solidification during the welding process.
[0024] Optionally, the synchronization pulse signal configured for the signal acquisition sensor can be generated by an FPGA.
[0025] S2. Based on the ultrasonic amplitude signal, analyze the energy entropy characteristics of the bonded silver wire, and based on the welding pressure signal, analyze the dynamic stiffness coefficient of the bonded silver wire.
[0026] Based on the ultrasonic amplitude signal, this invention analyzes the energy entropy characteristics of the bonded silver wire, which can serve as a core feature for evaluating welding quality and provide a quantitative basis for process optimization.
[0027] In detail, the analysis of the energy entropy characteristics of the bonded silver wire based on the ultrasonic amplitude signal includes: The ultrasonic amplitude signal is subjected to wavelet denoising to obtain a denoised ultrasonic amplitude signal; Calculate the analytical signal of the noise-reduced ultrasonic amplitude signal; Extract the envelope of the analyzed signal; Calculate the energy percentage of the equal-width energy range corresponding to the envelope; Based on the energy ratio, the energy entropy of the noise-reduced ultrasonic amplitude signal is calculated to construct the time-varying entropy curve of the noise-reduced ultrasonic amplitude signal; The energy entropy characteristics of the bonded silver wire are analyzed based on the time-varying entropy curve.
[0028] The denoised ultrasonic amplitude signal refers to the new signal obtained by applying wavelet denoising technology to the original ultrasonic amplitude signal (measured by a high-frequency laser vibrometer). The analytic signal refers to a complex-valued signal constructed from the denoised ultrasonic amplitude signal through Hilbert transform. The envelope refers to the modulus of the complex signal in the complex plane, which is composed of the real part (original signal) and the imaginary part (Hilbert transform) of the analytic signal. The equal-width energy interval refers to dividing the amplitude range covered by the envelope (e.g., from 0 to the maximum value A_max of the envelope) into several intervals evenly within the entire analysis time window. The energy proportion refers to the proportion of energy contained in the signal to the total energy within a specific equal-width energy interval. The energy entropy refers to a metric calculated based on the energy proportion (considered as a probability distribution). The time-varying entropy curve refers to a curve showing the change of a series of energy entropy values over time by sliding a time window (short-time analysis) on the signal and repeatedly calculating the energy entropy at each window position. The energy entropy feature refers to a representative statistic or pattern extracted from the time-varying entropy curve. For example, it may include the average energy entropy of the entire welding process, the maximum value, minimum value, and standard deviation of the energy entropy, or the shape characteristics of the time-varying entropy curve (such as rising or falling trends, fluctuation patterns, etc.).
[0029] Further, extracting the envelope of the parsed signal includes: The instantaneous amplitude of the analytic signal is calculated using the following formula: ; in, Indicates analytic signal Instantaneous amplitude at time, Indicates an analytical signal. This indicates that the analytical signal corresponds to the noise-reduced ultrasonic amplitude signal. Indicates the noise-reducing ultrasonic amplitude signal Hilbert transform, Represents the imaginary unit. This indicates the modulo operation of a complex number; Based on the instantaneous amplitude, the instantaneous amplitude of the analytical signal is extracted.
[0030] The instantaneous amplitude refers to the amplitude value of the analytic signal at any time t, used to analyze the energy distribution and dynamic changes of the signal, and is the basis for extracting the envelope. The Hilbert transform is a linear operation performed on the real-valued signal x(t), used to construct the analytic signal and extract the instantaneous phase and instantaneous amplitude of the signal. The imaginary unit is the mathematically defined unit j, which satisfies j^2 = -1, used to construct the complex signal (analytic signal), combining the real part of the signal with its Hilbert transform (imaginary part). The complex modulus operation refers to the operation of calculating the instantaneous amplitude of the analytic signal and converting the complex signal back to the envelope in real form.
[0031] Based on the welding pressure signal, this invention analyzes the dynamic stiffness coefficient of the bonding silver wire, which can further analyze the welding quality and provide a quantitative basis for process optimization.
