A method and system for controlling micro solder joints of radio frequency connectors based on improved genetic algorithm

Abstract

The present invention discloses a method and system for controlling micro-solder points of radio frequency connectors based on an improved genetic algorithm. Through various improvements to the traditional genetic algorithm, the limitations of the traditional solution are effectively overcome, including: through an adaptive genome adjustment mechanism, the present invention can respond to parameter fluctuations in the welding process in real time, dynamically adjust the gene length and structure of the population individuals, and ensure that the optimization process always adapts to changes in the current environment; multi-dimensional crossover operations enhance the diversity of the population, avoid the algorithm from falling into the local optimum, and improve the optimization effect and efficiency; by introducing a local environmental adaptation factor, the present invention can simulate and adapt to environmental interference in the actual welding process to ensure the stability of the solder joint quality. Finally, the present invention dynamically adjusts and optimizes the welding parameters through a further real-time feedback control mechanism to ensure the high precision and stability of the welding process, and significantly improves the reliability and performance of the radio frequency connector.

Description

Technical Field

[0001] The invention belongs to the field of welding technology and intelligent control, and in particular relates to a radio frequency connector micro-welding point control method and system based on an improved genetic algorithm. Background Art

[0002] In modern communication technology, RF connectors, as key components, are widely used in various communication equipment. Their quality and performance directly affect the stability and reliability of the entire communication system. In the manufacturing process of RF connectors, micro-welding technology is an important process. The precise control of micro-solder points is essential to ensure the electrical performance and mechanical strength of RF connectors. However, with the miniaturization and high frequency development of RF connectors, traditional micro-welding control methods face great challenges in terms of accuracy and stability.

[0003] Existing micro-welding technology usually relies on manual setting and adjustment of welding parameters, such as welding temperature, pressure, time, current and speed. The setting of these parameters is often based on experience and trial and error. Due to the complex coupling relationship between welding parameters, traditional methods are difficult to achieve high-precision control of micro-solder joints, resulting in unstable welding quality and affecting the final performance of RF connectors. For example, during the welding process, slight changes in temperature and pressure may cause deviations in solder joint size or insufficient welding strength, which in turn causes the electrical performance of the connector to deteriorate, and even cause problems such as poor contact. In addition, random interference factors in the welding process, such as ambient temperature fluctuations and welding equipment vibrations, will also have unpredictable effects on welding quality, making it difficult for traditional welding parameter optimization methods to adapt to the complex environment in actual production.

[0004] In order to improve the quality control accuracy of micro-weld joints, intelligent optimization algorithms have been introduced into welding parameter optimization in recent years. Among them, genetic algorithms, as an intelligent algorithm with global search capabilities, have gradually attracted attention. By simulating the process of natural selection and genetic variation, genetic algorithms can search for optimal solutions in a large parameter space, overcoming the defects of traditional methods that rely on experience and local optimization. However, standard genetic algorithms still have some shortcomings when applied to micro-weld joint control. First, the convergence speed of standard genetic algorithms is slow and they are prone to fall into local optimality, resulting in the final welding parameter combination not being the global optimal. Second, standard genetic algorithms have limited ability to deal with dynamic changes and random interference in the welding process, and it is difficult to adapt to parameter fluctuations and environmental changes in the welding process in real time. Finally, standard genetic algorithms usually use fixed gene encoding length and structure, which may not effectively capture the complex relationship between parameters in complex and changeable welding processes, resulting in unsatisfactory optimization results.

[0005] In summary, the existing technology still has many technical difficulties in the precise control of micro solder joints, which are mainly reflected in:

[0006] (1) Inefficient welding parameter optimization: Traditional welding parameter optimization methods mostly rely on manual experience and trial and error, which makes the optimization process time-consuming and difficult to find the optimal parameter combination, and cannot meet the high-precision control requirements of micro-weld joints.

[0007] (2) Lack of real-time adaptability to dynamic changes: The existing welding parameter optimization algorithm is unable to adjust and optimize the control parameters in a timely manner when faced with dynamic changes and random interference during the welding process, resulting in unstable weld quality.

[0008] (3) Limited algorithm optimization capability: The application of standard genetic algorithms in complex welding scenarios is limited. They are prone to falling into local optimality and have difficulty in handling the complex coupling relationship between parameters in the welding process, resulting in the optimization results failing to achieve the desired effect.

[0009] In summary, the existing technology has many defects in the precise control of RF connector micro-solder points, which directly affect the welding quality and overall performance of RF connectors. How to improve the efficiency of welding parameter optimization, enhance the adaptability of the algorithm to dynamic environments, and improve the quality of optimization results have become technical problems that need to be solved urgently. Summary of the invention

[0010] In order to solve the above technical problems, the present invention provides a method and system for controlling micro solder joints of a radio frequency connector based on an improved genetic algorithm.

