Adaptive optimization method for surface mounting process parameter based on automatic optical detection feedback

Abstract

An adaptive optimization method for a surface mounting process parameter based on automatic optical detection feedback. The present invention belongs to the technical field of electrical appliances and electrical engineering, and relates to an automatic optimization method for a surface mounting process parameter. The purpose of the present invention is to solve the problems with an existing surface mounting process parameter optimization method of relying on manual adjustment, not being able to effectively and dynamically adaptively regulate each parameter, and not being able to ensure a regulation effect, resulting in reduced mounting accuracy and efficiency, and increased mounting defects and manual repair work. Disclosed is an adaptive optimization method for a surface mounting process parameter based on automatic optical detection feedback. A mounting height is adjusted on the basis of a Bayesian optimization method, a blowing delay is adjusted on the basis of a particle swarm optimization method, and a mounting offset is automatically compensated on the basis of a moving average method. The adaptive optimization method based on automatic optical detection feedback is completely provided. Various parameters are adjusted to reduce defects in a mounting process, and mounting accuracy and efficiency are improved.

Description

Adaptive Optimization Method for Surface Mounting Process Parameters Based on Automatic Optical Inspection Feedback Technical Field

[0001] This invention relates to an automatic optimization method for surface mount process parameters, belonging to the field of electrical technology and electrical engineering. Background Technology

[0002] In modern life and production, various electrical devices are widely used in daily life and industrial production. The functionality of these devices depends on circuit boards, which contain various electronic components and chips, and coordinate the operation of each component through electrical connections. To efficiently assemble fully functional circuit boards, surface mount technology has been widely used, with pick-and-place machines being key equipment. These machines can quickly and accurately pick up various components and rapidly mount them to the target positions on the circuit board surface. For simplicity, the following text will refer to electronic components, electronic devices, and chips collectively as "components," and the picking and placing process will be simply referred to as "pick-and-place."

[0003] The pick-and-place machine in Figure 2 uses a dual-drive gantry platform, where X-axis and Y-axis motors drive the placement carrier to move along X-axis and Y-axis guide rails. A feeder base is fixed with a feeder that automatically supplies components and waits for the pick-up bar. Simultaneously, a conveyor belt transports the circuit board to the designated position for placement.

[0004] Inside the placement carrier shown in Figure 3, the vacuum suction rod is controlled by a Z-axis drive motor to achieve vertical movement, ensuring its end accurately contacts the component surface. The suction rod end is equipped with a nozzle, which reliably picks up and places components by generating suction and airflow. To precisely rotate the picked-up component to the required placement angle, a θ-axis motor is mounted behind the Z-axis motor for easy rotation around the central axis. Matching nozzles must be selected to accommodate various component types; nozzle replacement is achieved using an automatic nozzle changer.

[0005] The automated optical inspection equipment shown in Figure 4 employs a dual-drive gantry motion platform. The platform's X-axis and Y-axis motors drive a high-resolution automated optical inspection camera to inspect the placement results of components on the circuit board. This equipment evaluates the difference between the actual placement result and the standard placement result based on a template matching method to obtain the placement offset. It then uses a deep learning-based classification method to detect various types of defects and identify the source of errors, such as whether the defect is caused by insufficient placement height.

[0006] In traditional surface mount technology (SMT) production, technicians can use their expertise to adjust the surface mount process parameters of the pick-and-place machine based on the inspection results of automated optical inspection equipment. For each placement point, the surface mount process parameters include: X-axis offset compensation value, Y-axis offset compensation value, θ-axis deflection compensation value, placement height compensation value, and placement air blow delay. These parameters have a significant impact on component placement yield, placement accuracy, and placement efficiency.

[0007] To analyze the causes of various placement defects, it is first necessary to understand the inherent deviations in the mechanical structure. The machine itself has machining and assembly errors, therefore it is necessary to compensate for the offset of the X-axis placement coordinates, Y-axis placement coordinates, and θ-axis rotation angle of the placement point. As the pick-and-place machine operates continuously for extended periods, dynamic changes such as thermal deformation occur, necessitating real-time online compensation, which is currently impossible to achieve manually.

