A circuit board welding automation control method and system

Abstract

The present application relates to the technical field of electronic manufacturing, and discloses a kind of circuit board welding automation control method and system.The method includes obtaining the preheating temperature sequence of welding equipment in preheating stage, calculates the heating rate to determine the thermal load characteristic parameter;According to the thermal load characteristic parameter, the reference power curve is matched to drive equipment operation, and real-time temperature data stream is collected;The second derivative change of the real-time temperature data stream is monitored, and the phase change cut-in moment of solder heat absorption is identified;The deviation of real-time heating slope and ideal phase change slope in phase change stage is calculated, and dynamic power adjustment instruction is generated to compensate for latent heat lag;The recovery trend of real-time heating slope is detected to lock the phase change exit moment;According to the thermal load characteristic parameter and the temperature of phase change exit moment, the total heat demand is calculated, and wetting heating control signal is generated to drive equipment to preset peak value.The present application can realize the self-adaptive precision temperature control of different heat capacity pads.

Description

Technical Field

[0001] This invention relates to the field of electronic manufacturing technology, and in particular to an automated control method and system for circuit board welding. Background Technology

[0002] Currently, in the assembly process of electronic circuit board components, in order to ensure that the solder alloy can fully wet the pads without damaging heat-sensitive components, strict energy management must be carried out in each stage of the heating process, such as preheating, reflow melting, and cooling solidification.

[0003] In existing technologies, various automated control methods have been widely adopted in the industrial welding field. For example, in the metal joining process of large structural components, automatic and semi-automatic arc welding technology is quite mature. It typically achieves closed-loop control of welding quality by real-time monitoring of arc voltage and current fluctuations or by using visual sensors to track the molten pool morphology. However, unlike automatic and semi-automatic arc welding, which focuses on the control of a single macroscopic molten pool, circuit board reflow soldering faces a complex scenario where hundreds or thousands of tiny solder joints undergo simultaneous thermophysical changes. Most existing circuit board welding controls rely on pre-set fixed temperature-time curves or simple feedback regulation based on PID algorithms, that is, adjusting the heating power according to the deviation between the value collected by the temperature sensor and the preset target value.

[0004] Therefore, existing technologies suffer from problems such as slow perception of solder phase transformation processes and poor adaptability to differences in heat load. Summary of the Invention

[0005] This invention provides an automated control method and system for circuit board soldering to solve the problems of slow perception of solder phase transformation process and poor adaptability to thermal load differences in the prior art.

[0006] In a first aspect, to solve the above-mentioned technical problems, the present invention provides an automated control method for circuit board soldering, comprising: Obtain the preheating temperature sequence of the welding equipment when it is running at a preset constant test power, calculate the heating rate of the preheating temperature sequence, and determine the ratio of the preset constant test power to the heating rate as a heat load characteristic parameter. According to the heat load characteristic parameters, the corresponding reference power curve is matched from the preset reference power curve library, and the reference power curve is used to drive the welding equipment to run, while the real-time temperature data stream of the welding process is collected. When the value of the real-time temperature data stream enters the preset phase transition temperature range, the second derivative operation is performed on the real-time temperature data stream. When the calculation result shows that the value of the second derivative changes from positive to negative and satisfies the preset zero crossover condition, the phase transition cut-off time is determined. The phase transition cut-off time is set as the starting point, and the real-time temperature rise slope of the real-time temperature data stream is continuously calculated. The real-time temperature rise slope is then compared with the preset ideal phase transition slope to obtain the slope deviation value. The latent heat compensation amount is calculated using the slope deviation value, a dynamic power adjustment command superimposed with the latent heat compensation amount is generated, and the welding equipment is driven to execute the dynamic power adjustment command. The change trend of the real-time heating slope is detected, and when the real-time heating slope exceeds the preset liquid phase slope threshold, the phase change exit time is determined. Based on the heat load characteristic parameters and the current temperature value corresponding to the phase transition exit time, the total heat demand required to reach the preset welding target state is calculated, the total heat demand is converted into a wetting heating control signal, and the welding equipment is driven to run until the temperature reaches the preset wetting peak threshold.

[0007] In a second aspect, the present invention provides an automated control system for circuit board welding, comprising: The heat load parameter determination module is used to obtain the preheating temperature sequence of the welding equipment when it is running at a preset constant test power, calculate the heating rate of the preheating temperature sequence, and determine the ratio of the preset constant test power to the heating rate as the heat load characteristic parameter. The reference drive module is used to match the corresponding reference power curve from the preset reference power curve library according to the thermal load characteristic parameters, and use the reference power curve to drive the welding equipment to run, while collecting real-time temperature data streams during the welding process. The phase transition entry determination module is used to perform second derivative calculation on the real-time temperature data stream when the value of the real-time temperature data stream enters the preset phase transition temperature range, and determine the phase transition entry time when the calculation result shows that the value of the second derivative changes from a positive value to a negative value and satisfies the preset zero crossover condition. The slope deviation calculation module is used to set the phase change cut-in time as the starting point, continuously calculate the real-time temperature rise slope of the real-time temperature data stream, and compare the real-time temperature rise slope with the preset ideal phase change slope to obtain the slope deviation value. The dynamic power adjustment module is used to calculate the latent heat compensation amount using the slope deviation value, generate a dynamic power adjustment command superimposed with the latent heat compensation amount, and drive the welding equipment to execute the dynamic power adjustment command. The phase change exit determination module is used to detect the changing trend of the real-time heating slope, and determine the phase change exit time when the real-time heating slope exceeds the preset liquid phase slope threshold. The wetting and heating control module is used to calculate the total heat demand required to reach the preset welding target state based on the heat load characteristic parameters and the current temperature value corresponding to the phase change exit time, convert the total heat demand into a wetting and heating control signal, and drive the welding equipment to run until the temperature reaches the preset wetting peak threshold.

[0008] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention solves the technical problem of poor adaptability of traditional soldering processes to the thermal characteristics of circuit boards of different specifications by introducing a constant power detection and heating rate calculation mechanism in the preheating stage. This invention quantifies the thermal load characteristic parameters characterizing the inherent heat capacity of the solder pads by calculating the ratio of constant test power to actual heating rate, and automatically matches the optimal reference power curve accordingly. This feedforward thermal characteristic identification mechanism significantly improves the adaptability of the soldering process to multi-variety, small-batch production scenarios.

[0009] (2) This invention solves the problem of temperature lag and cold solder joint caused by the heat absorption of the solder phase transition during the welding process by phase transition identification based on the zero crossover of the second derivative and latent heat compensation mechanism based on the slope deviation. This closed-loop control based on physical mechanism ensures that the solder completely completes the transformation from solid to liquid state within the specified time window, effectively eliminating the hidden danger of cold solder joint caused by incomplete melting.

