Multi-temperature-zone welding temperature intelligent control method and system based on deep learning

Abstract

The invention provides a multi-temperature-zone welding temperature intelligent control method and system based on deep learning, and relates to the technical field of welding temperature control, and the method comprises the steps: obtaining multi-temperature-zone real-time temperature data, carrying out the feature extraction and time sequence analysis, carrying out the temperature change prediction, dividing a control level, calculating a temperature deviation value, generating an initial control parameter, and calculating the temperature deviation value; and iteratively optimizing the parameter group through an adaptive weight algorithm, carrying out sliding window analysis on historical temperature data, and carrying out temperature compensation based on proportional integral control and a fuzzy rule.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of welding temperature control, in particular to a multi-temperature-zone welding temperature intelligent control method and system based on deep learning. BACKGROUND

[0002] In the electronic manufacturing industry, the welding process is a key link in the production of circuit boards, and the temperature control of multi-temperature-zone welding equipment directly affects the welding quality and product yield. Traditional welding temperature control mainly relies on preset temperature curves and simple feedback adjustment mechanisms. With the application of the concept of intelligent manufacturing in circuit board welding production lines, welding processes that set temperature target values, real-time collect actual temperature of temperature zones, and control output power of heating elements according to error proportional, integral, and differential operations are widely used. However, existing multi-temperature-zone welding temperature control technology still has problems such as relying only on real-time feedback adjustment, being unable to effectively cope with temperature lag, leading to large temperature fluctuations, easily causing temperature overshoot or undershoot when switching temperature zones, failing to optimize parameters for different thermal characteristics of temperature zones, being difficult to meet the fine requirements of temperature uniformity and stability in complex circuit board welding, having limited adaptability to interference factors, and requiring manual intervention to adjust parameters. SUMMARY

[0003] The present application provides a multi-temperature-zone welding temperature intelligent control method and system based on deep learning, which can at least solve some of the problems in the prior art.

[0004] In a first aspect, the present application provides a multi-temperature-zone welding temperature intelligent control method based on deep learning, comprising: Obtaining real-time temperature data of each temperature zone of a multi-temperature-zone welding equipment in a circuit board welding production line and performing feature extraction and time series analysis to obtain temperature change prediction results; Based on the temperature change prediction results, dividing each temperature zone into multiple control levels, calculating temperature deviation values of each control level and generating initial weight coefficients, and fitting temperature gradient change data by least squares method to iteratively calculate temperature compensation coefficients and initial control parameters of each temperature zone; Determining quantitative indicators according to the initial control parameters, calculating influence factors of the quantitative indicators and comprehensive evaluation scores of the initial control parameters by combining an adaptive weight algorithm, and iteratively optimizing the initial control parameters based on the comprehensive evaluation scores to obtain an optimized temperature zone control parameter group; The historical temperature data is analyzed based on the optimized temperature zone control parameter set in a sliding window, the temperature change rate and the temperature fluctuation standard deviation are calculated, when the temperature change rate exceeds a preset threshold or the temperature fluctuation standard deviation is greater than an allowed range, a temperature compensation value is calculated based on a proportional integral control algorithm and a heating power adjustment amount is determined based on a fuzzy control rule to generate a temperature compensation control parameter; The temperature control instruction is generated based on the optimized temperature zone control parameter set and the temperature compensation control parameter and is executed.

[0005] In an optional embodiment, Real-time temperature data of each temperature zone of a multi-temperature zone soldering device in a circuit board soldering production line is acquired and feature extraction and time series analysis are performed to obtain a temperature change prediction result including: Real-time temperature data of each temperature zone of a multi-temperature zone soldering device in a circuit board soldering production line is acquired and feature extraction is performed on the real-time temperature data to obtain a temperature change law; A temperature change trend matrix is established based on the temperature change law and time series analysis is performed on the temperature change trend matrix; A temperature change prediction value of each temperature zone is calculated based on the result of the time series analysis to obtain a temperature change prediction result.

[0006] In an optional embodiment, Based on the temperature change prediction result, each temperature zone is divided into multiple control levels, a temperature deviation value of each control level is calculated and an initial weight coefficient is generated, and a temperature gradient change data is fitted by a least square method, and a temperature compensation coefficient and an initial control parameter of each temperature zone are iteratively calculated including: Based on the temperature change prediction result, a temperature coupling degree between different temperature zones of a circuit board and a difference in thermal capacity of components are calculated, and each temperature zone is divided into multiple control levels, a temperature deviation value is calculated by subtracting a pre-set target temperature value from a pre-acquired real-time temperature value for each control level, and an initial weight coefficient is calculated based on the temperature deviation value; A temperature gradient change data is calculated based on a pre-acquired temperature change trend matrix, the temperature deviation value and the temperature gradient change data of each temperature zone are fitted by a least square method, a fitting result is obtained, and a temperature compensation coefficient corresponding to the current iteration is calculated based on the fitting result and a temperature compensation coefficient of a previous iteration; Based on the initial weight coefficient, the temperature deviation value and the temperature compensation coefficient, an initial control parameter is calculated by a pre-set temperature control performance evaluation function combined with a gradient descent method.

[0007] In an optional embodiment, Based on the initial weight coefficient, the temperature deviation value and the temperature compensation coefficient, an initial control parameter is calculated by a pre-set temperature control performance evaluation function combined with a gradient descent method, including: The initial weight coefficient, the temperature deviation value and the temperature compensation coefficient are combined as an input vector and input into a pre-set adaptive fuzzy neural network, and a fuzzy processing is performed by a Gaussian membership function to obtain a membership parameter; A wavelet transform is performed on the membership parameter to obtain a multi-scale feature coefficient, a hierarchical analysis decision tree is constructed based on the multi-scale feature coefficient, and a local extreme point under different scales is calculated recursively, a membership subset of the local extreme point is determined, mutual conditional information of the membership subset is calculated and is reconstructed by dimension reduction, and an optimized membership parameter is obtained by fusing the reconstructed membership subset; A fuzzy rule base is established according to the optimized membership parameter and a pre-set temperature control performance evaluation function, a frequent item set in the fuzzy rule base is mined by an Apriori algorithm and a support degree between different fuzzy rules is calculated, an association matrix is constructed based on the support degree and is singular value decomposed to construct a rule activation sequence, the fuzzy rule base is layered based on the rule activation sequence, a rule output value is obtained by processing the optimized membership parameter combined with a local linear function, and an optimized intermediate parameter is obtained by solving; When the temperature deviation value exceeds a pre-set deviation threshold, a compensation control parameter is calculated based on the optimized intermediate parameter combined with an exponential moving average and is output as an initial control parameter.

[0008] In an alternative embodiment, A quantitative index is determined according to the initial control parameter, an influence factor of the quantitative index and a comprehensive evaluation score of the initial control parameter are calculated combined with an adaptive weight algorithm, and an optimized temperature zone control parameter group is obtained by iteratively optimizing the initial control parameter based on the comprehensive evaluation score, including: A control performance index corresponding to the initial control parameter is obtained, a quantitative index is mapped combined with a pre-set mapping rule, an environmental temperature fluctuation value and a process parameter fluctuation value are obtained, and a state space vector is constructed combined with the quantitative index; A proportion corresponding to each quantitative index in the state space vector is calculated and an information entropy is determined, an influence factor corresponding to each quantitative index is calculated based on the information entropy, a temperature change sequence is collected within a pre-set time window and a parameter sensitivity index is calculated based on a time sequence feature and a nonlinear feature extracted; The initial control parameter is encoded as the position information of a firefly individual based on the adaptive firefly algorithm, a fluorescence intensity evaluation function is constructed based on the control performance index, and an attraction coefficient is initialized based on the environmental temperature fluctuation value, an optimal individual is determined, and the position information of the optimal individual is decoded as a parameter estimation value, a Lyapunov function value corresponding to the parameter estimation value is calculated, convergence is judged and system stability is determined, a comprehensive evaluation score is obtained by fuzzy evaluation of the quantization index based on the parameter sensitivity index and the system stability, a gradient of the initial control parameter with respect to the comprehensive evaluation score is calculated, and the initial control parameter is updated according to the gradient direction, and the updating is repeated until the comprehensive evaluation score converges, thereby obtaining an optimized temperature zone control parameter set.

[0009] In an optional embodiment, The historical temperature data is analyzed based on the optimized temperature zone control parameter set, the temperature change rate and the temperature fluctuation standard deviation are calculated, when the temperature change rate exceeds a preset threshold or the temperature fluctuation standard deviation is greater than an allowed range, a temperature compensation value is calculated based on a proportional integral control algorithm, and a heating power adjustment amount is determined in combination with a fuzzy control rule, and a temperature compensation control parameter is generated, including: A sliding window and a step length are set based on the optimized temperature zone control parameter set, historical temperature data collected and stored during the welding process is sampled to obtain a temperature sequence, and a temperature change rate and a temperature fluctuation standard deviation corresponding to the historical temperature data are calculated based on the temperature sequence; When the absolute value of the temperature change rate is greater than a preset temperature change rate threshold or the temperature fluctuation standard deviation is greater than a preset fluctuation standard deviation threshold, a temperature error is calculated by calculating the difference between a pre-set expected temperature and a current temperature, the temperature error is divided into a preheating section, a reflow section and a cooling section by a segmented adaptive Smith prediction compensation algorithm, compensation parameters are set for each section, the lag time of each section is calculated by a particle swarm algorithm, and a temperature compensation value is obtained by feedforward compensation on the lag time; The temperature compensation value is input into a preset fuzzy control rule, a domain division value corresponding to the optimized temperature zone control parameter set is calculated, and a temperature field topological feature map is constructed in combination with the corresponding thermal stress of the circuit board material and a pre-set deformation control constraint, the weight distribution of different regions of the temperature field is calculated based on the temperature field topological feature map, a heating power adjustment amount is calculated in combination with the temperature compensation value, and a temperature compensation control parameter is solved based on the heating power adjustment amount and the expected temperature.