[0008] At the micrometer scale, any assembly deviation significantly affects the contact between the circuit board surface and the suction cup tip. Figure 5 illustrates the assembly deviation angles of the suction cup, where α and β represent the deviation angles of the Z-axis and θ-axis motors in three-dimensional space, respectively. These angular deviations cause the suction cup tip to no longer remain stationary when the θ-axis motor rotates 360 degrees, but instead form an approximately elliptical trajectory in a non-horizontal plane. This not only indicates the existence of linear and angular deviations, but also implies a height difference Δz between the suction cup's highest and lowest points. Therefore, different mounting height compensations are required for different mounting angles. Previous mounting height adjustments relied on extensive manual trial and error. Furthermore, the heights of different components vary with the thickness of the circuit board, and the circuit board may experience height variations due to warping; traditional adjustment methods struggle to adapt to these changes.

[0009] When the suction rod reaches the designated placement position and height, the air blowing process begins. This action stops the suction of the vacuum nozzle, allowing the component to fall and completing the placement. The duration of the air blowing process directly affects the reliability of the placement. While extending the air blowing time can improve the placement result, excessively long air blowing times will reduce the overall circuit board placement efficiency. Conversely, if the air blowing time is too short, components that have already been placed may be lifted again by residual vacuum suction, leading to placement deviations or even defects. As the air blowing delay approaches a critical threshold, increasing the delay time can improve placement accuracy, which can be used as an indicator for adjusting the air blowing delay. Typically, a longer air blowing time is set initially, and then gradually shortened. In the past, the air blowing delay was manually set, but this method yielded coarse adjustments and made it difficult to find the shortest air blowing delay that guarantees placement results.

[0010] By leveraging Industrial Internet of Things (IIoT) technology, pick-and-place machines and automated optical inspection (OIO) equipment can operate seamlessly as a unified system, significantly improving placement quality and efficiency. EtherCAT, an Ethernet technology designed for real-time control applications in automation, is used to integrate these two machines. This integration allows direct transmission of component registration information from the pick-and-place machine to the OIO, simplifying and accelerating the data preparation process. By associating and storing the placement accuracy and defect type of each placement point with its corresponding process parameters, it is possible to independently adjust the parameters of each placement point, enabling personalized adjustments under various production factors, thereby improving placement accuracy and yield.

[0011] The main drawback of current research is that existing surface mount process parameter optimization methods rely on manual adjustment, which cannot effectively and dynamically adapt to control each parameter, and cannot guarantee the control effect. This leads to reduced mounting accuracy and efficiency, and an increase in mounting defects and manual repair work.

[0012] Summary of the Invention

[0013] The purpose of this invention is to address the problems that existing surface mount process parameter optimization methods rely on manual adjustment, which cannot effectively and dynamically adapt to control each parameter and cannot guarantee the control effect, resulting in reduced mounting accuracy and efficiency, and increased mounting defects and manual repair work. Therefore, this invention proposes an adaptive optimization method for surface mount process parameters based on automatic optical detection feedback.

[0014] The specific process of the adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback is as follows:

[0015] Step 1: Obtain the data of the circuit board to be mounted, initialize the surface mount process parameters and the search range of the surface mount process parameters, initialize the parameters of the mounting height adaptive optimization method based on Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on moving average method, and the parameters of the air blowing delay adaptive optimization method based on particle swarm optimization algorithm; the specific process is as follows:

[0016] Step 11: Obtain the data file of the circuit board to be mounted. The data file of the circuit board to be mounted includes the serial number of each mounting point, the X-axis coordinate, the Y-axis coordinate, and the θ-axis angle.