[0010] (3) This invention locks the phase transition exit time by detecting the rise in the liquid phase slope and accurately calculates the wetting energy pulse based on the principle of energy conservation, thus solving the problems of temperature overshoot and low control accuracy in the liquid phase wetting stage. This "tailor-made" energy delivery strategy ensures that the solder pad surface receives sufficient wetting and spreading energy while avoiding overheating and burning caused by excess heat accumulation, thereby ensuring the growth quality of intermetallic compounds (IMC) and the mechanical strength of the solder joint at the microscopic level. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of a circuit board welding automation control method provided in the first embodiment of the present invention; Figure 2 This is a schematic diagram of an automated control system for circuit board welding provided in the second embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] Reference Figure 1 The first embodiment of the present invention provides an automated control method for circuit board soldering, comprising the following steps: S11, Obtain the preheating temperature sequence of the welding equipment when it is running at a preset constant test power, calculate the heating rate of the preheating temperature sequence, and determine the ratio of the preset constant test power to the heating rate as a heat load characteristic parameter. S12, Match the corresponding reference power curve from the preset reference power curve library according to the heat load characteristic parameters, and use the reference power curve to drive the welding equipment to run, while collecting real-time temperature data stream of the welding process. S13, when the value of the real-time temperature data stream enters the preset phase transition temperature range, the second derivative operation is performed on the real-time temperature data stream. When the calculation result shows that the value of the second derivative changes from positive to negative and satisfies the preset zero crossover condition, the phase transition cut-in time is determined. S14, set the phase change cut-in time as the starting point, continuously calculate the real-time temperature rise slope of the real-time temperature data stream, and compare the real-time temperature rise slope with the preset ideal phase change slope to obtain the slope deviation value. S15, calculate the latent heat compensation amount using the slope deviation value, generate a dynamic power adjustment command superimposed with the latent heat compensation amount, and drive the welding equipment to execute the dynamic power adjustment command; S16, detect the changing trend of the real-time heating slope, and determine the phase change exit time when the real-time heating slope exceeds the preset liquid phase slope threshold. S17. Based on the heat load characteristic parameters and the current temperature value corresponding to the phase change exit time, calculate the total heat demand required to reach the preset welding target state, convert the total heat demand into a wetting heating control signal, and drive the welding equipment to run until the temperature reaches the preset wetting peak threshold.

[0014] In step S11, the preheating temperature sequence of the welding equipment under a preset constant test power is obtained, the heating rate of the preheating temperature sequence is calculated, and the ratio of the preset constant test power to the heating rate is determined as a heat load characteristic parameter, including: Extract the data segment from the preheating temperature sequence whose temperature falls within the preset linear temperature rise range; The least squares method is used to perform linear fitting on the data segment, and the slope of the fitted line is calculated as the heating rate. Calculate the ratio of the preset constant test power to the heating rate, and normalize the ratio to obtain the thermal load characteristic parameters.

[0015] It should be noted that a dual-threshold method is used to extract data segments within a preset linear temperature rise range. First, a lower temperature threshold and an upper temperature threshold are set for this range. Then, each discrete sampling point in the preheating temperature sequence is traversed in chronological order. The system locates the starting data point where the value first exceeds or equals the lower temperature threshold, and the ending data point where the value first exceeds or equals the upper temperature threshold. Subsequently, the system extracts all continuous sampling data between the starting and ending data points (including both endpoints) to construct an independent time-temperature subsequence, i.e., the data segment.

[0016] It should be noted that the least squares method is used for linear fitting and slope calculation. A univariate linear regression model is established with time as the independent variable and temperature as the dependent variable. During the calculation, the residual between the temperature value of each actual sampling point in the data segment and the theoretical prediction value of the regression model at the corresponding time is calculated, and the sum of squares of all residuals is obtained. The model parameters that minimize this sum of squares of residuals are solved through mathematical operations. The proportionality coefficient describing the rate of temperature change over time is determined as the slope of the fitted line, and this slope value is defined as the heating rate. This method can statistically filter out the interference of single-point sensor noise on the rate calculation.

[0017] It should be noted that the calculation of the ratio and the maximum-minimum normalization process first involves reading the power value of the constant test power currently being performed by the welding equipment, performing a division operation, using this power value as the dividend, and the previously calculated heating rate as the divisor. The resulting quotient is the original thermal resistance value. Subsequently, a maximum-minimum normalization operation is performed on this original thermal resistance value. The system calculates the difference between this original thermal resistance value and the preset minimum thermal resistance reference value, and simultaneously obtains the difference between the preset maximum thermal resistance reference value and this minimum thermal resistance reference value as the range span. Finally, the aforementioned difference is divided by this range span to obtain a dimensionless value between zero and one, which is the thermal load characteristic parameter.

[0018] Specifically, the selection of the preset constant test power value is based on a comprehensive evaluation of the rated power of the soldering equipment and the thermal safety threshold of the circuit board. This value is determined according to two core constraints: first, a lower limit constraint, meaning the power must be sufficient to drive a heavy-load (high heat capacity) circuit board to generate a significant temperature rise exceeding the sensor noise floor (e.g., greater than 0.5 degrees Celsius per second), ensuring the signal-to-noise ratio in the heating rate calculation; second, an upper limit constraint, meaning the power must be limited to a safe preheating rate for light-load (low heat capacity) circuit boards (e.g., less than 3.0 degrees Celsius per second), to prevent premature excessive flux evaporation or thermal shock damage to the substrate due to excessively rapid heating during the testing phase. Typically, the system selects 20% to 40% of the equipment's rated maximum output power as the preset value of this constant test power (e.g., 500 watts for a 2000-watt rated power device), ensuring both the validity of the test data and the preheating safety of various types of circuit boards.

[0019] It is worth noting that the determination of the preset linear temperature rise range is based on statistical analysis of flux thermogravimetric analysis (TGA) data. The system analyzes the volatility characteristic curves of flux components in commonly used solder pastes, and selects a stable temperature rise range (usually 50 to 100 degrees Celsius) before the solvent begins to evaporate and far from room temperature fluctuations as the preset range to ensure that the linearity of the temperature rise data is not affected by the endothermic reaction.

[0020] It is worth noting that the determination of the preset minimum and maximum thermal impedance reference values ​​is based on extreme value statistical analysis of historical production data. The system collects the original thermal impedance data of circuit boards of different thicknesses and layers during preheating tests in historical production. After removing outliers, the lower edge value (e.g., the 5th percentile) of the statistical distribution is selected as the minimum reference value, and the upper edge value (e.g., the 95th percentile) is selected as the maximum reference value. This setting ensures that the normalized parameters can cover the vast majority of production situations and have good discriminative power.