[0010] In an optional embodiment, A temperature control instruction is generated based on the optimized temperature zone control parameter set and the temperature compensation control parameter, and the temperature control instruction is executed, including: The control cycle and temperature sampling interval are obtained from the optimized temperature zone control parameter group. Periodic temperature sampling is performed on each temperature zone of the circuit board to obtain temperature sampling values ​​and circuit board status parameters. The sampled temperature value is compared with the temperature compensation control parameters to obtain the sampled temperature deviation value. The sampled temperature deviation value is then corrected according to the optimized temperature zone control parameter group to obtain the temperature control adjustment amount. Based on the temperature control adjustment amount and the circuit board status parameters, a temperature control command containing heating power and heating time is generated, and the temperature control command is executed to adjust the temperature of the temperature zone.

[0011] A second aspect of the present invention provides a deep learning-based intelligent control system for multi-temperature zone welding temperature, comprising: The first unit is used to acquire real-time temperature data of each temperature zone of the multi-temperature zone welding equipment in the circuit board welding production line, and to perform feature extraction and time series analysis to obtain temperature change prediction results. The second unit is used to divide each temperature zone into multiple control levels based on the temperature change prediction results, calculate the temperature deviation value of each control level and generate initial weight coefficients, and fit the temperature gradient change data by the least squares method to iteratively calculate the temperature compensation coefficient and the initial control parameters of each temperature zone. The third unit is used to determine the quantitative index based on the initial control parameters, calculate the influence factor of the quantitative index and the comprehensive evaluation score of the initial control parameters by combining the adaptive weight algorithm, and iteratively optimize the initial control parameters based on the comprehensive evaluation score to obtain the optimized temperature zone control parameter set. The fourth unit is used to perform sliding window analysis on historical temperature data based on the optimized temperature zone control parameter group, calculate the temperature change rate and temperature fluctuation standard deviation. When the temperature change rate exceeds the preset threshold or the temperature fluctuation standard deviation is greater than the allowable range, the temperature compensation value is calculated based on the proportional-integral control algorithm and the heating power adjustment amount is determined in combination with the fuzzy control rules to generate temperature compensation control parameters. The fifth unit is used to generate and execute temperature control commands based on the optimized temperature zone control parameter set and temperature compensation control parameters.

[0012] A third aspect of the present invention provides an electronic device, comprising: A processor and a memory for storing processor-executable instructions, wherein the processor is configured to invoke instructions stored in the memory to perform the aforementioned method.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0014] In the present application, the real-time temperature data of the multi-temperature zone welding equipment is extracted and analyzed by deep learning, the accurate prediction of temperature change is realized, the temperature fluctuation trend is identified, the temperature deviation is effectively reduced, the welding quality and the first pass rate are improved, the control levels are divided, and the intelligent optimization of the control parameters of different temperature zones is realized by combining the adaptive weight algorithm, the temperature control precision is significantly improved, the dynamic adjustment of the temperature compensation control is realized, the temperature abnormal fluctuation can be quickly responded, the energy consumption is reduced, the production efficiency is improved, and the service life of the equipment is prolonged. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 A flowchart of the method for intelligent control of multi-temperature zone welding temperature based on deep learning according to an embodiment of the present application is shown in Figure 2 A flowchart of the generation of temperature compensation parameters of the method for intelligent control of multi-temperature zone welding temperature based on deep learning according to an embodiment of the present application is shown in DETAILED DESCRIPTION

[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0017] The technical solutions of the present application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described in some embodiments.

[0018] Figure 1 A flowchart of the method for intelligent control of multi-temperature zone welding temperature based on deep learning according to an embodiment of the present application is shown in Figure 1 As shown, the method comprises: Obtaining real-time temperature data of each temperature zone of a multi-temperature zone welding equipment in a circuit board welding production line and performing feature extraction and time series analysis to obtain temperature change prediction results; Based on the temperature change prediction results, each temperature zone is divided into multiple control levels, the temperature deviation value of each control level is calculated, the initial weight coefficient is generated, the least square method is used to fit the temperature gradient change data, and the temperature compensation coefficient and the initial control parameters of each temperature zone are iteratively calculated; determining a quantification index according to the initial control parameter, calculating an influence factor of the quantification index and a comprehensive evaluation score of the initial control parameter in combination with an adaptive weight algorithm, and iteratively optimizing the initial control parameter based on the comprehensive evaluation score to obtain an optimized temperature zone control parameter group; performing a sliding window analysis on the historical temperature data based on the optimized temperature zone control parameter group, calculating a temperature change rate and a temperature fluctuation standard deviation, calculating a temperature compensation value based on a proportional-integral control algorithm and determining a heating power adjustment amount in combination with a fuzzy control rule when the temperature change rate exceeds a preset threshold or the temperature fluctuation standard deviation is greater than an allowed range, and generating a temperature compensation control parameter; generating a temperature control instruction based on the optimized temperature zone control parameter group and the temperature compensation control parameter and executing the temperature control instruction.

[0019] In an alternative embodiment, obtaining real-time temperature data of each temperature zone of a multi-temperature zone soldering device in a circuit board soldering production line and performing feature extraction and time series analysis on the real-time temperature data to obtain a temperature change prediction result, including: obtaining real-time temperature data of each temperature zone of a multi-temperature zone soldering device in a circuit board soldering production line and performing feature extraction on the real-time temperature data to obtain a temperature change rule; establishing a temperature change trend matrix based on the temperature change rule and performing time series analysis on the temperature change trend matrix; calculating a temperature change prediction value of each temperature zone based on the result of the time series analysis to obtain a temperature change prediction result.

[0020] The temperature data is collected by a network of temperature sensors distributed in each temperature zone, with a sampling frequency of 1 per second. The data collected by the temperature sensors is transmitted to a data processing unit through an industrial Internet of Things module, forming a complete temperature data stream. For example, a certain soldering device has 5 temperature zones, labeled Z1 to Z5. During continuous production, the temperature data collected by each temperature zone forms a time series {T1,1, T1,2,..., T1,n} to {T5,1, T5,2,..., T5,n}, where Ti,j represents the temperature value of the i-th temperature zone at the j-th time point.

[0021] The collected real-time temperature data is subjected to feature extraction, which includes data preprocessing, trend analysis, and fluctuation feature recognition. In the data preprocessing stage, the original temperature data is subjected to filtering to eliminate random noise. The sliding average filtering method is used to take the average of the temperature data in each time window, with a window size of 10 sampling points. The filtered data is used to calculate the temperature change rate, which is the temperature difference between adjacent time points divided by the time interval. By analyzing the distribution characteristics of the temperature change rate, the acceleration mode of temperature change is identified. In addition, the temperature statistical characteristics of each temperature zone in different time periods are calculated, including mean, maximum, minimum, standard deviation, etc. For example, in a certain welding process, the average temperature of Z3 temperature zone in stable working state is 210°C, the standard deviation is 1.2°C, and the average temperature change rate is 0.05°C / s. These data reflect the temperature change characteristics of the temperature zone.

[0022] Based on the extracted temperature change characteristics, a temperature change trend matrix is constructed, which is a three-dimensional data structure containing temperature zone number, time point, and temperature feature. For each temperature zone, key parameters representing temperature change trends are extracted based on the historical temperature data change mode, including trend slope, periodic fluctuation amplitude, and phase information. The key parameters are arranged in time sequence in the matrix to form a complete trend description. The construction of the temperature change trend matrix takes into account the mutual influence between temperature zones, and the temperature conduction relationship between different temperature zones is determined through correlation analysis and quantified as a temperature transfer coefficient. For example, in practical application, the temperature change of Z2 temperature zone usually affects Z3 temperature zone after 3-5 seconds, with an influence coefficient of about 0.7, which is encoded into the trend matrix.

[0023] The temperature change trend matrix is subjected to time series analysis, which adopts a deep learning method, specifically a long short-term memory network model. The long short-term memory network model includes an input layer, multiple long short-term memory layers, and an output layer. The input layer receives normalized temperature trend matrix data, each long short-term memory layer contains 64 neurons to capture the time dependence of temperature change, and the historical temperature data is used as the training set to optimize the model parameters through stochastic gradient descent method, with mean square error as the loss function. To prevent overfitting, the dropout technique is used with a dropout rate of 0.2. After training, the average prediction error on the test set is less than 1.5°C, meeting the precision requirements of industrial welding.

[0024] The temperature change prediction value of each temperature zone is calculated based on the result of the time series analysis, and the prediction process is divided into short-term prediction and long-term prediction. The short-term prediction focuses on the temperature change in the next 30 seconds, mainly for real-time control adjustment; the long-term prediction focuses on the temperature trend in the next 5 minutes, for process optimization and abnormal warning, and the prediction algorithm considers the current temperature state, historical change trend and environmental factors. For each temperature zone, the temperature prediction value at each time point in the future is calculated based on the output of the long and short term memory network and combined with the temperature zone characteristic parameters. The calculation of the prediction value considers the heat conduction effect between temperature zones, and the temperature change of adjacent temperature zones will affect the future temperature of the target temperature zone in a certain proportion. In an actual application case, the current temperature of the Z4 temperature zone of a certain welding equipment is 250°C, based on the prediction algorithm, it is predicted that the temperature of the temperature zone will stabilize in the range of 248-252°C in the next 30 seconds, and will slowly decrease to about 245°C after 3 minutes.

[0025] In this embodiment, by extracting features from real-time temperature data of each temperature zone, representative temperature change characteristics can be extracted from the original data, improving the effectiveness and usability of temperature information representation. By establishing a temperature change trend matrix based on the temperature change law and performing time series analysis, the dynamic characteristics of temperature evolution over time can be more accurately reflected, improving the accuracy of temperature change law modeling and prediction. By calculating the temperature change prediction value of each temperature zone based on the time series analysis result, the temperature change trend can be obtained, and the prediction foresight and control stability can be improved in the temperature control process.