[0011] The technical solution provided by the present invention is as follows:

[0012] A method for controlling micro solder joints of radio frequency connectors based on an improved genetic algorithm comprises the following steps:

[0013] S1. Collect welding parameters during welding and build key parameter data sets;

[0014] S2, initializing the population of the genetic algorithm model, setting the number of individuals and individual gene codes of the initial population, wherein the individual gene codes are different combinations of parameter vectors in the key parameter data set;

[0015] S3. Introduce an adaptive genome adjustment mechanism to dynamically adjust the gene length and structure of individuals in the population;

[0016] S4, evaluate the fitness of the adjusted population individuals and select the population individuals with high fitness;

[0017] S5. Use multidimensional crossover operation to cross the gene fragments of two or more population individuals;

[0018] S6. Introduce local environmental adaptation factors and select population individuals with high adaptability to environmental disturbance;

[0019] S7, introducing dynamic mutation probability, by randomly changing the parameter value in the individual gene code, performing gene mutation operation on the individuals in the population, and the mutation probability is adjusted in real time according to the welding quality;

[0020] S8, re-evaluate the fitness of the individuals in the population after the gene mutation operation, and perform selection, crossover and mutation operations, repeating multiple iterations until the fitness evaluation function value or the number of iterations reaches a preset threshold;

[0021] S9. Apply the optimized key parameters to the actual welding process, and further adjust and optimize the control parameters according to the real-time monitoring data through the feedback mechanism.

[0022] Furthermore, in step S1, the welding parameters include the instantaneous temperature T(t) of the welding point, the welding pressure P(t) applied to the welding point, the welding duration t w , welding instantaneous current I(t) and welding head moving speed v(t); the collected welding parameters are collected in chronological order to form a parameter vector X(t) = [T(t), P(t), t w ,I(t),v(t)], and finally form a key parameter data set containing n parameter vectors.

[0023] Furthermore, in step S2, an initial genetic code is generated for each individual in the population. The genetic code G of the i-th individual in the population is i It is expressed as:

[0024]

[0025] Among them, α n is a weight factor used to adjust the effect of parameters at different time points on gene coding; T(t n )、P(t n ) and I(t n ) are the time series of temperature, pressure and current, respectively, I mean is the average value of the current, σ is the standard deviation of the current; N is the number of individuals in the initial population.

[0026] Furthermore, step S3 specifically includes the following steps:

[0027] S301, monitoring the fluctuation of welding parameters during welding, and obtaining the fluctuation amplitude of welding parameters in real time ΔX(t)=[ΔT(t),ΔP(t),Δt w ,ΔI(t),Δv(t)], where the fluctuation amplitudes represent the changes in temperature, pressure, welding time, current and speed respectively;

[0028] S302, calculating the adjustment factor λ(t)=1+α||ΔX(t)||||X(t)|| of the gene length of the individual in the population, the adjustment factor is used to dynamically adjust the length of the gene code of the individual, wherein α is a proportional constant, ||ΔX(t)|| represents the modulus of the fluctuation amplitude, and ||X(t)|| represents the modulus of the key parameter;

[0029] S303, dynamically adjust the gene encoding length L of each individual in the population according to the adjustment factor λ(t) i (t) = λ(t)·L i,0 , L i (t) represents the gene coding length of the i-th individual at time t, L i,0 is the initial gene coding length of the i-th individual;

[0030] S304. While adjusting the gene coding length, a structural adjustment factor β(t) is introduced to change the arrangement order of gene segments or add / delete specific gene segments; Among them, γ is the proportional constant, and β(t) represents the number of structural adjustments, which is used to adaptively adjust the individual gene structure according to real-time fluctuations.

[0031] Furthermore, in step S4, the fitness of the individuals in the population is evaluated by setting a fitness function, wherein the fitness function is constructed based on quality indicators of the welds, and the quality indicators include strength, dimensional consistency, and surface flatness.

[0032] Furthermore, in step S5, the gene codes of the selected two or more high-fitness individuals are subjected to a multi-dimensional crossover operation, a crossover point is randomly selected in each gene dimension, and the gene fragments corresponding to the crossover point are exchanged to generate population individuals with new gene codes; after the multi-dimensional crossover operation, the fitness of the population individuals is evaluated respectively, and the individuals with high fitness are retained to form a new population for subsequent iterations.

[0033] Further, in step S6, the local environmental adaptation factor is expressed as ζ(t)=1 / (1+||R(t)||·ε), which is used to measure the adaptability of the population individuals under interference conditions; wherein ε is the adjustment factor of the environmental noise, which is used to control the influence of the interference factor on the adaptability factor; ||R(t)|| is the modulus of the interference factor, and the interference factor R(t)=[r T (t),r I (t),r v (t)],r T (t) is the temperature fluctuation, r I (t) is the current disturbance, r v (t) is a random function of mechanical vibration;

[0034] Calculate the fitness value of individuals in the population under interference conditions ζ (i) = Fitness(i)·ζ(t), where Fitness(i) is the original fitness value of individual i, and the population individuals with high environmental adaptability are retained.