[0021] For example, a constant test power of 500 watts is set. The system extracts a data segment from the preheating temperature sequence between 50°C and 100°C, containing 50 sampling points over a time span of 20 seconds. Using least squares fitting, the slope of the fitted line (i.e., the heating rate) for this data segment is 2.5°C per second. The system performs a division operation, dividing 500 by 2.5 to obtain an initial thermal impedance value of 200. Assuming the system presets a minimum thermal impedance reference value of 100 and a maximum thermal impedance reference value of 400, the system performs a normalization calculation. First, it calculates 200 minus 100 to obtain 100, then calculates 400 minus 100 to obtain 300, and finally divides 100 by 300 to obtain 0.33. The final thermal load characteristic parameter of the circuit board is determined to be 0.33.

[0022] In step S12, a corresponding reference power curve is matched from a preset reference power curve library according to the heat load characteristic parameters, and the welding equipment is driven to run using the reference power curve. Simultaneously, real-time temperature data streams of the welding process are collected, including: According to the preset reference power curve library, each set of reference power curves corresponds to a heat load range. The larger the heat load characteristic parameter, the higher the preset basic power output of the corresponding reference power curve in the phase change stage. Determine the target heat load range to which the heat load characteristic parameters belong; The system retrieves a curve from the preset reference power curve library that matches the target heat load range as the reference power curve, and controls the energy output unit of the welding equipment according to the time power setting value of the reference power curve to collect the real-time temperature data stream.

[0023] It should be noted that the preset reference power curve library is constructed in local memory using a lookup table. The numerical range of the heat load interval is used as the index key, and the corresponding reference power curve data file pointer is used as the index value. Each reference power curve is essentially a two-dimensional array containing a timestamp sequence and a power percentage sequence. The setting that "the larger the heat load characteristic parameter, the higher the base power output" is based on the principle of thermodynamic compensation. For heavy-duty boards with high heat capacity, more heat energy is required to maintain a constant temperature during the solder melting stage. Therefore, the system significantly increases the power setting value during this stage in the curve design to prevent temperature drops.

[0024] It should be noted that determining the target heat load range and retrieving the curve is done by traversing all heat load ranges in the lookup table and checking whether the heat load characteristic parameters calculated in step S11 are greater than or equal to the lower limit of the range and less than the upper limit of the range. Once this condition is met, the system immediately locks the range as the target heat load range and loads the corresponding power curve file into memory according to the associated pointer, thus completing the curve loading.

[0025] It should be noted that controlling the energy output unit and acquiring the real-time temperature data stream involves starting a high-precision hardware timer to traverse the loaded reference power curve with millisecond-level time granularity. At each time step, the system reads the power setpoint defined in the curve, converts it into the duty cycle of the corresponding pulse width modulation (PWM) signal, and drives the energy output unit, such as the infrared heating lamp or hot air motor, to operate. Simultaneously, the system triggers the analog-to-digital converter (ADC) to read the voltage signal from the temperature sensor, converts it into a digital temperature value, adds a current timestamp, and stores it sequentially in a circular buffer, forming the real-time temperature data stream.

[0026] It is worth noting that, based on the analysis of massive historical process data, specific heat load intervals and baseline power curves are constructed. During the initialization phase, production data corresponding to all historically excellent welded products are collected, and their heat load characteristic parameters and actual power curves are extracted. Using the K-means clustering algorithm, these data samples are divided into k (e.g., three or five) typical clusters. The number of typical clusters (i.e., the k value) is determined using the elbow method; the system iterates through consecutive candidate k values, calculating the sum of squared errors (SSE) within each cluster for different k values. When the decrease in SSE with increasing k value becomes sharp, and the curve exhibits an elbow inflection point, the k value corresponding to this inflection point is determined as the optimal number of typical clusters. The statistical distribution range of all heat load parameters in each cluster is calculated as the heat load interval, and the arithmetic mean trajectory of all power curves in that cluster is calculated as the baseline power curve for that interval. This method ensures that the preset curves can objectively reflect the optimal process rules for different load types of boards.

[0027] For example, assume the heat load characteristic parameter output in step S11 is 0.65. The system has three preset curves: a light load range (0.0 to 0.4), a medium load range (0.4 to 0.7), and a heavy load range (0.7 to 1.0). The system compares and finds that 0.65 falls within the medium load range, so it retrieves the curve numbered "Curve_Medium_Std". This curve is set at 100 to 120 seconds after welding starts (the expected phase transition zone), with a power output of 55% of the rated power. Control commands are generated accordingly. When the time reaches 110 seconds, the system outputs a PWM signal with a 55% duty cycle to the heater. At the same time, the sensor acquires the current real-time temperature of 175.5 degrees Celsius and records it in the data stream.

[0028] In step S13, when the value of the real-time temperature data stream enters the preset phase transition temperature range, a second derivative operation is performed on the real-time temperature data stream. When the calculation result shows that the value of the second derivative changes from positive to negative and satisfies the preset zero-crossing condition, the phase transition cut-off time is determined, including: Gaussian smoothing filtering is applied to the real-time temperature data stream entering the preset phase transition temperature range to generate a smooth temperature sequence. Calculate the first and second derivative sequences of the smoothed temperature sequence with respect to time; The second derivative sequence is monitored in real time. When the value of the second derivative changes from positive to negative after crossing zero, and the number of subsequent consecutive preset stable verification cycles is less than the preset negative judgment threshold, the preset zero crossover condition is determined to be met. The starting time point that satisfies the preset zero-crossing condition is marked as the phase transition cut-in time.

[0029] It should be noted that Gaussian smoothing filtering of the real-time temperature data stream first involves constructing a one-dimensional Gaussian convolution kernel of a preset length, with weights distributed according to a normal distribution. When a new data point arrives in the real-time temperature data stream, the system selects a local data window centered on that new data point, multiplies each temperature value within the window with its corresponding weight value in the convolution kernel, and sums all the product results to obtain a weighted average. This weighted average is the smoothed temperature value at the current moment, and arranging these values ​​in chronological order constitutes the smoothed temperature sequence.

[0030] It is worth noting that the preset length was determined based on the spectral analysis of the temperature sensor's background noise. The sensor's raw output signal was acquired under isothermal steady-state conditions, and a Fast Fourier Transform (FFT) was performed to identify the dominant frequency period of high-frequency thermal noise. To achieve the optimal balance between noise suppression and signal detail preservation, a value was selected whose corresponding time-domain window width must cover at least one complete noise dominant frequency period, while being less than the solder phase transition characteristic time, i.e., one-tenth of the typical temperature plateau duration. This setting principle ensures that the filter can effectively remove random electronic noise while avoiding excessive smoothing due to an overly wide window, thus fully preserving the geometric characteristics of the abrupt bending of the temperature curve at the phase transition inflection point.

[0031] It should be noted that for the first derivative, the system calculates the difference between the smoothed temperature value at the current moment and the smoothed temperature value at the previous sampling moment, and divides this difference by the sampling time interval to obtain the instantaneous heating rate. For the second derivative, the system calculates the difference between the instantaneous heating rate at the current moment and the instantaneous heating rate at the previous sampling moment, and divides this difference again by the sampling time interval to obtain the heating acceleration. The system stores the calculated acceleration values ​​in the second derivative sequence in real time.