[0026] In an alternative embodiment, Based on the temperature change prediction result, each temperature zone is divided into multiple control levels, the temperature deviation value of each control level is calculated and the initial weight coefficient is generated, and the least square method is used to fit the temperature gradient change data, and the temperature compensation coefficient and the initial control parameter of each temperature zone are calculated iteratively, including: Based on the temperature change prediction result, the temperature coupling degree between different temperature zones of the circuit board and the difference in thermal capacity of the components are calculated, and each temperature zone is divided into multiple control levels. For each control level, the difference between the pre-set target temperature value and the pre-acquired real-time temperature value is obtained to obtain the temperature deviation value, and the initial weight coefficient is calculated based on the temperature deviation value; The temperature gradient change data is calculated based on the pre-acquired temperature change trend matrix, the temperature deviation value and the temperature gradient change data of each temperature zone are fitted by the least square method, the fitting result is obtained, and the temperature compensation coefficient corresponding to the current iteration is calculated combined with the pre-acquired temperature compensation coefficient of the last iteration; Based on the initial weight coefficient, the temperature deviation value and the temperature compensation coefficient, an initial control parameter is calculated by a pre-set temperature control performance evaluation function combined with a gradient descent method.

[0027] After obtaining the temperature change prediction result, the temperature coupling degree between different temperature zones of the circuit board is calculated, which represents the heat conduction influence intensity between different temperature zones. In the calculation, a correlation analysis method is adopted, and by comparing the time series data of temperature change of each temperature zone, a correlation coefficient matrix of temperature change is calculated. For example, in the welding process of a certain circuit board, the temperature coupling matrix of five temperature zones Z1 to Z5 shows that the coupling coefficient between Z2 and Z3 is 0.85, which means that the temperature change of Z2 will directly affect the temperature state of Z3, and the coupling degree is high; while the coupling coefficient between Z1 and Z5 is only 0.32, which means that the heat influence between the remote temperature zones is weak. At the same time, the calculation of the difference in thermal capacity of components is based on the thermal characteristic parameters of different types of components on the circuit board. By analyzing the circuit board design drawing and the distribution of components, combined with the density distribution and material characteristics of components in each temperature zone, the equivalent thermal capacity of each temperature zone is calculated. For example, the thermal capacity of the area containing large capacitors or transformers is as high as 245 J / K, while the thermal capacity of the area containing only small resistors and diodes may be only 85 J / K.

[0028] Based on the temperature coupling degree and the difference in thermal capacity, each temperature zone is divided into multiple control levels. The control level division adopts a clustering analysis method, and comprehensively considers the coupling relationship of the temperature zones, the thermal capacity characteristics and the importance of temperature control. Generally, the temperature zones are divided into three levels: key control layer, coordination control layer and auxiliary control layer. The key control layer includes the temperature zones with the highest requirement for welding quality, such as the peak temperature zone of reflow; the coordination control layer includes the temperature zones with high coupling relationship with the key temperature zones; the auxiliary control layer includes the remaining temperature zones. For example, the level division result of a certain six-temperature-zone welding equipment is: Z3 and Z4 belong to the key control layer, with a control accuracy requirement of ±1℃; Z2 and Z5 belong to the coordination control layer, with a control accuracy requirement of ±2℃; Z1 and Z6 belong to the auxiliary control layer, with a control accuracy requirement of ±3℃. After the control level division, the temperature deviation value of each level is calculated, which is the difference between the pre-set target temperature value and the real-time temperature value. Taking the key control layer Z3 as an example, the target temperature is set to 235℃, and the real-time temperature is 232℃, so the temperature deviation value is 3℃.

[0029] The initial weight coefficient is calculated based on the temperature deviation value, which reflects the influence degree of temperature deviation in different temperature zones on the overall control target. The weight is allocated according to the weighted principle, and the weight is related to the absolute value of the temperature deviation value and the control level to which the temperature zone belongs. The weight base of the key control layer is 1.0, the weight base of the coordination control layer is 0.7, and the weight base of the auxiliary control layer is 0.4. The actual weight coefficient also needs to be adjusted according to the size of the temperature deviation. The larger the deviation, the higher the weight coefficient. The initial weight coefficient of each temperature zone is obtained by multiplying the basic weight by the normalized result of the temperature deviation value. For example, the temperature deviation of the key control layer Z3 is 3°C, and after normalization, it is 0.6. Therefore, the initial weight coefficient of the key control layer Z3 is 0.6*1.0=0.6. The temperature deviation of the coordination control layer Z2 is 2°C, and after normalization, it is 0.4. Therefore, the initial weight coefficient of the coordination control layer Z2 is 0.4*0.7=0.28.

[0030] The temperature gradient change data is calculated based on the pre-acquired temperature change trend matrix. The temperature gradient change data represents the change amount of temperature per unit time, which is obtained by calculating the temperature difference between adjacent time points in the temperature change trend matrix. For a system with a sampling frequency of 1 Hz, the temperature gradient is the change amount of temperature per second. For example, the temperature record of the Z4 temperature zone in the last 10 seconds is {245.2, 246.1, 246.8, 247.3, 247.7, 248.0, 248.2, 248.3, 248.4, 248.5}°C. The calculated temperature gradient sequence is {0.9, 0.7, 0.5, 0.4, 0.3, 0.2, 0.1, 0.1, 0.1}°C / s, which shows a gradually decreasing trend, indicating that the temperature rising rate gradually slows down and approaches a stable state.

[0031] The relationship model between the temperature change trend and the control demand is obtained by fitting the temperature deviation value and the temperature gradient change data using the least squares method. In the fitting process, the temperature deviation value is taken as the dependent variable, and the temperature gradient change data is taken as the independent variable to construct a linear or nonlinear fitting equation. By minimizing the sum of squared residuals, the best fitting parameters are determined. In a certain welding equipment example, the fitting result of the Z3 temperature zone shows that there is a clear quadratic relationship between the temperature deviation value and the temperature gradient. The fitting parameters are: constant term 1.25, first-order term coefficient 0.78, and second-order term coefficient -0.15. The fitting accuracy R 2 value is 0.92, indicating good fitting effect. The temperature compensation coefficient corresponding to the current iteration is calculated by combining the fitting result with the temperature compensation coefficient of the last iteration. The iterative calculation of the temperature compensation coefficient uses the weighted average method. The compensation coefficient of the current iteration = the compensation coefficient of the last iteration * 0.7 + the newly calculated compensation coefficient * 0.3. For example, the compensation coefficient of the Z3 temperature zone in the last iteration is 1.15, and the newly calculated compensation coefficient is 1.23. Therefore, the compensation coefficient of the current iteration = 1.15*0.7 + 1.23*0.3 = 1.174.

[0032] Based on the initial weight coefficient, the temperature deviation value and the temperature compensation coefficient, the initial control parameter is calculated by a pre-set temperature control performance evaluation function combined with a gradient descent method. The temperature control performance evaluation function comprehensively considers control accuracy, response speed and stability. The control accuracy is represented by the square of the temperature deviation, the response speed is represented by the control time length, and the stability is represented by the temperature fluctuation amplitude. These three indexes are weighted and summed according to the ratio of 5:3:2 to form the value of the evaluation function. The gradient descent method is used to find the control parameter that makes the evaluation function minimum. Exemplarily, the initial value and learning rate of the control parameter are set, and the parameter is updated along the negative gradient direction of the evaluation function through multiple iterations until convergence. In a certain welding temperature zone control case, the initial control parameters are set as P=2.5, I=0.8, and D=0.3. After 50 gradient descent iterations, the parameters are optimized as P=2.78, I=0.95, and D=0.42, and the control performance is significantly improved.

[0033] In this embodiment, by calculating the temperature coupling degree between temperature zones and the difference in thermal capacity of components based on the temperature change prediction result, and dividing the temperature zones into multiple control levels, the temperature control can be more in line with the physical characteristics of the production process, and the refinement degree of layered regulation and control can be improved. The temperature deviation value is obtained by calculating the difference between the target temperature value and the real-time temperature value, and further combined with the temperature gradient change data to fit using the least squares method, which can more accurately depict the relationship between the temperature change trend and the deviation, and improve the accuracy and stability of the compensation calculation. By combining the initial weight coefficient, the temperature deviation value and the temperature compensation coefficient, and calculating the initial control parameter using the temperature control performance evaluation function and the gradient descent method, the convergence speed can be accelerated while optimizing the control parameter, and the response efficiency and overall stability of the temperature control can be improved.

[0034] In an alternative embodiment, Based on the initial weight coefficient, the temperature deviation value and the temperature compensation coefficient, the initial control parameter is calculated by a pre-set temperature control performance evaluation function combined with a gradient descent method. The temperature control performance evaluation function comprehensively considers control accuracy, response speed and stability. The control accuracy is represented by the square of the temperature deviation, the response speed is represented by the control time length, and the stability is represented by the temperature fluctuation amplitude. These three indexes are weighted and summed according to the ratio of 5:3:2 to form the value of the evaluation function. The gradient descent method is used to find the control parameter that makes the evaluation function minimum. Exemplarily, the initial value and learning rate of the control parameter are set, and the parameter is updated along the negative gradient direction of the evaluation function through multiple iterations until convergence. In a certain welding temperature zone control case, the initial control parameters are set as P=2.5, I=0.8, and D=0.3. After 50 gradient descent iterations, the parameters are optimized as P=2.78, I=0.95, and D=0.42, and the control performance is significantly improved. The initial weight coefficient, the temperature deviation value and the temperature compensation coefficient are combined as an input vector and input into a pre-set adaptive fuzzy neural network, and a Gaussian membership function is used for fuzzy processing to obtain a membership parameter; The membership parameter is wavelet transformed to obtain multi-scale feature coefficients, an analytic hierarchy process decision tree is constructed based on the multi-scale feature coefficients, and local extreme points at different scales are recursively calculated to determine a membership subset of the local extreme points, calculate mutual conditional information of the membership subset, and perform dimensionality reduction reconstruction to obtain an optimized membership parameter by fusing the reconstructed membership subset; The fuzzy rule base is established according to the optimized membership parameters and a preset temperature control performance evaluation function, a frequent item set in the fuzzy rule base is mined through an Apriori algorithm, and support degrees between different fuzzy rules are calculated, an association matrix is constructed based on the support degrees, singular value decomposition is performed on the association matrix to construct a rule activation sequence, the fuzzy rule base is layered based on the rule activation sequence, and a rule output value is obtained by processing the optimized membership parameters in combination with a local linear function, and an optimized intermediate parameter is obtained by solving. When the temperature deviation value exceeds a preset deviation threshold, a compensation control parameter is calculated based on the optimized intermediate parameter in combination with an exponential moving average, and the compensation control parameter is output as an initial control parameter.