[0032] Specifically, the step of analyzing the dynamic stiffness coefficient of the bonding silver wire based on the welding pressure signal includes: The welding pressure signal is baseline-corrected to obtain a corrected welding pressure signal. Identify the key feature points of the corrective welding pressure signal; Based on the aforementioned key feature points, a pressure-displacement relationship model for the bonded silver wire is established. Based on the pressure-displacement relationship model, the frequency domain stiffness coefficient and time domain stiffness coefficient of the bonded silver wire are extracted. The dynamic stiffness coefficient of the bonded silver wire is determined by combining the frequency domain stiffness coefficient and the time domain stiffness coefficient.
[0033] The corrected welding pressure signal refers to the signal after baseline correction of the original welding pressure signal. The key feature points refer to points with significant physical meaning in the corrected welding pressure signal, such as extreme points (peak values, valley values), zero-crossing points, inflection points, etc. The pressure-displacement relationship model refers to a mathematical model describing the relationship between welding pressure and bonding wire displacement. The frequency domain stiffness coefficient refers to a parameter describing the system stiffness in the frequency domain, reflecting the transmission characteristics of pressure and displacement in the frequency domain. The time domain stiffness coefficient refers to a parameter describing the system stiffness in the time domain, usually the slope or parameter of the pressure-displacement relationship model. The dynamic stiffness coefficient refers to a combination of frequency domain and time domain stiffness coefficients, reflecting the stiffness characteristics of the system under dynamic loads.
[0034] Optionally, based on the pressure-displacement relationship model, the frequency domain stiffness coefficient and time domain stiffness coefficient of the bonded silver wire are extracted. The frequency domain stiffness coefficient can be determined by calculating the amplitude of the frequency domain transfer function, and the time domain stiffness coefficient can be analyzed by extracting the slope from the pressure-displacement relationship model. For details, please refer to [link to relevant documentation]. Figure 2This is a schematic diagram of stiffness coefficient analysis for implementing the process optimization method based on ultrasonic welding signal of bonded silver wire according to an embodiment of the present invention. The results were obtained through analysis in the frequency domain and time domain, respectively.
[0035] S3. Using the transducer current signal, establish a nonlinear impedance model of the bonding silver wire to analyze the energy loss rate of the bonding silver wire, and analyze the microscopic deformation gradient of the bonding silver wire based on the solder joint deformation image sequence.
[0036] The present invention utilizes the transducer current signal to establish a nonlinear impedance model of the bonded silver wire, which can improve the accuracy of subsequent energy loss analysis.
[0037] In detail, establishing the nonlinear impedance model of the bonded silver wire using the transducer current signal includes: Calculate the instantaneous power and resonant frequency offset of the transducer current signal; Based on the instantaneous power and resonant frequency offset, an equivalent circuit for the bonding silver wire is established. Identify the impedance parameters of the equivalent circuit; Based on the impedance parameters, a nonlinear impedance model of the bonded silver wire is established.
[0038] Wherein, the instantaneous power of the transducer current signal at a certain moment represents the rate of energy conversion at that moment; the resonant frequency offset of the transducer during ultrasonic welding is the offset of its resonant frequency relative to the unloaded state; the equivalent circuit abstracts the electromechanical characteristics of the bonded silver wire during ultrasonic welding into a combination model of circuit elements (resistors, inductors, capacitors, etc.); the impedance parameter is the impedance value of each element in the equivalent circuit, including resistors, inductors, capacitors, etc.; and the nonlinear impedance model is an impedance model that considers the nonlinearity of the silver wire material (such as hysteresis, saturation, contact nonlinearity).
[0039] Optionally, establishing a nonlinear impedance model of the bonded silver wire based on the impedance parameters is an iterative process. Typically, it involves analyzing the nonlinear characteristics in the impedance parameters to select a suitable model structure (physical basis or data-driven), identifying the model parameters through optimization algorithms, and finally conducting rigorous verification.
[0040] This invention analyzes the energy loss rate of the bonded silver wire to determine the bonding loss of the bonded silver wire.
[0041] In detail, the analysis of the energy loss rate of the bonded silver wire includes: Based on the nonlinear impedance model of the bonded silver wire, the dynamic impedance of the bonding interface of the bonded silver wire is calculated. Based on the dynamic impedance of the bonding interface, the energy loss rate of the bonding silver wire is calculated using the following formula: ; in, Indicates bonding silver wire Energy loss rate at any given time This represents the linear weighting coefficient for loss. Indicates the dynamic impedance of the bonding interface. This represents the secondary loss coefficient. This indicates the resonant frequency shift of the bonded silver wire. This represents the loss coefficient due to the rate of change of frequency. This represents the rate of change of the absolute value of the frequency offset. This represents a differential operator.