[0032] It should be noted that the real-time monitoring and determination of the zero-crossing condition is implemented using a logic combining sign detection and counter verification. The system continuously reads the latest values ​​in the second derivative sequence. First, sign detection is performed to determine whether the second derivative value at the previous time step is greater than or equal to zero, and whether the second derivative value at the current time step is less than zero. If this condition is met, it indicates that a crossover from positive to negative has occurred. Subsequently, the system starts a counter and continuously checks each subsequently sampled second derivative value. If the subsequent value is consistently lower than the preset negative determination threshold, the counter is incremented; if a value exceeds the threshold, the counter is reset to zero and the detection state is reset. Only when the accumulated counter value reaches the preset number of stable verification periods does the system finally confirm that the zero-crossing condition is met.

[0033] It is worth noting that the determination of the preset phase transformation temperature range is based on the solidus temperature characteristics of the solder alloy. Consult the material property table of the solder (e.g., SAC305) to obtain its theoretical solidus temperature (e.g., 217 degrees Celsius). Select a safety margin below this temperature (e.g., minus 30 degrees Celsius) as the lower limit of the range, and select a margin above this temperature (e.g., add 10 degrees Celsius) as the upper limit.

[0034] It is worth noting that the determination of the preset number of stable verification cycles and the negative judgment threshold is based on the spectral analysis of the sensor's thermal noise. The system collects background noise data from the sensor under constant temperature conditions and calculates the standard deviation of the noise amplitude. The negative value of three times this standard deviation is selected as the negative judgment threshold to shield against random fluctuations. Simultaneously, the dominant frequency period of the noise is analyzed, and a number of sampling points greater than the dominant frequency period (e.g., 5 to 10 consecutive sampling points) is selected as the number of stable verification cycles to ensure the robustness of the judgment.

[0035] For example, suppose the system collects a real-time temperature of 190 degrees Celsius, entering a preset phase transition range. After applying Gaussian filtering, the second derivative calculated at 100.5 seconds is +0.02 degrees Celsius per square second. At 100.6 seconds, the second derivative becomes -0.15 degrees Celsius per square second, indicating a sign flip. The system's preset negative threshold is -0.1 degrees Celsius per square second, and the number of stable verification cycles is 5. Further monitoring reveals that from 100.7 seconds to 101.1 seconds, the second derivative values ​​at five consecutive sampling points are -0.18, -0.20, -0.22, -0.21, and -0.19, all below the -0.1 threshold. The system determines that the zero-crossing condition is met and marks the start time of the sign flip (100.6 seconds) as the phase transition initiation time.

[0036] In step S14, the phase transition cut-off time is set as the starting point, and the real-time temperature rise slope of the real-time temperature data stream is continuously calculated. The real-time temperature rise slope is then compared with the preset ideal phase transition slope to obtain the slope deviation value, including: Set the phase transition cutoff time as the time anchor point and start the sliding monitoring window; The slope of the temperature data within the sliding monitoring window is calculated using a linear regression algorithm and used as the real-time temperature rise slope. The target temperature rise rate corresponding to the current reference power curve is read as the ideal phase transition slope. The slope deviation value is obtained by calculating the difference between the real-time heating slope and the ideal phase transition slope.

[0037] It should be noted that the initiation of the sliding monitoring window and the calculation of the real-time temperature rise slope are implemented using dynamic caching and statistical regression techniques for time series data. The system allocates a fixed-capacity First-In-First-Out (FIFO) queue in memory as the sliding monitoring window. From the moment of phase transition, whenever the sensor collects a new real-time temperature data point, the system pushes it into the queue while removing the oldest data point, maintaining a constant queue length. Then, the least squares method is executed, using the timestamps of all data points in the queue as independent variables and the temperature value as the dependent variable to construct a univariate linear regression model. Through iterative calculation, the system finds the model parameters that minimize the sum of squared residuals for all data points; the regression coefficients representing the rate of change are then determined as the real-time temperature rise slope. This method effectively smooths the random fluctuations of single-point data and reflects the true trend of temperature change.

[0038] It should be noted that obtaining the ideal phase transformation slope and calculating the slope deviation is achieved using synchronous table lookup and differential calculation techniques. The system maintains a global clock that runs synchronously with the welding process. Based on the current timestamp index, it directly reads the preset theoretical temperature rise rate value at that moment from the reference power curve data structure loaded in step S12 and defines it as the ideal phase transformation slope. Subsequently, the system performs a subtraction operation, subtracting the value of the ideal phase transformation slope from the previously calculated real-time temperature rise slope value. The resulting algebraic difference is the slope deviation value. The sign and absolute value of this value directly quantify the degree of deviation of the current welding process from the ideal process path.

[0039] It is worth noting that the length of the sliding monitoring window, i.e., the queue capacity, is determined based on the analysis of the temperature sensor's sampling frequency and thermal response time constant. By analyzing the sensor's sampling frequency (e.g., 10 Hz) and the characteristic time of pad heat transfer (e.g., 0.5 seconds), the number of sampling points (e.g., 5 to 10 points) that can cover at least one complete thermal response cycle is selected as the window length. This ensures that the calculated slope filters out high-frequency noise while maintaining a sensitive response to temperature changes, avoiding numerical lag caused by an excessively long window.

[0040] It is worth noting that the determination of the ideal phase transition slope is based on thermodynamic finite element analysis (FEA). During the process development phase, a three-dimensional thermal model of the standard pads and solder is constructed to simulate the phase transition heat absorption process under ideal constant heat flux input. The rate of temperature change during the phase transition plateau period of the simulation curve is extracted, which is usually a small positive value close to zero, representing the dynamic equilibrium state between latent heat absorption and heat input. This value is then solidified as the target parameter in the baseline curve.

[0041] For example, assume the phase transition initiation time is at 120 seconds, and the current system time is at 125 seconds. The system extracts 10 temperature sampling points between 124.5 seconds and 125.0 seconds to form a sliding monitoring window. Linear regression calculations show that the slope (real-time temperature rise slope) corresponding to the temperature rise trend at these 10 points is 0.2 degrees Celsius per second. Simultaneously, the system queries the baseline power curve and learns that at 125 seconds, to maintain the ideal phase transition progress, the target temperature rise rate (ideal phase transition slope) should be 0.8 degrees Celsius per second. The system performs a subtraction operation, subtracting 0.8 from 0.2, resulting in a slope deviation of -0.6 degrees Celsius per second. This negative value indicates that the current actual temperature rise rate significantly lags behind the ideal target, requiring system intervention for compensation.