[0017] Step 1 & 2: Set the total number K of circuit boards to be mounted;

[0018] Step 13: Initialize surface mount process parameters;

[0019] The surface mount process parameters include the X-axis offset compensation value δX, Y-axis offset compensation value δY, θ-axis deflection compensation value δθ, mounting height compensation value δZ, and mounting air blowing delay T for each mounting point. b ;

[0020] Step 1, Section 4: Initialize the search range for surface mount process parameters. The search range for δX is SS. δX =[-100,100], the search range of δY is SS δY =[-100,100], the search range of δθ is SS δθ =[-1,1], the search range of δZ is SS δZ =[-500,500]、T b The search scope is SS Tb =[5,50];

[0021] SS δX The unit is μm, SS δY The unit is μm, SS δθ The unit is °, SS δZ The units are μm and SS. Tb The unit is ms;

[0022] Step 15: For the adaptive optimization method of mounting height based on Bayesian optimization algorithm, initialize the exploration-utilization parameter ξ = 3, the initial mounting height compensation value δZ = 0; initialize the number of mounting height observation points OB = 30: SZ = linspace(SS δZ ,OB);

[0023] OB represents the total number of mounting height observation points in the Bayesian optimization algorithm, and SZ represents the observation point array. δZ ,OB) refers to SS δZ Within the range, generate an array of OB elements at equal intervals;

[0024] Step 16: For the adaptive optimization method of mounting alignment based on the moving average method, initialize the size of the moving average window W = 20; the accumulated value of the X-axis offset ε X =0, Y-axis offset accumulated value ε Y =0, cumulative θ-axis offset ε θ =0;

[0025] Step 17: For the adaptive optimization method of blowing delay based on particle swarm optimization algorithm, initialize the number of particles N=5, randomly select N mounting points for blowing delay testing, set the particle count variable n=1, the individual learning factor c1=1 for each particle, the social learning factor c2=1 for each particle, and the inertia factor w=0.5. In the search space SS... Tb Within, the positions pb of each particle are randomly initialized. n and speed VB n ;

[0026] Step 2: Based on the initialized surface mount process parameters, complete the mounting of the first circuit board. Perform automatic optical inspection on the mounted circuit board, and update the parameters of the mounting height adaptive optimization method based on Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on moving average method, the X-axis offset compensation value, the Y-axis offset compensation value and the θ-axis deflection compensation value, and the parameters of the air blowing delay adaptive optimization method based on particle swarm optimization algorithm based on particle swarm optimization algorithm based on automatic optical inspection.

[0027] Step 3: Determine whether the circuit board counting variable k is less than or equal to the total number of circuit boards K;

[0028] If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five.

[0029] If not, it means that all circuit boards have been mounted and production has stopped;

[0030] Step 4: Execute the adaptive optimization method for mounting height based on Bayesian optimization algorithm to update the mounting height compensation value;

[0031] Step 5: Execute the adaptive optimization method for air blowing delay based on particle swarm optimization algorithm to update the mounting air blowing delay;

[0032] Step 6: Based on the updated surface mount process parameters from Steps 4 and 5, mount the next circuit board and perform automatic optical inspection on the mounted circuit board.

[0033] Step 7: Update the parameters of the mounting height adaptive optimization method based on the automatic optical inspection results from Step 6;

[0034] Step 8: Update the parameters of the blowing delay adaptive optimization method based on the automatic optical detection results from Step 6;

[0035] Step 9: Based on the offset of each mounting point detected by automatic optical detection in Step 6, execute the mounting alignment adaptive optimization method based on the moving average method to update the X-axis offset compensation value, Y-axis offset compensation value and θ-axis deflection compensation value.