[0042] Wherein, the energy loss rate refers to the rate of energy loss at time t due to electrical and mechanical losses (such as resistance heating, internal friction of materials, sound radiation, etc.) of the bonded silver wire and its interface; the dynamic impedance of the bonded interface refers to the time-dependent impedance value obtained from the nonlinear impedance model; the loss linear weighting coefficient refers to the contribution weight of the real part of the dynamic impedance of the bonded interface to the total energy loss rate; the quadratic loss coefficient refers to the contribution weight of the square of the amplitude of the dynamic impedance of the bonded interface to the total energy loss rate; the frequency change rate loss coefficient refers to the contribution weight of the absolute value change rate of the resonant frequency offset to the total energy loss rate; the differential operator refers to the mathematical derivative operator, representing the rate of change with respect to time t; and the absolute value change rate refers to the absolute value of the derivative of the resonant frequency offset (Δf(t)) with respect to time (t).
[0043] Based on the solder joint deformation image sequence, this invention analyzes the solder joint micro-deformation gradient of the bonding silver wire, which can accurately quantify the micro-deformation characteristics of the solder joint and provide a core evaluation method for high-reliability bonding processes.
[0044] Specifically, the analysis of the microscopic deformation gradient of the bonding silver wire solder joints based on the solder joint deformation image sequence includes: The weld joint deformation image sequence is filtered to obtain a filtered weld joint deformation image; Locate the solder joint region in the filtered solder joint deformation image; Calculate the displacement field of the solder joint region to calculate the Jacobian matrix of the bonding silver wire; The microscopic deformation gradient of the bonding silver wire joints is analyzed based on the Jacobian matrix.
[0045] Wherein, the filtered solder joint deformation image is the solder joint deformation image after image filtering processing, the solder joint region refers to the location region in the filtered image that is determined to be the location of the solder joint, the Jacobian matrix refers to the matrix used to describe the local rate of change of the displacement field, and the solder joint micro-deformation gradient refers to the vector field describing the relative deformation difference between different positions inside the solder joint.
[0046] Optionally, the image filtering of the weld joint deformation image sequence to obtain the filtered weld joint deformation image can be achieved by Gaussian filtering, median filtering, bilateral filtering, etc.
[0047] S4. Combining the energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and micro-deformation gradient of the solder joint, analyze the solder joint failure coefficient and optimization constraints of the bonding silver wire.
[0048] This invention combines the energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and micro-deformation gradient of the solder joint to analyze the solder joint failure coefficient and optimize the constraints of the bonding silver wire. This can quantify the solder joint failure risk and provide clear constraints for the optimization of the welding process, thereby improving the reliability of the bonding silver wire solder joint.
[0049] In detail, the analysis of the solder joint failure coefficient and optimization constraints of the bonded silver wire, combining the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient, and the micro-deformation gradient of the solder joint, includes: The energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and weld joint micro-deformation gradient are analyzed to determine the weld joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and weld joint deformation failure coefficient. The failure coefficient of the bonding silver wire is analyzed by combining the energy entropy solder joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and solder joint deformation failure coefficient. Based on the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient, and the micro-deformation gradient of the solder joint, the single-feature constraints and coupling constraints of the bonding silver wire are determined. Based on the single feature constraint and coupling constraint, the optimal constraint conditions for the bonded silver wire are determined.
[0050] The energy entropy weld joint failure coefficient refers to the uniformity and stability of the energy distribution of the weld joint; the energy loss failure coefficient refers to the rate of energy dissipation of the weld joint; the dynamic stiffness failure coefficient refers to the degree of stiffness degradation of the weld joint under dynamic load; the weld joint deformation failure coefficient refers to the influence of the micro-deformation gradient of the weld joint on failure; the weld joint failure coefficient refers to the total failure risk index that combines the above single-feature failure coefficients; the single-feature constraint refers to an independent threshold condition set for a single feature, such as the energy distribution stability threshold: 0.7 ≤ energy entropy H ≤ 1.0, maximum allowable energy loss: energy loss rate ≤ 10%; the coupling constraint refers to the constraint condition that considers the interaction between features, such as preventing high energy loss and local deformation concentration from occurring simultaneously, and when energy entropy decreases, stiffness needs to be increased accordingly to compensate for stability; the optimization constraint refers to the optimization goal that combines single-feature constraints and coupling constraints to guide the design of welding processes or materials.