[0042] In step S15, the latent heat compensation amount is calculated using the slope deviation value, a dynamic power adjustment command superimposed with the latent heat compensation amount is generated, and the welding equipment is driven to execute the dynamic power adjustment command, including: Construct a power compensation model that includes proportional and integral components; The slope deviation value is input into the proportional element of the power compensation model to calculate the instantaneous deviation compensation component. The slope deviation value is integrated and accumulated on the time axis after the phase transition cut-in time to calculate the thermal hysteresis compensation component; The latent heat compensation is obtained by summing the instantaneous deviation compensation component and the thermal hysteresis compensation component, and then superimposed on the base power at the current moment to generate the dynamic power adjustment command.

[0043] It should be noted that a positional PI (proportional-integral) control architecture is specifically used to construct the model. The construction process mainly includes three stages: memory space initialization, parameter loading, and logic mapping. First, a dedicated structure data area is allocated in the microcontroller's random access memory (RAM) to maintain the model's operating state. This data area contains two core state variables: a proportional register for real-time refreshing of the current input value, and an integral accumulator for persistently storing all accumulated deviation values ​​since the phase transition. The integral accumulator is automatically cleared at each phase transition (i.e., when step S3 is triggered) to ensure control consistency. Second, preset proportional gain coefficients and integral gain coefficients are read from non-volatile flash memory and loaded into the model's read-only parameter area. These two coefficients define the system's sensitivity to the current error and the strength of its correction for historical errors, respectively. Finally, the operation flow of the distributed arithmetic logic unit (ALU) is defined. The process stipulates that within each control cycle, the model first reads the current slope deviation value and stores it in the proportional register. This value is then multiplied by the proportional gain coefficient to obtain the instantaneous deviation compensation component. Subsequently, the current slope deviation value is added to the integral accumulator for updating, and the updated accumulator value is multiplied by the integral gain coefficient to obtain the thermal hysteresis compensation component. The system output logic is set to the algebraic sum of these two components. Furthermore, the model construction includes an output limiting circuit to ensure that the calculated final power adjustment command does not exceed the physical output range of the hardware device (i.e., 0% to 100% duty cycle), preventing system runaway due to integral saturation. The system does not introduce a differential (D) circuit to avoid excessive amplification of the sensor's high-frequency thermal noise, thereby ensuring the smoothness of the power output.

[0044] It should be noted that the calculation of the instantaneous deviation compensation component is achieved using a multiplication operation technique. The system reads the slope deviation value output by step S14 at the current moment and reads the preset proportional gain coefficient. The system performs a multiplication operation, multiplying the slope deviation value by the proportional gain coefficient, and the calculated product is directly defined as the instantaneous deviation compensation component. The physical meaning of this component is to quickly "reverse correct" the current heating rate deviation; if the heating is too slow (the deviation is negative), it generates positive power thrust.

[0045] It should be noted that the calculation of the thermal hysteresis compensation component is achieved using a combination of summation and multiplication. The system establishes an accumulator with the phase transition initiation time determined in step S13 as the time origin. Whenever a new slope deviation value is generated, the system adds it to the accumulator, updating the total accumulated value. Subsequently, the system reads a preset integral gain coefficient, multiplies the current total accumulated value by this coefficient, and the resulting product is defined as the thermal hysteresis compensation component. This component is used to eliminate the persistent steady-state error caused by latent heat absorption, ensuring that the temperature eventually rises to the target trajectory.

[0046] It should be noted that the generation and execution of dynamic power adjustment commands are implemented using superposition modulation and hardware driving technology. The system first arithmetically adds the instantaneous deviation compensation component and the thermal hysteresis compensation component to obtain the total latent heat compensation. Next, based on the current timestamp, the system queries the corresponding base power setpoint from the reference power curve selected in step S12. The system performs an addition operation, superimposing the latent heat compensation onto the base power setpoint to obtain the corrected target power value. Finally, the system converts this target power value into a corresponding current control signal or duty cycle signal, and sends it to a solid-state relay or thyristor controller via the I / O interface to drive the heating equipment to change its output energy.

[0047] It is worth noting that the determination of the proportional gain coefficient and integral gain coefficient is based on the Ziegler-Nichols critical proportional gain method combined with virtual simulation. In the digital twin environment, the control system is set to pure proportional control, and the gain is gradually increased until the temperature curve shows constant amplitude oscillation. The critical gain and oscillation period at this point are recorded. Subsequently, the initial parameters are calculated according to the empirical formula of the PI controller (for example, the proportional gain is 0.45 times the critical gain, and the integral time is 0.83 times the oscillation period), and fine-tuned according to the overshoot in the actual welding test, and finally fixed to the preset values. Since the deviation defined in step S14 is "actual minus ideal" (negative if there is lag), both gain coefficients are set to negative numbers to achieve negative feedback control.

[0048] For example, assume the current slope deviation is -0.6 degrees Celsius per second (indicating a temperature lag). The preset proportional gain coefficient is -20.0, and the integral gain coefficient is -5.0. The system performs a proportional calculation, multiplying -0.6 by -20.0 to obtain an instantaneous deviation compensation component of +12.0. Simultaneously, the total historical deviation stored in the accumulator since the phase transition is -1.0. The system performs an integral calculation, multiplying -1.0 by -5.0 to obtain a thermal lag compensation component of +5.0. The system sums up to obtain a total latent heat compensation of +17.0 (i.e., 12.0 plus 5.0). Looking up a table, the current reference power is 45% of the rated power. The system performs a superposition operation, adding 45 and 17 to obtain a final target power of 62%. The system then generates an instruction to increase the heater's output power from 45% to 62% to accelerate solder melting.

[0049] In step S16, the changing trend of the real-time heating slope is detected. When the real-time heating slope exceeds a preset liquid phase slope threshold, the phase transition exit time is determined, including: Temperature points are collected according to a preset number of sampling points, and the average slope of the temperature points is calculated in real time as the real-time temperature rise slope. Determine whether the real-time heating slope is greater than the liquid phase slope threshold within a preset number of sampling periods; If the judgment result is yes, then the first sampling time that meets the condition is determined as the phase transition exit time.

[0050] It should be noted that the real-time calculation of the average slope is achieved using a sliding window statistical analysis technique. The system maintains a first-in-first-out (FIFO) data queue with a length equal to the preset number of sampling points. Whenever the sensor acquires a new temperature data point, the system pushes it into the queue and removes the oldest data point. Subsequently, the system performs linear regression analysis on all data points in the queue, constructs a fitted straight line, and extracts the slope of this line as the current average slope. Compared to the simple two-point difference method, this method can effectively utilize multi-point data to smooth high-frequency random noise and reflect the true macroscopic trend of temperature changes.

[0051] It should be noted that the determination of the continuous sampling period and the exit time is implemented using state machine counting logic. A state counter is set up, with an initial value of zero. In each sampling period, the system compares the currently calculated average slope with the liquid phase slope threshold. If the average slope is greater than the threshold, the counter is incremented; if it is less than or equal to the threshold, the counter is immediately reset to zero. When the accumulated value of the counter reaches the preset number of confirmation periods, the state transition condition is met, and the phase transition process is confirmed to be completely over. At this time, the system backtracks the data log, finds the timestamp corresponding to the sampling period when the counter changed from zero to one, and locks it as the phase transition exit time.