[0035] Furthermore, in step S9, the welding parameters corresponding to the individual gene codes retained by the genetic algorithm optimization are applied to the actual welding process, and the welding quality index data are collected in real time during the welding process to form real-time feedback data; the real-time feedback data Q(t) is compared with the preset quality standard Q target Compare and calculate the deviation value ΔQ(t) = Q(t)-Q target ; Real-time optimization and adjustment of welding parameters based on deviation values: X adj (t) = X opt (t)+K·ΔQ(t), where X adj (t) is the adjusted welding parameter vector, X opt (t) is the welding parameter vector before adjustment, and K is the feedback gain matrix used to adjust the feedback response degree; the above feedback control and parameter adjustment operations are continuously performed during the welding process until the quality of the weld meets the preset quality standard; finally, the optimized and adjusted parameter data is stored in the welding parameter database for reference and further optimization of subsequent welding operations.

[0036] A radio frequency connector micro solder joint control system based on the above method, comprising:

[0037] A data set building module, used for forming and storing parameter vectors according to welding parameters collected during the welding process;

[0038] A genetic algorithm optimization module is used to construct the parameter vector in the module according to the data set, perform iterative optimization of welding parameters using the genetic algorithm, and output the optimal welding parameters;

[0039] The feedback adjustment module is used to adaptively adjust the optimal welding parameters output by the genetic algorithm optimization module according to real-time monitoring data during the welding process.

[0040] Furthermore, the genetic algorithm optimization module also includes:

[0041] Population initialization module, used to initialize the number of individuals and individual gene codes of the population;

[0042] Adaptive genome adjustment module, used to dynamically adjust the gene length and structure of individuals in the population;

[0043] The fitness evaluation module is used to select and retain population individuals with high fitness according to the set fitness evaluation function;

[0044] Genetic crossover module, used to perform multidimensional crossover operations on the genetic codes of individuals in the population;

[0045] Environmental disturbance fitness assessment module, used to measure the adaptability of population individuals under disturbance conditions and select population individuals with high adaptability to environmental disturbance;

[0046] Gene mutation module, used to randomly change the parameter values ​​in individual gene encoding to perform gene mutation operations;

[0047] The iteration control module is used to control the optimization iteration process of the genetic algorithm and set the iteration stop condition.

[0048] Compared with the prior art, the present invention has at least the following beneficial effects:

[0049] (1) Improved welding parameter optimization efficiency and ensured high-precision control of micro-solder joints: The present invention uses an improved genetic algorithm to globally optimize key welding parameters (temperature, pressure, welding time, current and speed), which greatly improves the efficiency and accuracy of parameter optimization compared to the empirical adjustment trial and error method and the traditional genetic optimization algorithm. Through diversified processing when initializing the population and multi-dimensional cross-operation, the present invention can search for the global optimal solution more quickly, thereby ensuring that the parameter settings during the welding process can accurately match the requirements of the micro-solder joints, significantly improving the consistency and quality stability of the solder joints.

[0050] (2) Enhanced the adaptability of the algorithm to dynamic environments and improved the stability of the welding process: The present invention enhances the adaptability of the genetic algorithm to dynamic changes and random interference in the welding process by introducing an adaptive genome adjustment mechanism and a local environmental adaptation factor. Specifically, the adaptive genome adjustment mechanism dynamically adjusts the gene length and structure of individuals in the population by real-time analysis of parameter fluctuations in the welding process, thereby ensuring that the algorithm can adapt to sudden changes that may occur during the welding process. The local environmental adaptation factor further improves the adaptability of individuals in the population in the actual production environment by simulating random interference factors in the actual welding environment, ensuring that the welding quality is not affected by environmental fluctuations and improving the stability of the weld quality.

[0051] (3) The present invention ensures continuous monitoring and optimization of welding quality by applying optimized key parameters to the actual welding process and combining with a real-time feedback control mechanism, collects quality data during the welding process in real time, and dynamically adjusts welding parameters using a feedback control mechanism to ensure that the quality of the solder joints always remains within the preset standard range. This not only effectively copes with parameter fluctuations that may occur during the welding process, but also ensures the reliability of RF connectors in high-frequency, high-precision applications by continuously optimizing control parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0053] Figure 1 It is a schematic diagram of a method flow provided by an embodiment of the present invention;

[0054] Figure 2 It is a schematic diagram of a genetic algorithm model framework provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, other embodiments obtained by ordinary technicians in this field without making creative work are all within the scope of protection of the present invention.

[0056] Embodiment 1

[0057] This embodiment provides a method for controlling micro solder joints of a radio frequency connector based on an improved genetic algorithm. Figure 1 and Figure 2 As shown, it mainly includes the following steps:

[0058] S1) Collect welding parameters during welding and build key parameter data sets

[0059] During the micro-welding process, high-precision sensors are used to collect the temperature T(t), pressure P(t), welding time t in real time. w , welding current I(t) and welding speed v(t) and other key parameters, temperature T(t) represents the instantaneous temperature of the welding point at time t, pressure P(t) represents the welding pressure applied to the welding point, welding time t w It represents the duration of the welding process, the welding current I(t) represents the instantaneous value of the current during the welding process, and the welding speed v(t) represents the speed at which the welding head moves during the welding process.