[0036] The beneficial effects of this invention are as follows:

[0037] This invention discloses an adaptive optimization method for surface mount technology (SMT) process parameters based on automatic optical inspection feedback. It proposes a complete sequential optimization framework, adjusting SMT process parameters from highest to lowest importance. This achieves integrated operation of the pick-and-place machine and automatic optical inspection equipment, thereby improving the overall SMT quality and efficiency. Within the optimization framework, Bayesian optimization is used to optimize the placement height, particle swarm optimization to optimize the air blowing delay, and moving average method to automatically compensate for placement offset. A complete adaptive optimization method based on automatic optical inspection feedback is presented. By adjusting various parameters, defects are reduced and placement accuracy and efficiency are significantly improved during the placement process.

[0038] Specifically, this invention has the following two innovative aspects:

[0039] This invention links the mounting quality with surface mount process parameters, performs precise calculations and adaptive adjustments, and achieves integrated operation of the pick-and-place machine and automatic optical inspection equipment, thereby improving the mounting quality and efficiency of circuit boards.

[0040] This invention proposes a complete sequential optimization framework, in which an adaptive optimization method guided by automatic optical detection feedback is introduced. For the control of various parameters, Bayesian optimization and particle swarm optimization methods are adapted respectively, thereby reducing defects and improving mounting accuracy and efficiency during the mounting process. Attached Figure Description

[0041] Figure 1 is a flowchart of the adaptive optimization method for surface mounting process parameters based on automatic optical detection feedback according to the present invention.

[0042] Figure 2 is a schematic diagram of a chip mounter;

[0043] Figure 3 is a schematic diagram of the patch carrier;

[0044] Figure 4 is a schematic diagram of an automated optical inspection device;

[0045] Figure 5 is a schematic diagram of the assembly deviation angle of the suction rod. Detailed Implementation

[0046] Specific Implementation Method 1: The specific process of the adaptive optimization method for surface mounting process parameters based on automatic optical detection feedback in this implementation method is as follows:

[0047] Step 1: Obtain the data of the circuit board to be mounted, initialize the surface mount process parameters and the search range of the surface mount process parameters, initialize the parameters of the mounting height adaptive optimization method based on Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on moving average method, and the parameters of the air blowing delay adaptive optimization method based on particle swarm optimization algorithm; the specific process is as follows:

[0048] Step 1: Obtain the data file of the circuit board to be mounted. The data file of the circuit board to be mounted includes the serial number of each mounting point, the X-axis coordinate, the Y-axis coordinate, and the θ-axis angle (the X-axis coordinate, Y-axis coordinate, and θ-axis of the existing dual-drive gantry motion platform are also introduced at the bottom of the first page); as shown in Table 1 Circuit Board Data File;

[0049] Step 1 & 2: Set the total number K of circuit boards to be mounted;

[0050] Step 13: Initialize surface mount process parameters;

[0051] The surface mount process parameters include the X-axis offset compensation value δX, Y-axis offset compensation value δY, θ-axis deflection compensation value δθ, mounting height compensation value δZ, and mounting air blowing delay T for each mounting point. b ;See Table 2 for surface mount process parameter file;

[0052] Step 1, Section 4: Initialize the search range for surface mount process parameters. The search range for δX is SS. δX =[-100,100], the search range of δY is SS δY =[-100,100], the search range of δθ is SS δθ =[-1,1], the search range of δZ is SS δZ =[-500,500]、T b The search scope is SS Tb =[5,50];

[0053] SS δX The unit is μm, SS δY The unit is μm, SS δθ The unit is °, SS δZ The units are μm and SS. Tb The unit is ms;

[0054] Step 15: For the adaptive optimization method of mounting height based on Bayesian optimization algorithm, initialize the exploration-utilization parameter ξ = 3, the initial mounting height compensation value δZ = 0; initialize the number of mounting height observation points OB = 30: SZ = linspace(SS δZ ,OB);

[0055] Note: The parameters set in step one five will be applied to the adaptive optimization method for mounting height based on the Bayesian optimization algorithm in step four;

[0056] OB represents the total number of mounting height observation points in the Bayesian optimization algorithm, and SZ represents the observation point array. δZ ,OB) refers to SSδZ Within the range, generate an array of OB elements at equal intervals;