[0052] It is worth noting that the determination of the preset number of sampling points is based on a comprehensive evaluation of the sensor sampling frequency and the signal-to-noise ratio. By performing spectral analysis on the sensor data under static isothermal conditions, the characteristic frequencies of high-frequency noise are identified, and a number of sampling points (e.g., 5 to 10) that can cover at least one complete noise cycle is selected to ensure the stability of the slope calculation, while avoiding a lag in response to rapid temperature rise due to an excessively long window.

[0053] It is worth noting that the determination of the preset number of confirmation cycles is based on statistical analysis of occasional disturbances in the welding environment. The duration of signal spikes caused by mechanical vibration or electromagnetic interference is tested in an actual workshop environment, and a number of cycles greater than the maximum spike duration (e.g., 3 to 5 consecutive cycles) is selected to prevent misjudgment of the phase transition's end due to a single instantaneous signal jump.

[0054] It is worth noting that the determination of the liquid phase slope threshold is based on calculations of the thermophysical properties of the solder alloy. According to the law of conservation of energy, in the liquid phase under full-power heating with no latent heat loss, the theoretical heating rate of the solder depends on its specific heat capacity. The system calculates the value of this theoretical rate and selects a percentage of this value (e.g., 60% to 80%) as the threshold. This threshold setting is significantly higher than the zero slope during the phase transition plateau, while also allowing for a safety margin for fluctuations in actual heating efficiency.

[0055] For example, the system sets the preset sampling point count to 5, the preset confirmation cycle count to 3, and the liquid phase slope threshold to 1.0 degrees Celsius per second. At the 130th second of welding, the real-time calculated average slope is 0.2 degrees Celsius per second (at a plateau). At the 132nd second, due to the end of latent heat absorption, the temperature begins to rise, and the calculated average slope is 1.1 degrees Celsius per second (exceeding the threshold for the first time), and the counter is recorded as 1. At the 132.1st second, the slope is 1.2 degrees Celsius per second, and the counter is recorded as 2. At the 132.2nd second, the slope is 1.3 degrees Celsius per second, and the counter is recorded as 3, reaching the preset confirmation cycle count. The system determines that the phase transition has ended and determines the moment the counter first triggers (the 132nd second) as the phase transition exit moment.

[0056] In step S17, based on the heat load characteristic parameters and the current temperature value corresponding to the phase transition exit time, the total heat requirement required to reach the preset welding target state is calculated, including: Calculate the temperature difference gap between the preset wetting peak threshold and the current temperature value corresponding to the phase change exit time; The equivalent heat capacity of the pads is estimated by combining the heat load characteristic parameters with the preset equipment thermal efficiency coefficient. According to the energy conservation formula, the product of the equivalent heat capacity and the temperature difference gap is calculated as the basic energy requirement; The basic energy requirement is corrected by introducing a preset heat loss compensation factor to obtain the total heat requirement.

[0057] In step S17, the total heat demand is converted into a wetting heating control signal, and the welding equipment is driven to operate until the temperature reaches a preset wetting peak threshold, including: Based on the rated power of the welding equipment, the total heat demand is converted into a heating pulse or a high duty cycle PWM signal of corresponding duration, which is used as the wetting heating control signal. The wetting and heating control signal is executed, and the temperature trend at the next moment is predicted using a Kalman filter algorithm during the heating process; When the temperature trend is about to reach or exceed the wetting peak threshold in the next control cycle, the output of the wetting heating control signal is terminated in advance.

[0058] It should be noted that the calculation of the temperature gap and basic energy requirement is achieved using differential and multiplication operations. The system first reads a preset wetting peak threshold and subtracts the actual temperature value collected by the sensor at the moment of phase transition exit; the difference is the temperature gap. Then, the system calls the normalized heat load characteristic parameter generated in step S11 and multiplies it with a preset equipment thermal efficiency coefficient (which characterizes the conversion ratio of a dimensionless parameter to physical heat capacity units). The calculation result is defined as the equivalent heat capacity value of the pad. Finally, according to the first law of thermodynamics, the system multiplies this equivalent heat capacity value by the temperature gap to calculate the theoretical energy required to raise the pad from the current temperature to the peak temperature, i.e., the basic energy requirement.

[0059] It should be noted that the introduction of a preset heat dissipation loss compensation factor for correction is achieved using weighted gain technology. Due to continuous heat convection and radiation losses in the welding environment, theoretical energy is usually insufficient to reach the target temperature. The system reads the preset heat dissipation loss compensation factor (typically greater than 1.0) and multiplies the basic energy requirement by this factor, thereby amplifying the energy budget. The final product obtained is the total heat requirement, which represents the total number of joules that the heating equipment actually needs to output to overcome environmental heat dissipation and complete the heating task.

[0060] It should be noted that the conversion and execution of the control signal are achieved using energy-time conversion and pulse width modulation technology. The system reads the rated output power parameters of the welding equipment, performs a division operation, divides the total heat demand by the rated power, and the quotient is the required full-power heating duration. Based on this, the system generates a drive pulse that remains high (or with a 100% duty cycle) during this duration as the wetting heating control signal, and sends it to the power controller to drive the heating element to run at full speed, thereby achieving rapid heating and spreading of the liquid solder.

[0061] It should be noted that the use of the Kalman filter algorithm to predict temperature trends and terminate early is achieved through state estimation and predictive control techniques. The system constructs a state-space model that includes both the temperature state and the temperature change rate state. During the heating process, the system inputs the sensor's measured values ​​into the Kalman filter every control cycle (e.g., 10 milliseconds), performs two steps: "prediction-correction," updates the current optimal state estimate, and deduces the predicted temperature value for the next control cycle based on this state. The system compares this predicted temperature value with the wetting peak threshold in real time. Once the predicted value is greater than or equal to the threshold, the system does not wait for the current cycle to end and immediately forces the control signal to a low level (0% duty cycle), utilizing the residual heat inertia of the heating element to complete the final surge, thereby avoiding temperature overshoot.

[0062] It is worth noting that the preset values ​​of the core process noise covariance matrix and measurement noise covariance matrix of the Kalman filter are based on offline statistical analysis of the hardware noise characteristics of the welding system. For the measurement noise covariance matrix, the system continuously samples the temperature sensor at high frequency under a static temperature condition, calculates the variance of the sampling sequence, and directly uses it as the intensity benchmark of the measurement noise. For the process noise covariance matrix, the system performs an open-loop constant power heating test under standard load, compares the actual temperature rise curve with the trajectory of the theoretical thermodynamic model, statistically analyzes the dynamic residual distribution between the two, and selects the variance of the residual as a measure of the uncertainty of the system model. This parameter initialization method based on measured statistical data ensures that the filter can dynamically balance the weights of model prediction and sensor observation, effectively filtering out high-frequency electromagnetic interference while maintaining a keen ability to track rapid temperature changes.