[0051] Optionally, the analysis of the solder joint failure coefficient of the bonding silver wire by combining the energy entropy solder joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and solder joint deformation failure coefficient can be obtained by defining different weights for the energy entropy solder joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and solder joint deformation failure coefficient, and performing weighted analysis.
[0052] S5. Combining the preset multi-objective optimization function, the solder joint failure coefficient, and the optimization constraints, optimize the process parameters of the bonding silver wire to obtain optimized process parameters, and perform process optimization of the bonding silver wire based on the optimized process parameters.
[0053] This invention combines a preset multi-objective optimization function, the solder joint failure coefficient, and the optimization constraints to optimize the process parameters of the bonding silver wire. The resulting optimized process parameters systematically improve the bonding silver wire's process parameters, balancing failure risk and process cost, and enhancing solder joint reliability. These optimized process parameters include welding temperature, welding pressure, welding time, and welding speed.
[0054] First, this method significantly improves the reliability and consistency of weld quality. Traditional welding processes often rely on experience or static parameter settings, making it difficult to cope with uncertainties caused by material differences and environmental fluctuations. This method, however, monitors ultrasonic amplitude, welding pressure, transducer current, and weld deformation in real time, and analyzes weld failure risks using advanced indicators such as energy entropy, energy loss rate, dynamic stiffness coefficient, and micro-deformation gradient. It can accurately identify potential factors that may lead to weld failure, making process optimization no longer a "black box" operation but a precise control based on scientific data. This significantly reduces the weld defect rate and ensures stable and reliable product performance. Second, this method achieves dynamic optimization and adaptive control of welding process parameters. By establishing a nonlinear impedance model of the transducer current signal to analyze energy loss, and combining other characteristics to analyze the failure coefficient, the system can automatically calculate the optimal combination of process parameters based on a preset multi-objective optimization function (which may simultaneously consider quality, efficiency, cost, etc.) and optimization constraints. The welding process can be dynamically adjusted based on real-time signal feedback. For example, when abnormal energy loss or excessive deformation gradient is detected, the welding power or pressure is automatically fine-tuned, avoiding over-welding or under-welding problems that may occur under traditional fixed parameters. This achieves more refined process control. Furthermore, this method improves production efficiency and reduces manufacturing costs. Intelligent process optimization reduces trial-and-error costs and rework and scrap losses caused by poor welding. Optimized process parameters can usually shorten welding time or reduce energy consumption while ensuring quality, thereby increasing production cycle time and saving energy. In addition, comprehensive monitoring and data analysis of the welding process provides valuable data support for equipment maintenance and process improvement, helping to achieve preventive maintenance and continuous improvement. Finally, this method promotes the intelligent upgrading of welding technology. It combines advanced signal processing, machine learning, and multi-objective optimization technologies with traditional welding processes, not only improving the intelligence level of individual processes but also laying the foundation for the digital and intelligent transformation of the entire manufacturing process. Therefore, this invention can improve the process optimization effect of bonding silver wires.
[0055] Example 2: like Figure 3 The diagram shown is a functional block diagram of a process optimization system based on ultrasonic welding signals of bonded silver wires according to the present invention.
[0056] Welding signal acquisition module and dynamic stiffness analysis module Deformation gradient analysis module, solder joint failure analysis module Bonding process optimization module The process optimization system 300 based on ultrasonic welding signals of bonded silver wires described in this invention can be installed in an electronic device. Depending on the functions implemented, the process optimization system based on ultrasonic welding signals of bonded silver wires may include a welding signal acquisition module 301, a dynamic stiffness analysis module 302, a deformation gradient analysis module 303, a weld joint failure analysis module 304, and a bonding process optimization module 305. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and is stored in the memory of the electronic device.