[0035] The initial weight coefficient, the temperature deviation value and the temperature compensation coefficient are combined as an input vector. Taking a five-zone welding device as an example, the input vector of each zone includes three elements: the initial weight coefficient (range 0-1), the temperature deviation value (range ±10℃) and the temperature compensation coefficient (range 0.8-1.5). For zone Z3, the actual input vector may be [0.75, 2.3, 1.12], indicating that the initial weight coefficient is 0.75, the temperature deviation is 2.3℃ and the compensation coefficient is 1.12. The combined input vector is input into a pre-set adaptive fuzzy neural network for processing. The adaptive fuzzy neural network adopts a five-layer structure: an input layer, a fuzzification layer, a rule layer, a normalization layer and an output layer. The input layer receives original data, and the number of nodes is equal to the dimension of the input vector; the fuzzification layer performs fuzzy transformation on the input, and a Gaussian membership function is used. For each input variable, five fuzzy language variables are set: very low, low, medium, high and very high. The center point and the width parameter of the Gaussian membership function are determined through historical data training. For example, for the temperature deviation value, the center point of the Gaussian function of "medium" is set to 0℃, and the width parameter is 2℃. After fuzzification, the input vector [0.75, 2.3, 1.12] is converted into a membership parameter matrix, which represents the membership degree of each input in different fuzzy sets. For example, the membership degree of the temperature deviation 2.3℃ in "medium" is 0.33, and the membership degree in "high" is 0.61.

[0036] The membership parameters are wavelet-transformed, and the wavelet transform uses a Haar wavelet basis function to perform three-level decomposition on the membership parameters to obtain low-frequency approximation coefficients and high-frequency detail coefficients. The wavelet transform is implemented through a filter bank, and a low-pass filter extracts the smooth trend of the signal, and a high-pass filter extracts local detail changes. The wavelet transform is applied to each row of the membership parameter matrix to obtain multi-scale feature coefficients. For example, wavelet-transforming the membership parameters of temperature zone Z3 yields three-level decomposition coefficients including: first-level approximation coefficients [0.65, 0.72, 0.58], first-level detail coefficients [0.12, -0.08, 0.05], second-level detail coefficients [0.03, -0.02, 0.01], and third-level detail coefficients [0.01, 0.00, -0.01]. A hierarchical analysis decision tree is constructed based on the multi-scale feature coefficients, and the decision tree uses the CART algorithm, each internal node represents a binary test on the feature coefficients, and the leaf node represents a local extreme region. The generation process of the decision tree uses a recursive method, and the optimal split feature and threshold are determined by calculating the Gini coefficient. For example, the first layer split can select the first element of the first-level approximation coefficients, and the threshold is 0.6; the second layer split selects the second element of the first-level detail coefficients, and the threshold is -0.05. Through recursive calculation of the decision tree, local extreme points at different scales are obtained, such as [0.65, 0.12, 0.03] and [0.72, -0.08, -0.02].

[0037] The membership subsets of the local extreme points are determined, and for each local extreme point, its corresponding membership subset is determined according to its position in the original membership space. The membership subset is a submatrix of the original membership parameters, and reflects the fuzzy characteristics of the local region. For example, the membership subset corresponding to the extreme point [0.65, 0.12, 0.03] can be a 3x3 submatrix in the upper left corner of the original membership matrix. The mutual conditional information of the membership subset is calculated to evaluate the information redundancy within the subset. The mutual conditional information is obtained by calculating the information entropy and joint entropy between the elements in the subset, and the subset with high information redundancy needs to be processed for dimension reduction. Dimensionality reduction reconstruction uses a principal component analysis method to retain principal components with a cumulative contribution rate of 95%. For example, after principal component analysis of a certain membership subset, the contribution rates of the first two principal components are 87% and 9%, for a total of 96%, so the two principal components are retained for reconstruction. The reconstructed membership subsets are fused to obtain optimized membership parameters. The fusion process uses a weighted average method, and the weight is proportional to the amplitude of the corresponding extreme point of the subset. For temperature zone Z3, the dimension of the optimized membership parameter matrix is reduced from the original 3x15 to 3x8, and the computational complexity is reduced by 47%.

[0038] The fuzzy rule base is established according to the optimized membership parameters and a preset temperature control performance evaluation function. The temperature control performance evaluation function comprehensively considers three indexes of overshoot, rise time and steady-state error, and is expressed in the form of weighted sum. The fuzzy rules are in the form of "if-then", for example, "if the temperature deviation is large and the change rate is positive, then the control output is large". The initial rule base is generated by expert experience and historical data analysis, and contains about 50 rules. The Apriori algorithm is applied to the rule base for frequent item set mining, and the minimum support threshold is set to 0.15 to mine the frequently occurring rule combinations. For example, it is found through analysis that the support of the two items "large temperature deviation" and "large control output" in the rule base is 0.35, which is higher than the threshold, and is determined as a frequent item set. The support matrix is calculated between different fuzzy rules. The association matrix is constructed based on the support matrix, and the element value of the association matrix represents the association strength between rules.

[0039] The association matrix is subjected to singular value decomposition to extract the main feature vectors and construct the rule activation sequence. The singular value decomposition decomposes the association matrix into the product of three matrices, and the retained feature dimension is determined by analyzing the singular value size. For example, the singular value sequence of a certain association matrix is [5.2, 2.8, 1.5, 0.9, 0.3], and the first three feature vectors are selected, which cover 86% of the information amount. The rule activation sequence is the priority order of rule triggering, which is obtained according to the element value of the feature vector. Based on the rule activation sequence, the fuzzy rule base is layered, and the rule base is divided into three layers of high priority, medium priority and low priority. The high-priority rules (such as 10 rules) are first evaluated in the control process, and if the triggering condition is met, the output is directly output; otherwise, the medium-priority rules (such as 25 rules) are evaluated, and finally the low-priority rules (such as 15 rules) are evaluated.

[0040] The rule output value is obtained by combining the local linear function processing of the optimized membership parameters, and the optimized intermediate parameter is solved. The local linear function divides the fuzzy input space into multiple local regions, and a linear function is used to describe the relationship between input and output in each region. For temperature zone Z4, when the temperature deviation is in the range of [-2, 2] °C, a linear function with a slope of 0.8 is used; when the temperature deviation is in the range of [2, 5] °C, a linear function with a slope of 1.2 is used. The output values of each rule are fused by the weighted average method, and the weight is equal to the triggering strength of the rule. The final control decision value is 0.68, which is converted into the optimized intermediate parameter.

[0041] When the temperature deviation value exceeds the preset deviation threshold, a compensation control mechanism is triggered. The preset deviation threshold is usually set to ±2% of the target temperature. For example, when the target temperature is 240°C, the deviation threshold is ±4.8°C. After exceeding the threshold, the compensation control parameter is calculated based on the optimization intermediate parameter combined with the exponential moving average method, which gives higher weight to recent data, and the smoothing coefficient is set to 0.2. The calculation formula is: current compensation value = last period compensation value × 0.8 + optimization intermediate parameter × 0.2. For example, if the last period compensation value is 1.25 and the current optimization intermediate parameter is 1.35, the calculated compensation control parameter is 1.27. This parameter is output as the initial control parameter of the PID controller, used to adjust the heating power or cooling intensity, to achieve precise temperature control.

[0042] In this embodiment, by combining the initial weight coefficient, temperature deviation value and temperature compensation coefficient into the adaptive fuzzy neural network and performing fuzzy processing combined with the Gaussian membership function, the temperature control characteristics can be more flexibly expressed in an uncertain environment, and the adaptability to complex nonlinear relationships is improved. By performing wavelet transform on the membership parameters and extracting multi-scale features, and combining the analytic hierarchy decision tree to recursively calculate the local extreme points and membership subsets, the detailed features of temperature changes can be captured at different scales, and the accuracy and robustness of membership parameter optimization are improved. By using the Apriori algorithm to mine frequent item sets from the fuzzy rule base and combining singular value decomposition to construct the rule activation sequence, the selection of redundant rules and the strengthening of effective rules can be realized.

[0043] In an alternative embodiment, According to the initial control parameter, a quantitative index is determined, an influence factor of the quantitative index and a comprehensive evaluation score of the initial control parameter are calculated by combining an adaptive weight algorithm, and the initial control parameter is iteratively optimized based on the comprehensive evaluation score to obtain an optimized temperature zone control parameter group, which includes: The control performance index corresponding to the initial control parameter is obtained, and a quantitative index is mapped by combining a pre-set mapping rule. The environmental temperature fluctuation value and the process parameter fluctuation value are obtained, and a state space vector is constructed by combining the quantitative index; The proportion of each quantitative index in the state space vector is calculated and the information entropy is determined, the influence factor corresponding to each quantitative index is calculated based on the information entropy, and the parameter sensitivity index is calculated by collecting the temperature change sequence in the preset time window and extracting the time sequence features and nonlinear features; The initial control parameters are encoded as the position information of firefly individuals based on the adaptive firefly algorithm, a fluorescence intensity evaluation function is constructed based on the control performance index, and an attractive coefficient is initialized based on the environmental temperature fluctuation value. The optimal individual is determined, and the position information of the optimal individual is decoded as the parameter estimation value. The Lyapunov function value corresponding to the parameter estimation value is calculated, the convergence is judged, and the system stability is determined. The comprehensive evaluation score is obtained by fuzzy evaluation of the quantization index based on the parameter sensitivity index and the system stability. The gradient of the initial control parameters with respect to the comprehensive evaluation score is calculated, and the initial control parameters are updated according to the gradient direction. The update is repeated until the comprehensive evaluation score converges, and the optimized temperature zone control parameter set is obtained.