[0063] It is worth noting that the preset wetting peak threshold is determined based on the solder alloy's process data sheet. A temperature point 30 to 40 degrees Celsius above the solder melting point (e.g., 245 degrees Celsius for SAC305 alloy) is selected as the threshold to ensure optimal intermetallic compound (IMC) growth thickness. The determination of the equipment's thermal efficiency coefficient is based on calorimetric experiments using standard samples. A standard copper block with known heat capacity is heated in an adiabatic environment, and the ratio of input electrical energy to actual heat absorption is calculated and averaged to serve as the preset coefficient. The determination of the preset heat dissipation loss compensation factor is based on the determination of Newton's law of cooling. The natural cooling rate of the sample is measured at the operating temperature, the heat loss per unit time is calculated, and the compensation ratio is then determined.

[0064] For example, at the phase transition exit moment (132 seconds), the sensor measures the current temperature as 220 degrees Celsius. The preset wetting peak threshold is 245 degrees Celsius. The system performs a subtraction operation, calculating the temperature difference gap to be 25 degrees Celsius. The heat load characteristic parameter determined in step S11 is 0.33, and the preset equipment thermal efficiency coefficient is 1515 joules per degree Celsius. The system performs a multiplication operation (0.33 multiplied by 1515), estimating the equivalent heat capacity of the pad to be 500 joules per degree Celsius. Next, the system calculates the basic energy requirement as 12500 joules (500 multiplied by 25). The preset heat loss compensation factor is 1.2. The system calculates the total heat requirement as 15000 joules (12500 multiplied by 1.2). The rated power of the welding equipment is known to be 2000 watts. The system performs a division operation (15000 divided by 2000), obtaining a heating duration of 7.5 seconds. The system generates a full-power pulse lasting 7.5 seconds and begins execution. At the 7.2-second mark of heating, the Kalman filter predicts that the temperature will reach 245.2 degrees Celsius at the next moment (7.21 seconds), exceeding the threshold of 245 degrees Celsius. The system then immediately cuts off the heating signal at the 7.2-second mark, using residual heat to precisely stabilize the final temperature at approximately 245 degrees Celsius.

[0065] In summary, the embodiments of the present invention construct thermal load characteristic parameters that characterize the inherent thermal capacity of the solder pads by performing constant power detection and calculating the heating rate during the preheating stage, and then matching the optimal reference power curve based on this feedforward, effectively overcoming the technical difficulty that traditional fixed processes cannot adapt to differences in circuit board loads. This method establishes a physical mechanism-driven fully closed-loop adaptive control system, which significantly improves the intelligence level, control accuracy, and product yield of the soldering process in complex electronic assembly scenarios.

[0066] Reference Figure 2 The second embodiment of the present invention provides an automated control system for circuit board welding, comprising: The heat load parameter determination module is used to obtain the preheating temperature sequence of the welding equipment when it is running at a preset constant test power, calculate the heating rate of the preheating temperature sequence, and determine the ratio of the preset constant test power to the heating rate as the heat load characteristic parameter. The reference drive module is used to match the corresponding reference power curve from the preset reference power curve library according to the thermal load characteristic parameters, and use the reference power curve to drive the welding equipment to run, while collecting real-time temperature data streams during the welding process. The phase transition entry determination module is used to perform second derivative calculation on the real-time temperature data stream when the value of the real-time temperature data stream enters the preset phase transition temperature range, and determine the phase transition entry time when the calculation result shows that the value of the second derivative changes from a positive value to a negative value and satisfies the preset zero crossover condition. The slope deviation calculation module is used to set the phase change cut-in time as the starting point, continuously calculate the real-time temperature rise slope of the real-time temperature data stream, and compare the real-time temperature rise slope with the preset ideal phase change slope to obtain the slope deviation value. The dynamic power adjustment module is used to calculate the latent heat compensation amount using the slope deviation value, generate a dynamic power adjustment command superimposed with the latent heat compensation amount, and drive the welding equipment to execute the dynamic power adjustment command. The phase change exit determination module is used to detect the changing trend of the real-time heating slope, and determine the phase change exit time when the real-time heating slope exceeds the preset liquid phase slope threshold. The wetting and heating control module is used to calculate the total heat demand required to reach the preset welding target state based on the heat load characteristic parameters and the current temperature value corresponding to the phase change exit time, convert the total heat demand into a wetting and heating control signal, and drive the welding equipment to run until the temperature reaches the preset wetting peak threshold.

[0067] It should be noted that the circuit board welding automation control system provided in this embodiment of the invention is used to execute all the process steps of the circuit board welding automation control method in the above embodiment. The working principle and beneficial effects of the two are one-to-one, so they will not be described again.

[0068] It should be noted that the system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0069] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. An automated control method for circuit board soldering, characterized in that, include: Obtain the preheating temperature sequence of the welding equipment when it is running at a preset constant test power, calculate the heating rate of the preheating temperature sequence, and determine the ratio of the preset constant test power to the heating rate as a heat load characteristic parameter. According to the heat load characteristic parameters, the corresponding reference power curve is matched from the preset reference power curve library, and the reference power curve is used to drive the welding equipment to run, while the real-time temperature data stream of the welding process is collected. When the value of the real-time temperature data stream enters the preset phase transition temperature range, the second derivative operation is performed on the real-time temperature data stream. When the calculation result shows that the value of the second derivative changes from positive to negative and satisfies the preset zero crossover condition, the phase transition cut-off time is determined. The phase transition cut-off time is set as the starting point, and the real-time temperature rise slope of the real-time temperature data stream is continuously calculated. The real-time temperature rise slope is then compared with the preset ideal phase transition slope to obtain the slope deviation value. The latent heat compensation amount is calculated using the slope deviation value, a dynamic power adjustment command superimposed with the latent heat compensation amount is generated, and the welding equipment is driven to execute the dynamic power adjustment command. The change trend of the real-time heating slope is detected, and when the real-time heating slope exceeds the preset liquid phase slope threshold, the phase change exit time is determined. Based on the heat load characteristic parameters and the current temperature value corresponding to the phase transition exit time, the total heat demand required to reach the preset welding target state is calculated, the total heat demand is converted into a wetting heating control signal, and the welding equipment is driven to run until the temperature reaches the preset wetting peak threshold.

2. The automated control method for circuit board welding according to claim 1, characterized in that, The process of acquiring the preheating temperature sequence of the welding equipment under a preset constant test power, calculating the heating rate of the preheating temperature sequence, and determining the ratio of the preset constant test power to the heating rate as a heat load characteristic parameter includes: Extract the data segment from the preheating temperature sequence whose temperature falls within the preset linear temperature rise range; The least squares method is used to perform linear fitting on the data segment, and the slope of the fitted line is calculated as the heating rate. Calculate the ratio of the preset constant test power to the heating rate, and normalize the ratio to obtain the thermal load characteristic parameters.