[0057] Step 16: For the adaptive optimization method of mounting alignment based on the moving average method, initialize the size of the moving average window W = 20; the accumulated value of the X-axis offset ε X =0, Y-axis offset accumulated value ε Y =0, cumulative θ-axis offset ε θ =0;

[0058] Note: The parameters set in step 16 will be applied to the adaptive optimization method for mounting alignment based on the moving average method in step 7;

[0059] Step 17: For the adaptive optimization method of air-blowing delay based on particle swarm optimization algorithm, initialize the number of particles N=5, randomly select N mounting points for air-blowing delay testing (referred to as "air-blowing test points"), particle count variable n=1, individual learning factor c1=1 for each particle, social learning factor c2=1 for each particle, inertia factor w=0.5, in the search space SS Tb Within, the positions pb of each particle are randomly initialized. n and speed VB n ;

[0060] Note: The parameters set in step 17 will be applied to the adaptive optimization method for blowing delay based on particle swarm optimization algorithm in step 5.

[0061] Step 2: Based on the initialized surface mount process parameters, complete the mounting of the first circuit board. Perform automatic optical inspection on the mounted circuit board, and update the parameters of the mounting height adaptive optimization method based on Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on moving average method, the X-axis offset compensation value, the Y-axis offset compensation value and the θ-axis deflection compensation value, and the parameters of the air blowing delay adaptive optimization method based on particle swarm optimization algorithm based on particle swarm optimization algorithm based on automatic optical inspection.

[0062] Step 3: Determine whether the circuit board counting variable k is less than or equal to the total number of circuit boards K;

[0063] If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five.

[0064] If not, it means that all circuit boards have been mounted and production has stopped;

[0065] Step 4: Execute the adaptive optimization method for mounting height based on Bayesian optimization algorithm to update the mounting height compensation value;

[0066] Step 5: Execute the adaptive optimization method for air blowing delay based on particle swarm optimization algorithm to update the mounting air blowing delay;

[0067] Step 6: Based on the updated surface mount process parameters from Steps 4 and 5, mount the next circuit board and perform automatic optical inspection on the mounted circuit board.

[0068] Step 7: Update the parameters of the mounting height adaptive optimization method based on the automatic optical inspection results from Step 6;

[0069] Step 8: Update the parameters of the blowing delay adaptive optimization method based on the automatic optical detection results from Step 6;

[0070] Step 9: Based on the offset of each mounting point detected by automatic optical detection in Step 6, execute the mounting alignment adaptive optimization method based on the moving average method to update the X-axis offset compensation value, Y-axis offset compensation value and θ-axis deflection compensation value.

[0071] Specific Implementation Method Two: This implementation method differs from Specific Implementation Method One in that, in step two, the first circuit board is mounted based on the initialized surface mount process parameters. The mounted circuit board is then subjected to automatic optical inspection. Based on the automatic optical inspection results, the mounting height adaptive optimization method based on the Bayesian optimization algorithm is updated, the mounting alignment adaptive optimization method parameters based on the moving average method are updated, the X-axis offset compensation value, the Y-axis offset compensation value, and the θ-axis deflection compensation value are updated, and the blowing delay adaptive optimization method parameters based on the particle swarm optimization algorithm are updated.

[0072] The specific process is as follows:

[0073] Step 2: Initialize surface mount process parameters, including: initial mount height compensation value δZ = 0, initial X-axis offset compensation value δX = 0, Y-axis offset compensation value δY = 0, θ-axis deflection compensation value δθ = 0, and set the air blowing delay at each air blowing test point to pb. n The blowing delay for all other mounting points (other mounting points are the remaining mounting points excluding the blowing test point) is set to a safety value T. ba =SS Tb (2); Initialize the historical mounting height array δz′=δZ;

[0074] The SS Tb (2) Delay time for air blowing during mounting b The upper limit of the search range; n∈[1:N];

[0075] Step 22: Based on the initialized surface mount process parameters, complete the mounting of the first circuit board; set the count variable k=1 for mounted circuit boards, perform automatic optical inspection on the mounted circuit boards, and obtain the results of the automatic optical inspection;

[0076] The results of automated optical inspection include the offset of each mounting point ∈ X ,∈ Y ,∈ θ Total number of mounting defects D 0 Defect sources are categorized as follows: defects caused by excessive mounting height, defects caused by insufficient air blowing delay, and defects caused by excessive mounting offset.