[0044] The control performance index corresponding to the initial control parameters is obtained, including overshoot, rise time, settling time, and steady-state error. Taking a certain six-temperature zone welding equipment as an example, after applying the initial control parameters P=2.8, I=0.4, D=0.2 to temperature zone Z2, the measured overshoot is 1.8%, the rise time is 15 seconds, the settling time is 35 seconds, and the steady-state error is 0.5°C. The performance index is mapped to the quantization index by combining the mapping rule set in advance. The mapping rule adopts a piecewise linear mapping method, and each performance index is normalized to the [0, 1] interval. For example, the overshoot mapping rule is: when the overshoot is less than 1%, the quantization value is 0.9-1.0; when the overshoot is between 1%-3%, the quantization value is 0.7-0.9; when the overshoot is between 3%-5%, the quantization value is 0.5-0.7; and so on. According to this rule, the overshoot of 1.8% is mapped to the quantization value of 0.84. Similarly, the rise time of 15 seconds is mapped to 0.76, the settling time of 35 seconds is mapped to 0.68, and the steady-state error of 0.5°C is mapped to 0.82. The environmental temperature fluctuation value and the process parameter fluctuation value are obtained. The environmental temperature fluctuation range in the past 4 hours is 23.5°C to 25.8°C, and the standard deviation is 0.7°C, which is recorded by the plant temperature sensor. The process parameter fluctuation value is obtained by monitoring the equipment operating parameters, such as the conveyor belt speed fluctuation of ±3% of the standard value and the solder supply amount fluctuation of ±2% of the standard value. The state space vector is constructed by combining the quantization index, the environmental temperature fluctuation value, and the process parameter fluctuation value. For temperature zone Z2, the state space vector is [0.84, 0.76, 0.68, 0.82, 0.7, 0.03, 0.02], the first four elements are the control performance quantization index, the fifth element is the quantization value of the environmental temperature fluctuation, and the last two elements are the quantization values of the process parameter fluctuation.

[0045] The importance of each quantified index in the state space vector is determined by the proportion of the index, which is calculated using the analytic hierarchy process (AHP) to determine the relative importance of each index. The judgment matrix is established based on expert experience and historical data analysis, and the rationality is ensured by consistency check. For welding temperature control, the importance of steady-state error and overshoot is usually higher than that of rise time and settling time. After calculation, the proportion of each index in temperature zone Z2 is: overshoot 0.28, rise time 0.15, settling time 0.12, steady-state error 0.30, environmental temperature fluctuation 0.10, and process parameter fluctuation 0.05. The information entropy is calculated based on the proportion, which reflects the uncertainty of the index distribution. When calculating the information entropy, the proportion of each index is normalized as a probability distribution, and the entropy value of the probability distribution is calculated. For example, the information entropy of each index in temperature zone Z2 is 1.62. The influence factor of each quantified index is calculated based on the information entropy, which represents the influence degree of the index on the overall performance. The calculation method is to multiply the index proportion by the normalized value of the information entropy. For example, the influence factor of overshoot is 0.28 x 1.62 ÷ 4.5 = 0.10, where 4.5 is the normalization coefficient. The temperature change sequence is collected within a preset time window, and the features are extracted, with the time window set to 120 seconds and the sampling frequency set to 1 Hz, a total of 120 temperature data points are collected. The sequence data collected is extracted for timing features, including statistical features such as mean, standard deviation, skewness, kurtosis, and frequency domain features extracted by Fourier transform. For example, the timing features of temperature zone Z4 include: mean 245.8°C, standard deviation 1.2°C, skewness 0.15, kurtosis 2.8, and main frequency component near 0.05 Hz. The nonlinear feature extraction uses phase space reconstruction and complexity analysis methods to calculate features such as Lyapunov exponent, entropy value, and fractal dimension. For example, the Lyapunov exponent of temperature zone Z4 is 0.024, indicating that it has certain chaotic characteristics. The parameter sensitivity index is calculated by combining the timing features and nonlinear features, which represents the influence degree of the control parameter change on the system performance. The calculation method is to make a small perturbation to the control parameter and observe the change rate of the performance index. For example, when the P parameter increases by 5%, the overshoot increases by 12%, indicating that the sensitivity of overshoot to P parameter is 2.4.

[0046] The adaptive firefly algorithm is used to optimize the control parameters. The firefly algorithm is a swarm intelligence-based optimization algorithm that finds the optimal solution by simulating the aggregation behavior of fireflies. The initial control parameters are encoded as the position information of firefly individuals. For a PID controller, each firefly individual is represented as a three-dimensional vector [P, I, D]. For example, the initial population contains 30 firefly individuals, and the position of each individual is randomly initialized within the allowed range. For temperature zone Z3, one individual in the initial population may have a position of [2.5, 0.4, 0.2]. A fluorescence intensity evaluation function is constructed based on the control performance indicators. The evaluation function adopts a weighted sum form, and the weight of each performance indicator is proportional to its influence factor. For example, the evaluation function is: 0.35 x overshoot score + 0.15 x rise time score + 0.15 x settling time score + 0.35 x steady-state error score. The fluorescence intensity is proportional to the evaluation function value, and the higher the evaluation function value, the stronger the fluorescence intensity. The attraction coefficient is initialized based on the environmental temperature fluctuation value. The attraction coefficient represents the mutual influence strength between firefly individuals. The attraction coefficient is inversely proportional to the environmental temperature fluctuation. The larger the environmental temperature fluctuation, the smaller the attraction coefficient, indicating that the individual needs more autonomous exploration in a complex environment. For example, when the standard deviation of the environmental temperature fluctuation is 0.7°C, the attraction coefficient is initialized to 0.3.

[0047] During the iterative optimization process, the optimal individual is determined and its position information is decoded into parameter estimates. In each iteration, all firefly individuals move towards individuals with higher fluorescence intensity, and the movement distance is related to the attraction coefficient and the distance between individuals. After 50 iterations, the optimal individual position converges to [2.35, 0.55, 0.25], which is decoded into parameter estimates P = 2.35, I = 0.55, and D = 0.25. The Lyapunov function value corresponding to the parameter estimates is calculated to judge the stability of the system. The Lyapunov function is constructed based on the system energy, and the change trend of the function value reflects the stability of the system. Through numerical simulation, the Lyapunov function value is calculated to be -0.35, and the function derivative is -0.08, indicating that the system is asymptotically stable. The convergence is judged and the stability of the system is determined. The convergence judgment standard is that the optimal value changes less than 0.01 for 10 consecutive iterations. The stability of the system is classified according to the Lyapunov function value. Less than -0.3 is high stability, -0.3 to -0.1 is medium stability, and greater than -0.1 is low stability. For example, the Lyapunov function value of temperature zone Z3 is -0.35, indicating that the system has high stability.

[0048] The fuzzy evaluation is based on the parameter sensitivity index and system stability. The double-layer evaluation model is used in the fuzzy evaluation. The first layer evaluates the satisfaction of each performance index, and the second layer evaluates the overall performance. The five-level fuzzy set is used in the evaluation: excellent, good, general, poor, and very poor. For the temperature zone Z3, the fuzzy evaluation results are: overshoot-good (0.7), rise time-excellent (0.85), settling time-good (0.75), and steady-state error-excellent (0.8). The comprehensive evaluation score is 0.78, which belongs to the "good" level. The gradient of the initial control parameters is calculated to reflect the sensitivity of the score to the parameter changes. The numerical differentiation method is used to calculate the change rate of the evaluation score by adding a small perturbation to each parameter. For example, when the P parameter increases by 0.1, the comprehensive evaluation score changes by -0.05, indicating a gradient of -0.5; when the I parameter increases by 0.1, the comprehensive evaluation score changes by 0.03, indicating a gradient of 0.3; and when the D parameter increases by 0.1, the comprehensive evaluation score changes by 0.02, indicating a gradient of 0.2. The initial control parameters are updated according to the gradient direction, and the update formula is: new parameter = old parameter + learning rate × gradient. The learning rate is set to 0.2, and the updated parameters are P = 2.35 - 0.2 × (-0.5) = 2.45, I = 0.55 + 0.2 × 0.3 = 0.61, and D = 0.25 + 0.2 × 0.2 = 0.29.

[0049] The parameter updating process is repeated until the comprehensive evaluation score converges. The convergence criterion is that the score change is less than 0.005 for 5 consecutive iterations. After 15 iterations, the control parameters of the temperature zone Z3 are optimized as P = 2.42, I = 0.58, and D = 0.26, and the comprehensive evaluation score reaches 0.83, which belongs to the "excellent" level. The optimization process is repeated for all temperature zones to obtain the optimized control parameter set for each temperature zone. For example, the optimized parameter set for the six-temperature-zone welding equipment is: Z1 [1.85, 0.38, 0.15], Z2 [2.20, 0.45, 0.22], Z3 [2.42, 0.58, 0.26], Z4 [2.35, 0.52, 0.24], Z5 [2.10, 0.48, 0.20], and Z6 [1.75, 0.35, 0.15].

[0050] In this embodiment, the control performance index is converted into a quantitative index by combining the mapping rule, and the state space vector is constructed together with the environmental temperature fluctuation value and the process parameter fluctuation value, which can comprehensively represent the multi-dimensional temperature control state, improve the description ability of the system operation characteristics, highlight the dominant role of key parameters on the control performance by calculating the influence factor based on information entropy and combining the time sequence and the parameter sensitivity index extracted by nonlinear characteristics, thereby improving the pertinence and effectiveness of parameter optimization, and the global optimization ability can be enhanced and the local optimum can be avoided by using the adaptive firefly algorithm to encode and iteratively optimize the initial control parameters, combining the fluorescence intensity evaluation function and the attraction coefficient to dynamically adjust the search process, thereby improving the optimization efficiency and the stability of parameter convergence.

[0051] In an alternative embodiment, Based on the optimized temperature zone control parameter group, the historical temperature data is analyzed by sliding window, the temperature change rate and the temperature fluctuation standard deviation are calculated, when the temperature change rate exceeds the preset threshold or the temperature fluctuation standard deviation is greater than the allowed range, the temperature compensation value is calculated based on the proportional integral control algorithm and the heating power adjustment amount is determined combined with the fuzzy control rule, and the temperature compensation control parameter is generated, including: Based on the optimized temperature zone control parameter group, a sliding window and a step length are set, the historical temperature data collected and stored during the welding process is sampled to obtain a temperature sequence, and the temperature change rate and the temperature fluctuation standard deviation corresponding to the historical temperature data are calculated based on the temperature sequence. When the absolute value of the temperature change rate is greater than the preset temperature change rate threshold or the temperature fluctuation standard deviation is greater than the preset fluctuation standard deviation threshold, the difference between the expected temperature and the current temperature is calculated to obtain a temperature error, the temperature error is divided into a preheating section, a reflow section and a cooling section by a segmented adaptive Smith prediction compensation algorithm, compensation parameters are set for each section, the lag time of each section is calculated by a particle swarm algorithm, and the lag time is fed forward to obtain a temperature compensation value; The temperature compensation value is input into a preset fuzzy control rule, the domain division value corresponding to the optimized temperature zone control parameter group is calculated, and the temperature field topological feature map is constructed by combining the thermal stress corresponding to the circuit board material and the pre-set deformation control constraint, the weight distribution of different regions of the temperature field is calculated based on the temperature field topological feature map, the heating power adjustment amount is calculated combined with the temperature compensation value, and the temperature compensation control parameter is solved based on the heating power adjustment amount and the expected temperature.