3. The automated control method for circuit board welding according to claim 1, characterized in that, The step of matching a corresponding reference power curve from a pre-set reference power curve library based on the heat load characteristic parameters, and using the reference power curve to drive the welding equipment, while simultaneously acquiring real-time temperature data streams during the welding process, includes: According to the preset reference power curve library, each set of reference power curves corresponds to a heat load range. The larger the heat load characteristic parameter, the higher the preset basic power output of the corresponding reference power curve in the phase change stage. Determine the target heat load range to which the heat load characteristic parameters belong; The system retrieves a curve from the preset reference power curve library that matches the target heat load range as the reference power curve, and controls the energy output unit of the welding equipment according to the time power setting value of the reference power curve to collect the real-time temperature data stream.

4. The automated control method for circuit board welding according to claim 1, characterized in that, When the value of the real-time temperature data stream enters the preset phase transition temperature range, a second derivative operation is performed on the real-time temperature data stream. When the calculation result shows that the second derivative value changes from positive to negative and a preset zero-crossing condition is met, the phase transition cut-off time is determined, including: Gaussian smoothing filtering is applied to the real-time temperature data stream entering the preset phase transition temperature range to generate a smooth temperature sequence. Calculate the first and second derivative sequences of the smoothed temperature sequence with respect to time; The second derivative sequence is monitored in real time. When the value of the second derivative changes from positive to negative after crossing zero, and the number of subsequent consecutive preset stable verification cycles is less than the preset negative judgment threshold, the preset zero crossover condition is determined to be met. The starting time point that satisfies the preset zero-crossing condition is marked as the phase transition cut-in time.

5. The automated control method for circuit board welding according to claim 1, characterized in that, The process begins by setting the phase transition cutoff time as the starting point, continuously calculating the real-time temperature rise slope of the real-time temperature data stream, and comparing the real-time temperature rise slope with the preset ideal phase transition slope to obtain the slope deviation value, including: Set the phase transition cutoff time as the time anchor point and start the sliding monitoring window; The slope of the temperature data within the sliding monitoring window is calculated using a linear regression algorithm and used as the real-time temperature rise slope. The target temperature rise rate corresponding to the current reference power curve is read as the ideal phase transition slope. The slope deviation value is obtained by calculating the difference between the real-time heating slope and the ideal phase transition slope.

6. The automated control method for circuit board welding according to claim 1, characterized in that, The step of calculating the latent heat compensation amount using the slope deviation value, generating a dynamic power adjustment command superimposed with the latent heat compensation amount, and driving the welding equipment to execute the dynamic power adjustment command includes: Construct a power compensation model that includes proportional and integral components; The slope deviation value is input into the proportional element of the power compensation model to calculate the instantaneous deviation compensation component. The slope deviation value is integrated and accumulated on the time axis after the phase transition cut-in time to calculate the thermal hysteresis compensation component; The latent heat compensation is obtained by summing the instantaneous deviation compensation component and the thermal hysteresis compensation component, and then superimposed on the base power at the current moment to generate the dynamic power adjustment command.

7. The automated control method for circuit board welding according to claim 1, characterized in that, The detection of the changing trend of the real-time heating slope, and the determination of the phase transition exit time when the real-time heating slope exceeds a preset liquid phase slope threshold, includes: Temperature points are collected according to a preset number of sampling points, and the average slope of the temperature points is calculated in real time as the real-time temperature rise slope. Determine whether the real-time heating slope is greater than the liquid phase slope threshold within a preset number of sampling periods; If the judgment result is yes, then the first sampling time that meets the condition is determined as the phase transition exit time.

8. The automated control method for circuit board welding according to claim 1, characterized in that, The calculation of the total heat requirement to reach the preset welding target state based on the heat load characteristic parameters and the current temperature value corresponding to the phase transition exit time includes: Calculate the temperature difference gap between the preset wetting peak threshold and the current temperature value corresponding to the phase change exit time; The equivalent heat capacity of the pads is estimated by combining the heat load characteristic parameters with the preset equipment thermal efficiency coefficient. According to the energy conservation formula, the product of the equivalent heat capacity and the temperature difference gap is calculated as the basic energy requirement; The basic energy requirement is corrected by introducing a preset heat loss compensation factor to obtain the total heat requirement.

9. The automated control method for circuit board welding according to claim 1, characterized in that, The step of converting the total heat demand into a wetting heating control signal and driving the welding equipment to operate until the temperature reaches a preset wetting peak threshold includes: Based on the rated power of the welding equipment, the total heat demand is converted into a heating pulse or a high duty cycle PWM signal of corresponding duration, which is used as the wetting heating control signal. The wetting and heating control signal is executed, and the temperature trend at the next moment is predicted using a Kalman filter algorithm during the heating process; When the temperature trend is about to reach or exceed the wetting peak threshold in the next control cycle, the output of the wetting heating control signal is terminated in advance.

10. An automated control system for circuit board welding, characterized in that, include: The heat load parameter determination module is used to obtain the preheating temperature sequence of the welding equipment when it is running at a preset constant test power, calculate the heating rate of the preheating temperature sequence, and determine the ratio of the preset constant test power to the heating rate as the heat load characteristic parameter. The reference drive module is used to match the corresponding reference power curve from the preset reference power curve library according to the thermal load characteristic parameters, and use the reference power curve to drive the welding equipment to run, while collecting real-time temperature data streams during the welding process. The phase transition entry determination module is used to perform second derivative calculation on the real-time temperature data stream when the value of the real-time temperature data stream enters the preset phase transition temperature range, and determine the phase transition entry time when the calculation result shows that the value of the second derivative changes from a positive value to a negative value and satisfies the preset zero crossover condition. The slope deviation calculation module is used to set the phase change cut-in time as the starting point, continuously calculate the real-time temperature rise slope of the real-time temperature data stream, and compare the real-time temperature rise slope with the preset ideal phase change slope to obtain the slope deviation value. The dynamic power adjustment module is used to calculate the latent heat compensation amount using the slope deviation value, generate a dynamic power adjustment command superimposed with the latent heat compensation amount, and drive the welding equipment to execute the dynamic power adjustment command. The phase change exit determination module is used to detect the changing trend of the real-time heating slope, and determine the phase change exit time when the real-time heating slope exceeds the preset liquid phase slope threshold. The wetting and heating control module is used to calculate the total heat demand required to reach the preset welding target state based on the heat load characteristic parameters and the current temperature value corresponding to the phase change exit time, convert the total heat demand into a wetting and heating control signal, and drive the welding equipment to run until the temperature reaches the preset wetting peak threshold.

Images