[0060] The collected key parameter data are collected in chronological order to form a parameter vector X(t) = [T(t), P(t), t w ,I(t),v(t)]. Then the parameter vector X(t) is preprocessed, including denoising, normalization and data smoothing, and finally a key parameter data set containing n parameter vectors is formed. Ensure the accuracy and consistency of the data set. Among them, X(t i ) is the time point ti The corresponding parameter vector.

[0061] S2) Initialize the population of the genetic algorithm model, set the number of individuals and individual gene codes of the initial population, and the individual gene codes are different combinations of parameter vectors in the key parameter data set

[0062] Use the TensorFlow framework to initialize the population of the genetic algorithm model and determine the number of individuals N in the initial population, where N represents the number of candidate solutions calculated simultaneously in the genetic algorithm, and the number is set according to the complexity of the optimization problem and the computing resources.

[0063] Generate an initial genetic code G for each individual in the population i , each gene encoding consists of multiple sets of parameter vectors, representing different combinations of key parameter data sets D, individual gene encoding G i It can be expressed as:

[0064]

[0065] Among them, α n is a weight factor used to adjust the effect of parameters at different time points on gene coding; T(t n )、P(t n ) and I(t n ) are the time series of temperature, pressure and current, respectively, I mean is the average value of the current, and σ is the standard deviation of the current.

[0066] Using the data flow graph feature of TensorFlow, each gene is encoded G i Converted into TensorFlow tensor form for subsequent parallel computing. Specifically, the tensor form of gene encoding G' i It can be expressed as:

[0067] G' i = tf.Tensor(X(t i1 ),X(t i2 ),…,X(t ik )).

[0068] Through the batch processing function of TensorFlow, all the generated individual gene codes G' i Perform batch initialization to ensure efficient calculation, all individual gene encoding Form an initial population. Batch processing further improves the computational efficiency of genetic algorithms in high-dimensional parameter spaces.

[0069] The initialized population is stored as a trainable variable of TensorFlow so that it can be updated in subsequent iterations. The individual gene codes in the population will be used as the input of the genetic algorithm model to participate in subsequent crossover, mutation and selection operations to ensure the continuous optimization of key parameters of the welding process.

[0070] S3) Introducing an adaptive genome adjustment mechanism to dynamically adjust the gene length and structure of individuals in the population

[0071] In each iteration of the genetic algorithm, the fluctuation of key parameters in the welding process is monitored in real time, and the amplitude of parameter fluctuation is determined by real-time data flow analysis ΔX(t)=[ΔT(t),ΔP(t),Δt w ,ΔI(t),Δv(t)], where the fluctuation amplitudes represent the changes in temperature, pressure, welding time, current and speed respectively.

[0072] The adjustment factor λ(t)=1+α||ΔX(t)||||X(t)|| is used to dynamically adjust the length of the individual gene code based on the fluctuation amplitude. α is a proportional constant, ||ΔX(t)|| represents the modulus of the fluctuation amplitude, and ||X(t)|| represents the modulus of the key parameters, which is used to measure the impact of parameter fluctuations in the current welding process on the gene length. The gene code length L of each individual in the population is dynamically adjusted according to the adjustment factor λ(t). i (t) = λ(t)·L i,0 , L i (t) represents the gene coding length of the i-th individual at time t, L i,0 is the initial gene coding length of the i-th individual. Through this step, the gene coding length is extended when the parameter fluctuation is large to enhance the adaptability of the individual in the parameter fluctuation environment.

[0073] While adjusting the gene coding length, the individual gene structure is adjusted and a structural adjustment factor β(t) is introduced to change the arrangement order of gene fragments or add / delete specific gene fragments. Among them, γ is the proportional constant, and β(t) represents the number of structural adjustments, which is used to adaptively adjust the individual gene structure according to the real-time fluctuations and to make structural adjustments in combination with gene coding.

[0074] Through the adaptive genome adjustment mechanism, the gene length and gene structure of individuals in the population are dynamically adjusted to ensure the robustness and optimization effect of the genetic algorithm under different welding conditions and improve the stability of the weld quality.

[0075] S4) Evaluate the fitness of individuals in the adjusted initial population and select individuals with high fitness

[0076] A preliminary evaluation is performed by setting the fitness function, which is constructed based on the quality indicators of the solder joints (strength, dimensional consistency, and surface flatness). The fitness evaluation process is implemented using the computational graph of TensorFlow to ensure that large-scale data in the population can be quickly processed and parallel computing can be performed.