[0057] In this embodiment of the invention, the functions of each module / unit are as follows: The welding signal acquisition module 301 is used to extract the ultrasonic welding signal of the bonded silver wire, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence. The dynamic stiffness analysis module 302 is used to analyze the energy entropy characteristics of the bonded silver wire based on the ultrasonic amplitude signal, and to analyze the dynamic stiffness coefficient of the bonded silver wire based on the welding pressure signal. The deformation gradient analysis module 303 is used to establish a nonlinear impedance model of the bonding silver wire using the transducer current signal, so as to analyze the energy loss rate of the bonding silver wire and analyze the micro-deformation gradient of the bonding silver wire based on the solder joint deformation image sequence. The solder joint failure analysis module 304 is used to analyze the solder joint failure coefficient and optimization constraints of the bonded silver wire by combining the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient and the solder joint micro-deformation gradient. The bonding process optimization module 305 is used to optimize the process parameters of the bonding silver wire by combining a preset multi-objective optimization function, the solder joint failure coefficient and the optimization constraint conditions, to obtain optimized process parameters, and to perform process optimization of the bonding silver wire based on the optimized process parameters.
[0058] In detail, the modules in the process optimization system 300 based on ultrasonic welding signals of bonded silver wire described in this embodiment of the invention employ the same methods as described above. Figure 1 The process optimization method based on the ultrasonic welding signal of bonded silver wire described in the article uses the same technical means and can produce the same technical effect, so it will not be repeated here.
[0059] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A process optimization method based on ultrasonic welding signals of bonded silver wires, characterized in that, The method includes: The ultrasonic welding signal of the bonded silver wire is extracted, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence; Based on the ultrasonic amplitude signal, the energy entropy characteristics of the bonded silver wire are analyzed, and the dynamic stiffness coefficient of the bonded silver wire is analyzed based on the welding pressure signal. Using the transducer current signal, a nonlinear impedance model of the bonding silver wire is established to analyze the energy loss rate of the bonding silver wire. Based on the solder joint deformation image sequence, the micro-deformation gradient of the solder joint of the bonding silver wire is analyzed. Based on the energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and micro-deformation gradient of the solder joint, the solder joint failure coefficient and optimization constraints of the bonding silver wire are analyzed. By combining the preset multi-objective optimization function, the solder joint failure coefficient, and the optimization constraints, the process parameters of the bonding silver wire are optimized to obtain optimized process parameters, and the process optimization of the bonding silver wire is performed based on the optimized process parameters.
2. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 1, characterized in that, The ultrasonic welding signal extracted from the bonded silver wire includes: The signal acquisition sensor configured with the bonded silver wire includes a high-frequency laser vibrometer, a miniature dynamic force sensor, a high-frequency current probe, and a high-speed camera. Construct the sensing topology network of the signal acquisition sensor; Based on the aforementioned sensing topology network, configure the synchronization pulse signal of the signal acquisition sensor; Based on the synchronous pulse signal, the ultrasonic welding signal of the bonded silver wire is extracted, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence.
3. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 2, characterized in that, The analysis of the energy entropy characteristics of the bonded silver wire based on the ultrasonic amplitude signal includes: The ultrasonic amplitude signal is subjected to wavelet denoising to obtain a denoised ultrasonic amplitude signal; Calculate the analytical signal of the noise-reduced ultrasonic amplitude signal; Extract the envelope of the analyzed signal; Calculate the energy percentage of the equal-width energy range corresponding to the envelope; Based on the energy ratio, the energy entropy of the noise-reduced ultrasonic amplitude signal is calculated to construct the time-varying entropy curve of the noise-reduced ultrasonic amplitude signal; The energy entropy characteristics of the bonded silver wire are analyzed based on the time-varying entropy curve.
4. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 3, characterized in that, Extracting the envelope of the analyzed signal includes: The instantaneous amplitude of the analytic signal is calculated using the following formula: ； in, Indicates analytic signal Instantaneous amplitude at time, Indicates an analytical signal. This indicates that the analytical signal corresponds to the noise-reduced ultrasonic amplitude signal. Indicates the noise-reducing ultrasonic amplitude signal Hilbert transform, Represents the imaginary unit. This indicates the modulo operation of a complex number; Based on the instantaneous amplitude, the instantaneous amplitude of the analytical signal is extracted.
5. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 4, characterized in that, The step of analyzing the dynamic stiffness coefficient of the bonding silver wire based on the welding pressure signal includes: The welding pressure signal is baseline-corrected to obtain a corrected welding pressure signal. Identify the key feature points of the corrective welding pressure signal; Based on the aforementioned key feature points, a pressure-displacement relationship model for the bonded silver wire is established. Based on the pressure-displacement relationship model, the frequency domain stiffness coefficient and time domain stiffness coefficient of the bonded silver wire are extracted. The dynamic stiffness coefficient of the bonded silver wire is determined by combining the frequency domain stiffness coefficient and the time domain stiffness coefficient.
6. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 5, characterized in that, The step of establishing a nonlinear impedance model of the bonded silver wire using the transducer current signal includes: Calculate the instantaneous power and resonant frequency offset of the transducer current signal; Based on the instantaneous power and resonant frequency offset, an equivalent circuit for the bonding silver wire is established. Identify the impedance parameters of the equivalent circuit; Based on the impedance parameters, a nonlinear impedance model of the bonded silver wire is established.
7. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 6, characterized in that, The analysis of the energy loss rate of the bonded silver wire includes: Based on the nonlinear impedance model of the bonded silver wire, the dynamic impedance of the bonding interface of the bonded silver wire is calculated. Based on the dynamic impedance of the bonding interface, the energy loss rate of the bonding silver wire is calculated using the following formula: ； in, Indicates bonding silver wire Energy loss rate at any given time This represents the linear weighting coefficient for loss. Indicates the dynamic impedance of the bonding interface. This represents the secondary loss coefficient. This indicates the resonant frequency shift of the bonded silver wire. This represents the loss coefficient due to the rate of change of frequency. This represents the rate of change of the absolute value of the frequency offset. This represents a differential operator.
8. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 7, characterized in that, The analysis of the micro-deformation gradient of the bonding silver wire solder joints based on the solder joint deformation image sequence includes: The weld joint deformation image sequence is filtered to obtain a filtered weld joint deformation image; Locate the solder joint region in the filtered solder joint deformation image; Calculate the displacement field of the solder joint region to calculate the Jacobian matrix of the bonding silver wire; The microscopic deformation gradient of the bonding silver wire joints is analyzed based on the Jacobian matrix.
9. The process optimization method based on ultrasonic welding signals of bonded silver wire as described in claim 8, characterized in that, The analysis of the solder joint failure coefficient and optimization constraints of the bonded silver wire, combining the energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and solder joint micro-deformation gradient, includes: The energy entropy characteristics, energy loss rate, dynamic stiffness coefficient, and weld joint micro-deformation gradient are analyzed to determine the weld joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and weld joint deformation failure coefficient. The failure coefficient of the bonding silver wire is analyzed by combining the energy entropy solder joint failure coefficient, energy loss failure coefficient, dynamic stiffness failure coefficient, and solder joint deformation failure coefficient. Based on the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient, and the micro-deformation gradient of the solder joint, the single-feature constraints and coupling constraints of the bonding silver wire are determined. Based on the single feature constraint and coupling constraint, the optimal constraint conditions for the bonded silver wire are determined.
10. A process optimization system based on ultrasonic welding signals of bonded silver wires, characterized in that, The system includes: The welding signal acquisition module is used to extract the ultrasonic welding signal of the bonded silver wire, wherein the ultrasonic welding signal includes ultrasonic amplitude signal, welding pressure signal, transducer current signal and weld joint deformation image sequence. The dynamic stiffness analysis module is used to analyze the energy entropy characteristics of the bonded silver wire based on the ultrasonic amplitude signal, and to analyze the dynamic stiffness coefficient of the bonded silver wire based on the welding pressure signal. The deformation gradient analysis module is used to establish a nonlinear impedance model of the bonding silver wire using the transducer current signal, so as to analyze the energy loss rate of the bonding silver wire and analyze the micro-deformation gradient of the bonding silver wire based on the solder joint deformation image sequence. The solder joint failure analysis module is used to analyze the solder joint failure coefficient and optimization constraints of the bonding silver wire by combining the energy entropy characteristics, the energy loss rate, the dynamic stiffness coefficient, and the solder joint micro-deformation gradient. The bonding process optimization module is used to optimize the process parameters of the bonding silver wire by combining the preset multi-objective optimization function, the solder joint failure coefficient, and the optimization constraints to obtain optimized process parameters, and to perform process optimization of the bonding silver wire based on the optimized process parameters.