[0077] Note: Automated optical inspection is performed by an automated optical inspection machine. This equipment can measure the positional and angular offsets of each mounting point, classify existing mounting defects, and determine the source of the defects.

[0078] Steps two and three: Update the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm according to the automatic optical inspection results; the specific process is as follows:

[0079] Step 231: Based on the results of the automatic optical inspection, determine whether the defect is caused by excessive mounting height. If so, update the upper limit of the mounting height δZ search range and assign it the value SS. δZ (2) = δZ; if not, update the lower limit of the search range for the mounting height δZ and assign SS. δZ (1) = δZ;

[0080] Among them, SS δZ (2) The search range for the mounting height δZ is SS. δZ The upper limit of SS δZ (1) The search range for the mounting height δZ is SS. δZ The lower limit;

[0081] Step 232: Calculate the defect rate R = D 0 / P;

[0082] Initialize the historical defect rate array r = R;

[0083] Initialize the number of mounting defects D on the previous circuit board l =D 0 ;

[0084] Where P is the total number of mounting points on the circuit board;

[0085] Step 24: Update the parameters of the adaptive optimization method for mounting alignment based on the moving average method according to the automatic optical inspection results, and update the X-axis offset compensation value, Y-axis offset compensation value, and θ-axis deflection compensation value; the specific process is as follows:

[0086] Step 241: Add the offset of each decoration point to the cache array: B X =[B X (∈ X +δX)]、B Y =[B Y (∈ Y +δY)]、B θ =[B θ (∈ θ +δθ)];

[0087] Where δX represents the X-axis offset compensation value, δY represents the Y-axis offset compensation value, and δθ represents the θ-axis deflection compensation value; ∈ X ,∈ Y ,∈ θ Indicates the offset of each mounting point; B X B represents a cache array storing historical data of the X-axis mounting offset when not compensated. Y B represents a cache array storing historical data of the Y-axis mounting offset when not compensated. θ This represents a cache array that stores historical data of the θ-axis mounting offset when it is not compensated.

[0088] Step 242: Update the accumulated X-axis offset value ε X =ε X +∈ X +δX, Y axis offset cumulative value ε Y =ε Y +∈ Y +δY, θ axis offset cumulative value ε θ =ε θ +∈ θ +δθ;

[0089] Steps 2, 4, and 3: Obtain the length n of the cache array. B =length(B X );

[0090] Here, length() means getting the number of elements in the array within the parentheses;

[0091] Step 244: Update the X-axis offset compensation value δX = ε X / n B Y-axis offset compensation value δY=ε Y / n B θ-axis deflection compensation value δZ=ε θ / n B ;

[0092] Step 25: Update the parameters of the blowing delay adaptive optimization method based on particle swarm optimization algorithm according to the automatic optical detection results; the specific process is as follows:

[0093] Step 2.51: Initialize the particle count variable n = 1, and set the penalty weight value V = 10. 90 ;

[0094] Note: In the particle swarm optimization algorithm, a "particle" represents a candidate solution;

[0095] Step 252: Determine if n is less than N. If yes, proceed to Step 253; otherwise, proceed to Step 257.