[0052] Based on the optimization of the temperature zone control parameter group setting sliding window and step size, the sliding window size is determined according to the welding process characteristics. For the standard reflow soldering process, the window size is set to 120 seconds, which can cover a complete temperature change cycle in the welding process. The step size is set to 10 seconds to ensure the continuity of the data coverage while reducing the calculation amount. Taking a certain eight-temperature zone welding equipment as an example, the optimization control parameters of each temperature zone have been determined, such as the parameters of Z3 region P = 2.42, I = 0.58, D = 0.26. The historical temperature data collected and stored during the welding process are sampled, and the sampling frequency is 2 Hz, that is, the temperature data is collected every 0.5 seconds. For temperature zone Z4, a sliding window contains 240 sampling points, forming a temperature sequence {T1, T2,..., T240}. For example, part of the temperature sequence of Z4 temperature zone in a certain welding process is {210.5, 211.2, 211.8, 212.5, 213.0, 213.6, 214.2, 214.7, 215.2, 215.8}℃. Based on the temperature sequence, the temperature change rate is calculated, that is, the temperature difference between adjacent time points divided by the time interval. For data with a sampling interval of 0.5 seconds, the change rate is calculated as the temperature difference between every two adjacent points multiplied by 2, with the unit of ℃ / s. The change rate sequence corresponding to the temperature sequence is {1.4, 1.2, 1.4, 1.0, 1.2, 1.2, 1.0, 1.0, 1.2}℃ / s. The temperature fluctuation standard deviation is obtained by calculating the average of the square sum of the deviation of the temperature sequence from its mean value, and then taking the square root. For the foregoing temperature sequence, the mean value is 213.25℃, and the standard deviation is 1.76℃.

[0053] When the absolute value of the temperature change rate is greater than a preset temperature change rate threshold or the temperature fluctuation standard deviation is greater than a preset fluctuation standard deviation threshold, a temperature compensation control mechanism is triggered, the preset temperature change rate threshold is determined according to process requirements, and for precision electronic component welding, the preset temperature change rate threshold is usually set to 2 ℃ / s; the fluctuation standard deviation threshold is set to 1% of the process temperature, for example, when the target temperature is 240 ℃, the fluctuation standard deviation threshold is 2.4 ℃. When it is detected that the temperature change rate of the temperature zone Z4 reaches 2.3 ℃ / s, which exceeds the preset threshold 2 ℃ / s, the difference between the expected temperature and the current temperature is calculated to obtain a temperature error. Assuming that the expected temperature at this moment is 225 ℃ and the current actual temperature is 219 ℃, the temperature error is 6 ℃. The temperature error is processed by a segmented adaptive Smith prediction compensation algorithm, the welding temperature curve is divided into a preheating section, a reflow section and a cooling section, and different compensation parameters are set according to the temperature characteristics of different sections. The preheating section (the temperature range is usually room temperature to 180 ℃) is characterized by rapid temperature rise, the compensation parameter a is set to 0.7; the reflow section (the temperature range is usually 180 ℃ to 250 ℃) is characterized by accurate temperature control, the compensation parameter β is set to 1.2; and the cooling section (the temperature decreases from the peak value) is characterized by controlled temperature decrease, and the compensation parameter γ is set to 0.5. According to the current temperature 219 ℃, it is judged that the reflow section is applicable, and the compensation parameter β = 1.2 is applicable.

[0054] The hysteresis time of each section is calculated in combination with the particle swarm algorithm, and the hysteresis time represents the delay between the issuance of a control signal and the actual temperature response. The parameter settings of the particle swarm algorithm are as follows: population size 30, maximum iteration number 50, inertia weight 0.8, individual learning factor 1.5, and group learning factor 2.0. The objective function of the algorithm is to minimize the mean square error between the predicted temperature and the actual temperature. For the reflow section of the temperature zone Z4, the hysteresis time calculated by the particle swarm algorithm is 4.5 s. The hysteresis time is fed forwardly compensated to calculate a temperature compensation value. The compensation calculation formula is: temperature compensation value = temperature error × compensation parameter × hysteresis time compensation coefficient. The hysteresis time compensation coefficient is inversely proportional to the hysteresis time, and the calculation method is 1 ÷ (1 + 0.2 × hysteresis time). For the temperature zone Z4, the hysteresis time compensation coefficient is 0.53, and the final temperature compensation value is 6 × 1.2 × 0.53 = 3.82 ℃.

[0055] The temperature compensation value is input into the preset fuzzy control rule to further optimize the control decision. The fuzzy control rule is generated by expert experience and historical data analysis, and adopts the "if-then" form. For example: "if the temperature compensation value is large and the temperature change rate is positive, then the heating power increases moderately". The fuzzy control uses five fuzzy sets: negative large, negative small, zero, positive small, and positive large. The domain partition value corresponding to the optimized temperature zone control parameter group is calculated, and the domain partition determines the boundary and overlap degree of the fuzzy set. The domain partition adopts a non-uniform partition method, which is finer in the accurate control area and coarser in the transition area. For the temperature zone Z4, the domain partition of the temperature compensation value is {-10, -5, -2, -1, 0, 1, 2, 5, 10}℃, and the domain partition of the heating power adjustment amount is {-20, -10, -5, -2, 0, 2, 5, 10, 20}%. The temperature field topology feature map is constructed in combination with the corresponding thermal stress of the circuit board material and the pre-set deformation control constraint. The thermal stress data is obtained through finite element analysis. For a standard PCB board of FR-4 material, when the temperature is 240℃, the thermal stress of the edge area is 15MPa, and the thermal stress of the center area is 12MPa. The maximum allowed deformation is set to 0.2mm. The temperature field topology feature map adopts a heat map form, and the color depth represents the temperature distribution, and the contour line represents the thermal stress distribution.

[0056] The weight distribution of different regions of the temperature field is calculated based on the temperature field topology feature map. The weight distribution principle is that the weight of the region with high thermal stress is high, the weight of the region with large temperature gradient is high, and the weight of the region with high component density is high. The weighted average method is used to calculate the weight of each region. For example, the temperature zone Z4 is divided into 9 sub-regions, the center sub-region weight is 0.18, the edge sub-region weight is 0.08, and the weight of the remaining sub-regions is between 0.10 and 0.15. The heating power adjustment amount is calculated in combination with the temperature compensation value. The adjustment amount calculation considers the heat capacity characteristics of the temperature zone and the current temperature state. The calculation formula is: heating power adjustment amount = temperature compensation value x power coefficient x region weight comprehensive value. The power coefficient is proportional to the heat capacity of the temperature zone and inversely proportional to the current temperature. For the temperature zone Z4, the heat capacity characteristic coefficient is 0.85, the current temperature coefficient is 0.95, and the region weight comprehensive value is 0.14. The calculated heating power adjustment amount is 3.82x0.85x0.95x0.14=0.43kW.

[0057] Solve the temperature compensation control parameters based on the heating power adjustment amount and the desired temperature. The temperature compensation control parameters include the compensated PID parameters and the prediction time constant. The adjustment of the PID parameters adopts an incremental method, and the original PID parameters are corrected according to the size and direction of the heating power adjustment amount. The calculation formula is: the compensated P parameter = the original P parameter × (1 + the heating power adjustment ratio × the P correction coefficient), wherein the heating power adjustment ratio is the percentage of the adjustment amount to the full power, and the P correction coefficient is set to 0.2. The compensated I parameter and D parameter are calculated. For the temperature zone Z4, the original PID parameters are P = 2.35, I = 0.52, and D = 0.24, the heating power adjustment ratio is 0.43 ÷ 2.0 = 0.215, and the compensated PID parameters are calculated as P = 2.45, I = 0.55, and D = 0.25. The prediction time constant is determined according to the dynamic characteristics of the system, and the calculation formula is: the prediction time constant = the lag time × (1 - the heating power adjustment ratio × 0.3). For the temperature zone Z4, the calculated prediction time constant is 4.18 seconds.

[0058] In this embodiment, by sampling the historical temperature data in a sliding window based on the optimized temperature zone control parameter group and calculating the temperature change rate and the temperature fluctuation standard deviation, the temperature change abnormality in the welding process can be identified in real time, the sensitivity and accuracy of temperature abnormality detection are improved, by using the segmented adaptive Smith prediction compensation algorithm to process the temperature error in segments and combining the particle swarm algorithm to optimize and compensate the lag time of each segment, the influence of system lag on the temperature control accuracy can be effectively reduced, the real-time performance and stability of the compensation control are improved, by inputting the temperature compensation value into the fuzzy control rule and combining the thermal stress and deformation control constraints to construct the temperature field topological feature map, the spatial modeling of the temperature distribution and the optimization regulation under the constraint conditions can be realized, and the adaptability of the control process to complex working conditions is improved.

[0059] Figure 2 A flow chart of the temperature compensation parameter generation process of the deep learning-based multi-temperature zone welding temperature intelligent control method of the embodiment of the application.

[0060] In an optional implementation, Generating the temperature control instruction based on the optimized temperature zone control parameter group and the temperature compensation control parameter and performing includes: Obtaining the control period and the temperature sampling interval from the optimized temperature zone control parameter group, periodically sampling the temperature of each temperature zone of the circuit board to obtain the temperature sampling value and the circuit board state parameter; Comparing the temperature sampling value with the temperature compensation control parameter to obtain the sampling temperature deviation value, correcting the sampling temperature deviation value according to the optimized temperature zone control parameter group to obtain the temperature control adjustment amount; generate temperature control instructions containing heating power and heating time based on the temperature control adjustment amount and the circuit board state parameter, and execute the temperature control instructions to perform temperature regulation on the temperature zones.