[0077] S5) Use multidimensional crossover operation to cross the gene fragments of two or more high fitness individuals

[0078] Select individuals with high fitness from the current population for multidimensional crossover operations and set the crossover probability P c , ensuring that individuals with high fitness participate in crossover first. The gene encoding of two or more selected individuals with high fitness is represented as G i ″ and G j ″, where G i ″=[g i1 ,g i2 ,Λ,g im ] and G j ″=[g j1 ,g j2 ,Λ,g jm ], m is the length of the gene code. i ″ and G j ″Perform a multi-dimensional crossover operation, randomly select a crossover point on each dimension d, and convert the gene fragment g id With g jd The crossover operation is performed in each dimension to make the new gene code diverse in multiple dimensions. The gene code after the crossover contains part of the information of the original individual in multiple dimensions to improve the adaptability of the offspring individuals.

[0079] The fitness of the offspring gene codes generated by the crossover is evaluated to ensure the effectiveness of the newly generated individual gene codes in the optimization of welding process parameters, and individuals with high fitness are retained to form a new population for subsequent iterations.

[0080] S6) Introducing local environmental adaptability factors, and conducting environmental adaptability tests on individuals in the population by simulating random interference factors in the actual welding environment;

[0081] In the optimization process of genetic algorithm, the random interference factors R(t) in the actual welding environment are simulated. The interference factors include temperature fluctuation, welding current disturbance, and mechanical vibration: R(t) = [r T (t),r I (t),r v (t)], where r T (t) is the temperature fluctuation, r I (t) is the current disturbance, rv (t) is a random function of mechanical vibration, simulating random changes in the actual environment.

[0082] The local environmental adaptation factor ζ(t) is introduced to measure the adaptability of individuals under simulated interference conditions: ζ(t) = 1 / (1+||R(t)||·ε), where ε is the adjustment factor of environmental noise, which controls the impact of interference factors on the adaptability factor, and ||R(t)|| is the modulus of the interference factor.

[0083] In each iteration, the environmental adaptation factor ζ(t) is used to test the fitness of each individual in the population and calculate the fitness value of the individual under interference conditions. ζ (i) Fitness ζ (i) = Fitness(i)·ζ(t), where Fitness(i) is the original fitness value of individual i and the fitness value after adjustment by the environmental adaptation factor is Fitness(t). ζ (i) It is more in line with the performance requirements under actual welding conditions. Based on the environmental adaptability test results, individuals with stronger adaptability are retained first to enter the next generation of populations, gradually improving the overall adaptability of the population to the actual welding environment and ensuring that the optimized welding parameters can be stably applied in actual production.

[0084] S7) introduces dynamic mutation probability, and performs gene mutation operation on individuals in the new generation population by randomly changing parameter values ​​in individual gene encoding. The mutation probability is adjusted in real time according to welding quality.

[0085] S8) The fitness of individuals in the new generation population is evaluated again, and selection, crossover and mutation operations are performed, and the process is repeated for multiple iterations until the fitness evaluation function value or the number of iterations reaches a preset threshold.

[0086] S9) Apply the optimized key parameters to the actual welding process, and further adjust and optimize the control parameters according to the real-time monitoring data through the feedback mechanism

[0087] The key parameters corresponding to the individuals retained by genetic algorithm optimization are applied to the actual welding process, and the welding quality data, including the weld strength, weld size consistency and weld surface flatness index, are collected in real time during the welding process to form real-time feedback data; the real-time feedback data is compared with the preset quality standards to calculate the deviation value ΔQ(t): ΔQ(t) = Q(t)-Q target , where Q target is the target quality index vector of the weld, and the deviation value ΔQ(t) is used to evaluate the actual effect of the welding process; the control parameters are adjusted based on the deviation value, and the temperature, pressure, welding time t, welding current and welding speed are optimized and adjusted in real time through the feedback control mechanism: Xadj (t) = X opt (t)+K·ΔQ(t), where X adj (t) is the adjusted control parameter vector, X opt (t) is the optimized parameter vector, and K is the feedback gain matrix, which is used to adjust the response degree of the feedback. The above feedback control and parameter adjustment operations are continuously performed during the welding process until the quality of the weld meets the preset standard to ensure the stability and consistency of the final welding quality. Finally, the optimized and adjusted parameter data is stored in the welding parameter database for reference and further optimization of subsequent welding operations.

[0088] To verify the effectiveness of the above method, an actual production practice case is given below.

[0089] In a production practice in July 2024, a communications equipment manufacturer was mass-producing a radio frequency connector for 5G communication systems. This connector requires extremely high welding precision, the diameter of the micro solder joint is controlled within 0.3 mm, and the welding strength must reach more than 10 N. However, in the actual production process, the factory has repeatedly found that the solder joint strength is insufficient and the solder joint size exceeds the standard due to improper welding parameter settings, which directly affects the product qualification rate.

[0090] In order to improve the welding quality, the factory decided to improve the consistency and strength of the solder joints by intelligently optimizing the welding parameters. During the production process from 9 pm on July 15, 2024 to 1 am the next day, the operator first set the welding parameters using traditional methods. These parameters include the welding temperature set to 320°C, the welding pressure to 3.5N, the welding time to 0.8 seconds, the welding current to 0.5A, and the welding speed to 2.5mm / s. However, in the first batch of 50 connectors welded, after quality inspection, it was found that 12 solder joints failed the strength test, the solder joint strength was less than 8N, and the diameter of 8 solder joints exceeded the specified 0.3 mm.