[0096] Step 253: Based on the automatic optical inspection results, determine whether the nth air blowing test point has a placement defect caused by excessively short air blowing delay or excessive placement offset. If so, set ψ n =1; if not, then ψ n =0;

[0097] Where, ψ n is a binary variable, indicating whether there is a mounting defect at the nth air blowing test point;

[0098] Step 254: Calculate the fitness value of the particle. n =pb n *ψ n *V+pb n ;

[0099] Where * represents a multiplication sign;

[0100] Note: Fitness refers to the degree of goodness or badness of a solution during the optimization process, and is usually represented by a numerical value, namely the fitness value;

[0101] Note: The smaller the fitness value given in step 264, the better. This calculation function uses a penalty weight V to penalize situations where defects occur due to insufficient blowing delay.

[0102] Step 2.55: Update the optimal fitness value fo of particle n. n =fit n Update the individual optimal position po of particle n. n =pb n ;

[0103] Step 256: Update the particle count variable n = n + 1, and return to step 252;

[0104] Step 257: Obtain the optimal fitness value OF and the particle number j that has reached the optimal fitness value, [OF,j] = min(fit);

[0105] The min(·) function refers to finding the minimum value of the array within the parentheses and the index of the minimum value in the array.

[0106] Step 258: Determine if OF is greater than V. If so, it means that all tested air-blowing delays will lead to defects. Set the optimal air-blowing delay to the safe value OT. b =T ba If not, then set the optimal blowing delay OT. b =pb j ;

[0107] Among them, Pb j This represents the position of particle j.

[0108] The other steps and parameters are the same as in Specific Implementation Method 1.

[0109] Specific Implementation Method 3: This implementation method differs from Specific Implementation Method 1 or 2 in that, in step 3, it is determined whether the circuit board counting variable k is less than or equal to the total number of circuit boards K.

[0110] If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five.

[0111] If not, it means that all circuit boards have been mounted and production has stopped;

[0112] The specific process is as follows:

[0113] Step 31: Update the circuit board mounting count variable k = k + 1;

[0114] Step 3.2: Determine whether the circuit board counting variable k is less than or equal to the total number of circuit boards K;

[0115] If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five.

[0116] If so, proceed to step three.

[0117] Step 33: All circuit boards have been mounted; production has ceased.

[0118] Other steps and parameters are the same as in specific implementation method one or two.

[0119] Specific Implementation Method Four: This implementation method differs from Specific Implementation Methods One to Three in that step four executes a mounting height adaptive optimization method based on the Bayesian optimization algorithm to update the mounting height compensation value; the specific process is as follows:

[0120] Step 4: Based on the observation results, use the squared exponent to fit the historical value array of mounting height δz′ with the historical value array of defect rate r to obtain the Gaussian process regression model gprm=GPRM(δz′,r);

[0121] Wherein, GPRM is a Gaussian process regression model, and gprm is a Gaussian process regression model;

[0122] The observed result refers to the defect rate r when the mounting height is δz′.

[0123] Kernel functions are used to describe the correlation between data points; the squared exponential kernel is the most commonly used kernel function. It assumes that data points that are close to each other are highly correlated, and that the correlation decays exponentially as the distance increases.

[0124] Note: Gaussian Process Regression Model (GPRM) is a powerful non-parametric machine learning method commonly used for regression and classification tasks. It predicts new data points by utilizing the correlation (covariance function) between data points. The Gaussian process model can not only predict the mean of the observations, but also provide the standard deviation of this prediction.

[0125] Step 42: Based on the Gaussian process regression model gprm obtained in Step 41, predict the predicted mean μ(ob) and standard deviation v(ob) at each observation point SZ(ob).

[0126] Where ob∈[1:OB] is the element index in the observation point array SZ;

[0127] Step 4.3: Calculate rb = min(r), d = (rb - μ - ξ) / v;

[0128] The ". / " symbol means that the elements in the array before and after the symbol are divided one by one, and the quotient is stored as an array.