[0061] obtain a control period and a temperature sampling interval from an optimized temperature zone control parameter set, the optimized temperature zone control parameter set containing various control parameters, wherein the control period determines the update frequency of the control instructions, and the temperature sampling interval determines the frequency of data acquisition. For precision electronic component welding processes, the control period is usually set to 200 milliseconds, and the temperature sampling interval is set to 50 milliseconds, i.e., 4 temperature sampling points are contained in each control period. Taking a certain eight-temperature-zone welding device as an example, the parameters of the Z5 temperature zone in the optimized temperature zone control parameter set include: P = 2.15, I = 0.48, D = 0.22, control period = 200 milliseconds, and sampling interval = 50 milliseconds. Periodic temperature sampling is performed on each temperature zone of the circuit board, and the sampling process is realized by using a network of high-precision thermocouple temperature sensors distributed in each temperature zone. Multiple sensors are configured for each temperature zone to ensure spatial coverage of temperature sampling. For example, the Z5 temperature zone is configured with 5 temperature sensors, which are respectively located at the center point and the four corner positions of the region. After being processed by a signal conditioning circuit, the temperature values collected by the five sensors of the Z5 temperature zone in a certain control period are transmitted to the control unit, which are 241.5°C, 240.8°C, 242.3°C, 241.2°C and 241.0°C, respectively. The temperature sampling value of the temperature zone is calculated by weighted average to be 241.3°C, wherein the center point sensor has a weight of 0.4, and the four corner sensors each have a weight of 0.15.

[0062] obtain a control period and a temperature sampling interval from an optimized temperature zone control parameter set, the optimized temperature zone control parameter set containing various control parameters, wherein the control period determines the update frequency of the control instructions, and the temperature sampling interval determines the frequency of data acquisition. For precision electronic component welding processes, the control period is usually set to 200 milliseconds, and the temperature sampling interval is set to 50 milliseconds, i.e., 4 temperature sampling points are contained in each control period. Taking a certain eight-temperature-zone welding device as an example, the parameters of the Z5 temperature zone in the optimized temperature zone control parameter set include: P = 2.15, I = 0.48, D = 0.22, control period = 200 milliseconds, and sampling interval = 50 milliseconds. Periodic temperature sampling is performed on each temperature zone of the circuit board, and the sampling process is realized by using a network of high-precision thermocouple temperature sensors distributed in each temperature zone. Multiple sensors are configured for each temperature zone to ensure spatial coverage of temperature sampling. For example, the Z5 temperature zone is configured with 5 temperature sensors, which are respectively located at the center point and the four corner positions of the region. After being processed by a signal conditioning circuit, the temperature values collected by the five sensors of the Z5 temperature zone in a certain control period are transmitted to the control unit, which are 241.5°C, 240.8°C, 242.3°C, 241.2°C and 241.0°C, respectively. The temperature sampling value of the temperature zone is calculated by weighted average to be 241.3°C, wherein the center point sensor has a weight of 0.4, and the four corner sensors each have a weight of 0.15.

[0062] obtain a control period and a temperature sampling interval from an optimized temperature zone control parameter set, the optimized temperature zone control parameter set containing various control parameters, wherein the control period determines the update frequency of the control instructions, and the temperature sampling interval determines the frequency of data acquisition. For precision electronic component welding processes, the control period is usually set to 200 milliseconds, and the temperature sampling interval is set to 50 milliseconds, i.e., 4 temperature sampling points are contained in each control period. Taking a certain eight-temperature-zone welding device as an example, the parameters of the Z5 temperature zone in the optimized temperature zone control parameter set include: P = 2.15, I = 0.48, D = 0.22, control period = 200 milliseconds, and sampling interval = 50 milliseconds. Periodic temperature sampling is performed on each temperature zone of the circuit board, and the sampling process is realized by using a network of high-precision thermocouple temperature sensors distributed in each temperature zone. Multiple sensors are configured for each temperature zone to ensure spatial coverage of temperature sampling. For example, the Z5 temperature zone is configured with 5 temperature sensors, which are respectively located at the center point and the four corner positions of the region. After being processed by a signal conditioning circuit, the temperature values collected by the five sensors of the Z5 temperature zone in a certain control period are transmitted to the control unit, which are 241.5°C, 240.8°C, 242.3°C, 241.2°C and 241.0°C, respectively. The temperature sampling value of the temperature zone is calculated by weighted average to be 241.3°C, wherein the center point sensor has a weight of 0.4, and the four corner sensors each have a weight of 0.15.

[0063] The temperature sampling value is compared with the temperature compensation control parameters, which include the target temperature value and the allowed deviation range. For temperature zone Z5, the target temperature is set to 240°C, and the allowed deviation range is ±2°C. The comparison process obtains the sampling temperature deviation value by calculating the difference between the temperature sampling value and the target temperature. Taking the aforementioned sampling data as an example, the sampling temperature deviation value of temperature zone Z5 is 241.3°C-240°C=1.3°C, indicating that the current temperature is slightly higher than the target temperature but still within the allowed range. Correcting the sampling temperature deviation value according to the optimized temperature zone control parameter set is an important link to improve control accuracy. The correction process is implemented based on the incremental PID control algorithm, which calculates the control increment according to the current deviation value, the historical deviation value change rate, and the deviation integral value. In specific calculation, the P, I, and D parameters in the optimized temperature zone control parameter set are multiplied by the deviation value, the deviation change rate, and the deviation integral value, respectively, and the sum is obtained to obtain the temperature control adjustment amount.

[0064] Taking temperature zone Z5 as an example, the current sampling temperature deviation value is 1.3°C, the last period deviation value is 1.5°C, the deviation change rate is -0.2°C / period, and the deviation integral value is 3.8°C·period. The optimized PID parameters P=2.15, I=0.48, and D=0.22 are applied for calculation: the proportional term is 2.15×1.3=2.795, the integral term is 0.48×3.8=1.824, and the differential term is 0.22×(-0.2)=-0.044. The sum of the three terms obtains the temperature control adjustment amount of 4.575°C. Considering the stability requirement, the maximum adjustment amount per period is set to 5°C, so the final temperature control adjustment amount is 4.575°C. The sign of the temperature control adjustment amount indicates the adjustment direction, and a positive value indicates the need to reduce the temperature, and a negative value indicates the need to increase the temperature.

[0065] Based on the temperature control adjustment amount and the circuit board state parameters, the temperature control instruction is generated, which includes two key parameters: heating power and heating time. The calculation of heating power is based on the temperature control adjustment amount and the circuit board thermal characteristic parameters. When calculating, the temperature control adjustment amount is converted into heat adjustment demand, combined with the circuit board material heat capacity, area, and heat transfer efficiency, to calculate the required power adjustment amount. The calculation of heating power also needs to consider the influence of the circuit board state parameters. For example, when the conveyor belt speed is fast, the heating power needs to be increased to compensate for heat loss; when the component distribution density is high, the problem of uneven heat distribution needs to be considered, and the heating power distribution needs to be adjusted appropriately.

[0066] In this embodiment, by setting the control period and sampling interval based on the optimized temperature zone control parameter group, periodic temperature sampling is performed, which can ensure continuous monitoring of the temperature changes of each temperature zone of the circuit board, improve the real-time and completeness of temperature collection, compare the sampling temperature value with the temperature compensation control parameter and correct the deviation, which can effectively reduce the influence of measurement error and environmental disturbance in the sampling process on the control accuracy, improve the accuracy of temperature regulation, generate heating power and heating time instructions based on the corrected temperature control adjustment amount and circuit board state parameters, which can realize fine regulation and control of the temperature zone, improve the stability and consistency of temperature control, ensure uniform heating of the circuit board during welding, improve the quality of the welding points and enhance the reliability of the production process.

[0067] In a second aspect of the embodiment of the present application, a multi-temperature zone welding temperature intelligent control system based on deep learning is provided, which comprises: A first unit is configured to acquire real-time temperature data of each temperature zone of a multi-temperature zone welding device in a circuit board welding production line, perform feature extraction and time series analysis, and obtain a temperature change prediction result; A second unit is configured to divide each temperature zone into multiple control levels based on the temperature change prediction result, calculate a temperature deviation value of each control level and generate an initial weight coefficient, and fit temperature gradient change data by using a least square method to iteratively calculate a temperature compensation coefficient and an initial control parameter of each temperature zone; A third unit is configured to determine a quantitative index according to the initial control parameter, calculate an influence factor of the quantitative index and a comprehensive evaluation score of the initial control parameter by using an adaptive weight algorithm, and iteratively optimize the initial control parameter based on the comprehensive evaluation score to obtain an optimized temperature zone control parameter group; A fourth unit is configured to perform sliding window analysis on historical temperature data based on the optimized temperature zone control parameter group, calculate a temperature change rate and a temperature fluctuation standard deviation, calculate a temperature compensation value based on a proportional integral control algorithm when the temperature change rate exceeds a preset threshold or the temperature fluctuation standard deviation is greater than an allowed range, determine a heating power adjustment amount based on a fuzzy control rule, and generate a temperature compensation control parameter; A fifth unit is configured to generate a temperature control instruction based on the optimized temperature zone control parameter group and the temperature compensation control parameter and execute the temperature control instruction.

[0068] In a third aspect of the embodiment of the present application, an electronic device is provided, which comprises: A processor and a memory for storing processor-executable instructions, wherein the processor is configured to invoke the instructions stored in the memory to execute the method described above.

[0069] In a fourth aspect, the present application provides a computer readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method described above.

[0070] The present application can be a method, an apparatus, a system, and / or a computer program product. The computer program product can include a computer readable storage medium (or media) having computer readable program instructions thereon for performing various aspects of the present application.