[0091] In this case, the operator adopted the RF connector micro-solder point control method based on the improved genetic algorithm provided in this embodiment. First, the welding parameters and solder point quality data of the first 50 connectors were collected to generate an initial data set. In the data analysis, it was found that the welding current and welding speed had a greater impact on the quality of the solder points, so these two parameters were taken as the focus of optimization. In the next 2 hours, the parameter range of the welding current and welding speed was dynamically adjusted by introducing an adaptive genome adjustment mechanism. During the optimization process, the parameters were optimized for multiple iterations by real-time collection of solder point temperature, pressure, current and speed data, and combined with a feedback mechanism. Finally, the parameter optimization was completed at 3:30 a.m., and the welding temperature was set to 318°C, the welding pressure was set to 3.4N, the welding time was set to 0.75 seconds, the welding current was adjusted to 0.48A, and the welding speed was set to 2.4mm / s. After applying the optimized parameters, the system began to mass produce new connectors. In the next 4 hours, a total of 200 connectors were produced. Quality inspection data showed that the solder joint strength of this batch of products reached more than 10N, and the solder joint diameter was all controlled within 0.3 mm, and the solder joint quality was significantly improved. At 3 pm on July 18, 2024, it was detected that the temperature fluctuation during the welding process exceeded the normal range, and the real-time feedback mechanism was activated. By adjusting the welding current and speed, the changes in ambient temperature were automatically compensated, ensuring that the solder joint quality of the 50 connectors produced subsequently was normal.

[0092] In order to more accurately evaluate the effectiveness of the method in this embodiment, the factory conducted a detailed experimental data comparison:

[0093] (1) Comparison of solder joint strength: Under the traditional method, among the 200 solder joints, 34 solder joints had strengths lower than 9N, the lowest strength was 7.5N, and the average strength was 8.8N. After using this method, among the 200 solder joints, none had strength lower than 10N, the lowest strength was 10.2N, and the average strength was 10.5N.

[0094] (2) Comparison of solder joint diameters: Under the traditional method, 26 of the 200 solder joints had diameters exceeding 0.3 mm, with an average diameter of 0.305 mm and a maximum diameter of 0.312 mm. After using this method, the diameters of the 200 solder joints did not exceed 0.3 mm, with an average diameter of 0.298 mm and a maximum diameter of 0.299 mm.

[0095] (3) Production efficiency comparison: Under the traditional method, the production time of 200 samples is 7 hours due to multiple rework and parameter adjustments. After using this method, the production time of 200 samples is shortened to 5 hours, and the production efficiency is improved by about 28.6%.

[0096] (4) Environmental adaptability test: Under the traditional method, when the workshop temperature fluctuates between 24℃ and 30℃, the quality of the solder joints decreases significantly, and the qualified rate is 72%. After using this method, within the same temperature fluctuation range, the quality of the solder joints is stable, and the qualified rate remains above 95%.

[0097] Through this production practice, the method provided in this embodiment has demonstrated its superiority in many aspects. Whether it is solder joint strength, solder joint diameter control, or production efficiency and environmental adaptability, it is significantly better than the traditional welding parameter setting method. Especially in actual production, in the face of unpredictable environmental changes and equipment fluctuations, the method of this embodiment can respond in time and automatically adjust parameters, ensuring the high quality and stability of the RF connector.

[0098] Embodiment 2

[0099] Corresponding to the method described in Example 1, this embodiment provides a radio frequency connector micro solder joint control system based on an improved genetic algorithm, which mainly includes:

[0100] A data set building module, used for forming and storing parameter vectors according to welding parameters collected during the welding process;

[0101] A genetic algorithm optimization module is used to construct the parameter vector in the module according to the data set, perform iterative optimization of welding parameters using the genetic algorithm, and output the optimal welding parameters;

[0102] The feedback adjustment module is used to adaptively adjust the optimal welding parameters output by the genetic algorithm optimization module according to real-time monitoring data during the welding process.

[0103] Wherein, the genetic algorithm optimization module also includes:

[0104] Population initialization module, used to initialize the number of individuals and individual gene codes of the population;

[0105] Adaptive genome adjustment module, used to dynamically adjust the gene length and structure of individuals in the population;

[0106] The fitness evaluation module is used to select and retain population individuals with high fitness according to the set fitness evaluation function;

[0107] Genetic crossover module, used to perform multidimensional crossover operations on the genetic codes of individuals in the population;

[0108] Environmental disturbance fitness assessment module, used to measure the adaptability of population individuals under disturbance conditions and select population individuals with high adaptability to environmental disturbance;

[0109] Gene mutation module, used to randomly change the parameter values ​​in individual gene encoding to perform gene mutation operations;

[0110] The iteration control module is used to control the optimization iteration process of the genetic algorithm and set the iteration stop condition.

[0111] The above-mentioned product can execute the RF connector micro-solder point control method based on the improved genetic algorithm described in Example 1, and has the corresponding functional modules and beneficial effects of the method. For technical details not described in detail in this embodiment, please refer to the specific contents provided in Example 1 of the present invention.

[0112] The present invention proposes a method for controlling micro-solder joints of RF connectors based on an improved genetic algorithm, aiming to solve the problem of accurate control of micro-solder joints in the manufacturing process of RF connectors by optimizing welding parameters and improving welding quality. Although other similar technical solutions can also optimize the control of micro-solder joints to some extent, they still have key defects and cannot replace the solution of the present invention.

[0113] First, traditional welding parameter optimization methods usually rely on manual experience or simple linear models. These methods are often unable to cope with the complex multivariate effects in the welding process and are difficult to achieve global optimization of welding parameters. In addition, although standard genetic algorithms have certain optimization capabilities, they have problems such as slow convergence and easy to fall into local optimality in practical applications. Especially in the face of dynamic changes and random interference in the welding process, it is difficult to respond and adjust quickly, resulting in unstable welding quality.

[0114] Secondly, existing welding control methods lack effective adaptive mechanisms when faced with environmental fluctuations and interference in actual production. They can usually only control welding through preset fixed parameters and cannot be dynamically adjusted according to real-time changes. This method ignores the randomness and complexity of the welding process, which can easily lead to fluctuations in solder joint quality and affect the performance and reliability of the connector. In addition, although some intelligent optimization algorithms can theoretically handle multivariable problems, their high computational complexity makes them difficult to apply in actual production, especially in scenarios that require real-time response and high-precision control.

[0115] In contrast, the present invention effectively overcomes the limitations of the above-mentioned technical solutions through various improvements to traditional genetic algorithms, including the introduction of innovative means such as adaptive genome adjustment mechanisms, multidimensional crossover operations, and local environmental adaptation factors. Through the adaptive genome adjustment mechanism, the algorithm can analyze and respond to parameter fluctuations in the welding process in real time, dynamically adjust the gene length and structure of the individuals in the population, and ensure that the optimization process always adapts to changes in the current environment. The multidimensional crossover operation further enhances the diversity of the population, prevents the algorithm from falling into the local optimum, and improves the optimization effect and efficiency. By introducing the local environmental adaptation factor, the algorithm can simulate and adapt to environmental interference during the actual welding process to ensure the stability of the quality of the solder joints. Finally, the present invention can dynamically adjust and optimize the welding parameters through a real-time feedback control mechanism, ensure the high precision and stability of the welding process, and significantly improve the reliability and performance of the RF connector.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may also be combined, the steps may be implemented in any order, and there are many other changes in different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for controlling micro solder joints of radio frequency connectors based on an improved genetic algorithm, characterized in that: Includes steps: S1. Collect welding parameters during welding and build key parameter data sets; S2, initializing the population of the genetic algorithm model, setting the number of individuals and individual gene codes of the initial population, wherein the individual gene codes are different combinations of parameter vectors in the key parameter data set; S3. Introduce an adaptive genome adjustment mechanism to dynamically adjust the gene length and structure of individuals in the population; S4, evaluate the fitness of the adjusted population individuals and select the population individuals with high fitness; S5. Use multidimensional crossover operation to cross the gene fragments of two or more population individuals; S6. Introduce local environmental adaptation factors and select population individuals with high adaptability to environmental disturbance; S7, introducing dynamic mutation probability, by randomly changing the parameter value in the individual gene code, performing gene mutation operation on the individuals in the population, and the mutation probability is adjusted in real time according to the welding quality; S8, re-evaluate the fitness of the individuals in the population after the gene mutation operation, and perform selection, crossover and mutation operations, repeating multiple iterations until the fitness evaluation function value or the number of iterations reaches a preset threshold; S9. Apply the optimized key parameters to the actual welding process, and further adjust and optimize the control parameters according to the real-time monitoring data through the feedback mechanism.

2. The RF connector micro solder joint control method according to claim 1, characterized in that: In step S1, the welding parameters include the instantaneous temperature T(t) of the welding point, the welding pressure P(t) applied to the welding point, the welding duration t w , welding instantaneous current I(t) and welding head moving speed v(t); the collected welding parameters are collected in chronological order to form a parameter vector X(t) = [T(t), P(t), t w ,I(t),v(t)], and finally form a key parameter data set containing n parameter vectors.

3. The RF connector micro solder joint control method according to claim 2, characterized in that: In step S2, an initial genetic code is generated for each individual in the population. The genetic code G of the i-th individual in the population is i It is expressed as: Among them, α n is a weight factor used to adjust the effect of parameters at different time points on gene coding; T(t n )、P(t n ) and I(t n ) are the time series of temperature, pressure and current, respectively, I mean is the average value of the current, σ is the standard deviation of the current; N is the number of individuals in the initial population.