[0071] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limit the present application; although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the technical solutions recorded in the above-mentioned embodiments can be modified, or some or all of the technical features thereof can be replaced equivalently; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A deep learning-based intelligent temperature control method for multi-temperature zone welding, characterized in that, include: Real-time temperature data of each temperature zone in the multi-temperature zone welding equipment of the circuit board welding production line are obtained, and feature extraction and time series analysis are performed to obtain temperature change prediction results. Based on the temperature change prediction results, each temperature zone is divided into multiple control levels. The temperature deviation value of each control level is calculated and an initial weighting coefficient is generated. The temperature gradient change data is fitted by the least squares method, and the temperature compensation coefficient and the initial control parameters of each temperature zone are obtained by iterative calculation. The quantitative index is determined based on the initial control parameters. The influence factor of the quantitative index and the comprehensive evaluation score of the initial control parameters are calculated by combining the adaptive weight algorithm. The initial control parameters are iteratively optimized based on the comprehensive evaluation score to obtain the optimized temperature zone control parameter set. Based on the optimized temperature zone control parameter group, a sliding window analysis is performed on historical temperature data to calculate the temperature change rate and temperature fluctuation standard deviation. When the temperature change rate exceeds the preset threshold or the temperature fluctuation standard deviation is greater than the allowable range, the temperature compensation value is calculated based on the proportional-integral control algorithm and combined with fuzzy control rules to determine the heating power adjustment amount, thereby generating temperature compensation control parameters. Temperature control commands are generated and executed based on the optimized temperature zone control parameter group and temperature compensation control parameters.

2. The method according to claim 1, characterized in that, Real-time temperature data of each temperature zone in a multi-temperature zone welding equipment in a circuit board welding production line is acquired, and feature extraction and time series analysis are performed to obtain temperature change prediction results, including: Real-time temperature data of each temperature zone in the multi-temperature zone welding equipment in the circuit board welding production line are obtained, and feature extraction is performed on the real-time temperature data to obtain the temperature change pattern. A temperature change trend matrix is ​​established based on the temperature change pattern, and a time series analysis is performed on the temperature change trend matrix. Based on the results of the time series analysis, the predicted temperature changes for each temperature zone are calculated, and the predicted temperature changes are obtained.

3. The method according to claim 1, characterized in that, Based on the temperature change prediction results, each temperature zone is divided into multiple control levels. The temperature deviation value of each control level is calculated and an initial weighting coefficient is generated. The temperature gradient change data is then fitted using the least squares method, and the temperature compensation coefficient and the initial control parameters for each temperature zone are obtained through iterative calculation. Based on the temperature change prediction results, the temperature coupling degree between different temperature zones of the circuit board and the difference in heat capacity of components are calculated, and each temperature zone is divided into multiple control levels. For each control level, the difference between the preset target temperature value and the preset real-time temperature value is calculated to obtain the temperature deviation value. The initial weighting coefficient is calculated based on the temperature deviation value. Based on the pre-acquired temperature change trend matrix, the temperature gradient change data is calculated. The temperature deviation value and the temperature gradient change data of each temperature zone are fitted by the least squares method to obtain the fitting result. Combined with the pre-acquired temperature compensation coefficient of the previous iteration, the temperature compensation coefficient corresponding to the current iteration is calculated. Based on the initial weighting coefficient, the temperature deviation value, and the temperature compensation coefficient, the initial control parameters are calculated using a pre-set temperature control performance evaluation function combined with the gradient descent method.

4. The method according to claim 3, characterized in that, Based on the initial weighting coefficient, the temperature deviation value, and the temperature compensation coefficient, the initial control parameters are calculated using a pre-set temperature control performance evaluation function and the gradient descent method, including: The initial weighting coefficients, temperature deviation values, and temperature compensation coefficients are combined into an input vector and input into a pre-set adaptive fuzzy neural network. The membership parameters are obtained by fuzzification using a Gaussian membership function. Wavelet transform is performed on the membership parameters to obtain multi-scale feature coefficients. A hierarchical analysis decision tree is constructed based on the multi-scale feature coefficients and local extreme points at different scales are recursively calculated. The membership subsets of the local extreme points are determined. The cross-condition information of the membership subsets is calculated and dimensionality reduction and reconstruction are performed. The optimized membership parameters are obtained by fusing the reconstructed membership subsets. A fuzzy rule base is established based on the optimized membership parameters and the preset temperature control performance evaluation function. The frequent itemsets in the fuzzy rule base are mined by the Apriori algorithm and the support between different fuzzy rules is calculated. An association matrix is ​​constructed based on the support and singular value decomposition is performed to construct a rule activation sequence. The fuzzy rule base is sealed based on the rule activation sequence. The optimized membership parameters are processed by local linear functions to obtain the rule output values ​​and the optimized intermediate parameters are obtained by solving. When the temperature deviation exceeds the preset deviation threshold, the compensation control parameters are calculated based on the optimized intermediate parameters and the exponential moving average, and then output as the initial control parameters.

5. The method according to claim 1, characterized in that, Based on the initial control parameters, a quantitative index is determined. The influence factor of the quantitative index and the comprehensive evaluation score of the initial control parameters are calculated using an adaptive weighting algorithm. Based on the comprehensive evaluation score, the initial control parameters are iteratively optimized to obtain an optimized temperature zone control parameter set, including: The control performance index corresponding to the initial control parameters is obtained, and the quantitative index is obtained by combining the pre-set mapping rules. The ambient temperature fluctuation value and the process parameter fluctuation value are obtained, and the state space vector is constructed by combining the quantitative index. Calculate the weight of each quantitative index in the state space vector and determine the information entropy. Based on the information entropy, calculate the influence factor corresponding to each quantitative index. Collect temperature change sequence within a preset time window and extract time-series and nonlinear features to calculate the parameter sensitivity index. Based on the adaptive firefly algorithm, the initial control parameters are encoded as the location information of individual fireflies. A fluorescence intensity evaluation function is constructed based on the control performance index, and the attraction coefficient is initialized based on the ambient temperature fluctuation value. The optimal individual is determined, and the location information of the optimal individual is decoded into parameter estimates. The Lyapunov function value corresponding to the parameter estimates is calculated, convergence is judged, and system stability is determined. Based on the parameter sensitivity index and the system stability, the quantitative index is fuzzy evaluated to obtain a comprehensive evaluation score. The gradient of the comprehensive evaluation score with respect to the initial control parameters is calculated, and the initial control parameters are updated according to the gradient direction. This update is repeated until the comprehensive evaluation score converges, resulting in an optimized temperature zone control parameter set.

6. The method according to claim 1, characterized in that, Based on the optimized temperature zone control parameter set, a sliding window analysis is performed on historical temperature data to calculate the temperature change rate and temperature fluctuation standard deviation. When the temperature change rate exceeds a preset threshold or the temperature fluctuation standard deviation is greater than the allowable range, a temperature compensation value is calculated based on the proportional-integral control algorithm, and the heating power adjustment is determined in conjunction with fuzzy control rules. The temperature compensation control parameters include: Based on the optimized temperature zone control parameter group, a sliding window and step size are set, and the historical temperature data collected and stored during the welding process are sampled to obtain a temperature sequence. Based on the temperature sequence, the temperature change rate and temperature fluctuation standard deviation corresponding to the historical temperature data are calculated. When the absolute value of the temperature change rate is greater than the preset temperature change rate threshold or the temperature fluctuation standard deviation is greater than the preset fluctuation standard deviation threshold, the temperature error is obtained by calculating the difference between the preset expected temperature and the current temperature. The temperature error is divided into a preheating section, a recirculation section and a cooling section by a segmented adaptive Smith prediction compensation algorithm. Compensation parameters are set for each section. The lag time of each section is calculated by combining the particle swarm algorithm and the lag time is fed forward to obtain the temperature compensation value. The temperature compensation value is input into a preset fuzzy control rule, the universe partition value corresponding to the optimized temperature zone control parameter group is calculated, and the temperature field topology feature map is constructed by combining the thermal stress corresponding to the circuit board material and the preset deformation control constraints. The weight distribution of different regions of the temperature field is calculated based on the temperature field topology feature map, and the heating power adjustment amount is calculated by combining the temperature compensation value. The temperature compensation control parameters are obtained by solving based on the heating power adjustment amount and the desired temperature.

7. The method according to claim 1, characterized in that, Based on the optimized temperature zone control parameter set and temperature compensation control parameters, temperature control commands are generated and executed, including: The control cycle and temperature sampling interval are obtained from the optimized temperature zone control parameter group. Periodic temperature sampling is performed on each temperature zone of the circuit board to obtain temperature sampling values ​​and circuit board status parameters. The sampled temperature value is compared with the temperature compensation control parameters to obtain the sampled temperature deviation value. The sampled temperature deviation value is then corrected according to the optimized temperature zone control parameter group to obtain the temperature control adjustment amount. Based on the temperature control adjustment amount and the circuit board status parameters, a temperature control command containing heating power and heating time is generated, and the temperature control command is executed to adjust the temperature of the temperature zone.

8. A deep learning-based intelligent control system for multi-temperature zone welding, used to implement the method of any one of claims 1-7, characterized in that, include: The first unit is used to acquire real-time temperature data of each temperature zone of the multi-temperature zone welding equipment in the circuit board welding production line, and to perform feature extraction and time series analysis to obtain temperature change prediction results. The second unit is used to divide each temperature zone into multiple control levels based on the temperature change prediction results, calculate the temperature deviation value of each control level and generate initial weight coefficients, and fit the temperature gradient change data by the least squares method to iteratively calculate the temperature compensation coefficient and the initial control parameters of each temperature zone. The third unit is used to determine the quantitative index based on the initial control parameters, calculate the influence factor of the quantitative index and the comprehensive evaluation score of the initial control parameters by combining the adaptive weight algorithm, and iteratively optimize the initial control parameters based on the comprehensive evaluation score to obtain the optimized temperature zone control parameter set. The fourth unit is used to perform sliding window analysis on historical temperature data based on the optimized temperature zone control parameter group, calculate the temperature change rate and temperature fluctuation standard deviation. When the temperature change rate exceeds the preset threshold or the temperature fluctuation standard deviation is greater than the allowable range, the temperature compensation value is calculated based on the proportional-integral control algorithm and the heating power adjustment amount is determined in combination with the fuzzy control rules to generate temperature compensation control parameters. The fifth unit is used to generate and execute temperature control commands based on the optimized temperature zone control parameter set and temperature compensation control parameters.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images