4. The RF connector micro solder joint control method according to claim 2, characterized in that: Step S3 specifically includes the following steps: S301, monitoring the fluctuation of welding parameters during welding, and obtaining the fluctuation amplitude of welding parameters in real time ΔX(t)=[ΔT(t),ΔP(t),Δt w ,ΔI(t),Δv(t)], where the fluctuation amplitudes represent the changes in temperature, pressure, welding time, current and speed respectively; S302, calculating the adjustment factor λ(t)=1+α||ΔX(t)||||X(t)|| of the gene length of the individual in the population, the adjustment factor is used to dynamically adjust the length of the gene code of the individual, wherein α is a proportional constant, ||ΔX(t)|| represents the modulus of the fluctuation amplitude, and ||X(t)|| represents the modulus of the key parameter; S303, dynamically adjust the gene encoding length L of each individual in the population according to the adjustment factor λ(t) i (t) = λ(t)·L i,0 , L i (t) represents the gene coding length of the i-th individual at time t, L i,0 is the initial gene coding length of the i-th individual; S304. While adjusting the gene coding length, a structural adjustment factor β(t) is introduced to change the arrangement order of gene segments or add / delete specific gene segments; Among them, γ is the proportional constant, and β(t) represents the number of structural adjustments, which is used to adaptively adjust the individual gene structure according to real-time fluctuations.

5. The RF connector micro solder joint control method according to claim 1, characterized in that: In step S4, the fitness of the individuals in the population is evaluated by setting a fitness function, wherein the fitness function is constructed based on quality indicators of the solder joints, and the quality indicators include strength, dimensional consistency, and surface flatness.

6. The RF connector micro solder joint control method according to claim 1, characterized in that: In step S5, the gene codes of two or more selected individuals with high fitness are subjected to a multi-dimensional crossover operation, a crossover point is randomly selected in each gene dimension, and the gene fragments corresponding to the crossover point are exchanged to generate population individuals with new gene codes; after the multi-dimensional crossover operation, the fitness of the population individuals is evaluated respectively, and the individuals with high fitness are retained to form a new population for subsequent iterations.

7. The RF connector micro solder joint control method according to claim 1, characterized in that: In step S6, the local environmental adaptation factor is expressed as ζ(t)=1 / (1+||R(t)||·ε), which is used to measure the adaptability of the population individuals under interference conditions; wherein ε is the adjustment factor of the environmental noise, which is used to control the influence of the interference factor on the adaptability factor; ||R(t)|| is the modulus of the interference factor, and the interference factor R(t)=[r T (t),r I (t),r v (t)],r T (t) is the temperature fluctuation, r I (t) is the current disturbance, r v (t) is a random function of mechanical vibration; Calculate the fitness value of individuals in the population under interference conditions ζ (i) = Fitness(i)·ζ(t), where Fitness(i) is the original fitness value of individual i, and the population individuals with high environmental adaptability are retained.

8. The RF connector micro solder joint control method according to claim 1, characterized in that: In step S9, the welding parameters corresponding to the individual gene codes retained after optimization by the genetic algorithm are applied to the actual welding process, and welding quality index data are collected in real time during the welding process to form real-time feedback data; Compare the real-time feedback data Q(t) with the preset quality standard Q target Compare and calculate the deviation value ΔQ(t) = Q(t)-Q target ; Real-time optimization and adjustment of welding parameters based on deviation values: X adj (t) = X opt (t)+K·ΔQ(t), where X adj (t) is the adjusted welding parameter vector, X opt (t) is the welding parameter vector before adjustment, and K is the feedback gain matrix used to adjust the feedback response degree; feedback control and parameter adjustment operations are continuously performed during the welding process until the quality of the weld meets the preset quality standard; finally, the optimized and adjusted parameter data is stored in the welding parameter database for reference and further optimization of subsequent welding operations.

9. A radio frequency connector micro solder joint control system based on the method according to any one of claims 1 to 8, characterized in that: include: A data set building module, used for forming and storing parameter vectors according to welding parameters collected during the welding process; A genetic algorithm optimization module is used to construct the parameter vector in the module according to the data set, perform iterative optimization of welding parameters using the genetic algorithm, and output the optimal welding parameters; The feedback adjustment module is used to adaptively adjust the optimal welding parameters output by the genetic algorithm optimization module according to real-time monitoring data during the welding process.

10. The radio frequency connector micro solder joint control system according to claim 9, characterized in that: The genetic algorithm optimization module also includes: Population initialization module, used to initialize the number of individuals and individual gene codes of the population; Adaptive genome adjustment module, used to dynamically adjust the gene length and structure of individuals in the population; The fitness evaluation module is used to select and retain population individuals with high fitness according to the set fitness evaluation function; Genetic crossover module, used to perform multidimensional crossover operations on the genetic codes of individuals in the population; Environmental disturbance fitness assessment module, used to measure the adaptability of population individuals under disturbance conditions and select population individuals with high adaptability to environmental disturbance; Gene mutation module, used to randomly change the parameter values ​​in individual gene encoding to perform gene mutation operations; The iteration control module is used to control the optimization iteration process of the genetic algorithm and set the iteration stop condition.

Images