[0129] Where rb is the minimum defect rate in the historical defect rate array, r is the historical defect rate array, d is an array (used to guide the improvement amount of the next decision of the Bayesian optimization algorithm, reflecting the possible degree of improvement at the current position compared to the current known optimal solution), and ξ is the explore-utilize parameter, which is the parameter that balances the exploration of new solutions (exploration) and the utilization of known optimal solutions (utilization).

[0130] μ is the predicted mean array, and v is the standard deviation array;

[0131] Step 4: Calculate the expected improvement EI = d*Φ(d) + v*φ(d);

[0132] Where Φ(d) is the cumulative distribution function of the standard normal distribution, and φ(d) is the probability density function of the standard normal distribution;

[0133] Note: The cumulative distribution function Φ(d) of the standard normal distribution describes the cumulative probability that a random variable is less than or equal to each point in array d under the standard normal distribution.

[0134] The probability density function φ(d) of the standard normal distribution describes the probability density of each point in array d under the standard normal distribution;

[0135] Steps four and five: Weight the EI values ​​within the range of the optimal value; the specific process is as follows:

[0136] Note: Steps four and five aim to increase the probability of selecting observation points within the potential optimal value region, thereby more efficiently locating the optimal mounting height;

[0137] Step 451: Initialize the observation point index array obs = [1:OB];

[0138] obs = [1:OB] is an array used to refer to all observation points in the process of searching for the mounting height. Each element is 1:2:…:OB, where ob is the index variable of the mounting height observation point, and ob∈[1:OB] is a numerical value with a range of 1:2:…:OB.

[0139] Step 452: Obtain the smallest index outside the range of the optimal value: ob1 = find(SZ) <SS δZ (1) and the maximum index ob2 = find(SZ>SS) δZ (2));

[0140] The optimal value is the optimal mounting height that the optimization algorithm aims to achieve, and the range of the optimal value is the interval in which the optimal mounting height may exist.

[0141] Where SZ is the observation point array, SS δZ (2) The search range for the mounting height δZ is SS. δZ The upper limit of SS δZ (1) The search range for the mounting height δZ is SS. δZ The lower limit;

[0142] The find(·) function retrieves the index of the element in the array within the parentheses that satisfies the condition within the parentheses.

[0143] Step 453: Delete the indices outside the range of the optimal value: obs([ob1:ob2]) = [];

[0144] The [] is an empty array. Assigning the values ​​of ob1 and ob2 in the array to [] indicates clearing the ob1 and ob2 elements.

[0145] Step 454: Weight EI within the range of the optimal value, EI(obs) = EI(obs) * V;

[0146] The element index within the range is obs. EI(obs) is weighted, and the weighted EI also includes other elements besides obs, but those elements remain unchanged. Only the elements at the obs index are weighted.

[0147] EI = d*Φ(d) + v(obs)*φ(d), where EI is an array and obs is used as an index to retrieve the elements in EI corresponding to these indices;

[0148] Step 46: Select the next mounting height value δZ from the observation points corresponding to the maximum expected improvement; the specific process is as follows:

[0149] Step 461: Obtain the maximum value of the desired improvement, mei = max(EI);

[0150] The max(·) function retrieves the maximum value in the array within the parentheses.

[0151] EI represents an array containing all elements, including the weighted EI(obs) and other unweighted EI() arrays.

[0152] Step 462: Obtain the index array iδzs = find(EI == mei) of all elements that are equal to the maximum value mei to be improved;

[0153] Here, == represents a relational operator used to compare whether two values ​​are equal;

[0154] Step 463: Obtain the number of elements in the array iδzs: nδzs = length(iδzs);

[0155] Step 464: Randomly select the index of the next observation point iδz = ceil(rand(1)*nδzs);

[0156] rand(1) represents randomly obtaining a real number within (0,1); ceil(·) represents rounding up the array within the parentheses;

[0157] Step 465: Update the mounting height compensation value δZ=δz′(iδz).

[0158] The other steps and parameters are the same as those in one of the specific implementation methods one to three.

[0159] Specific Implementation Method Five: This implementation method differs from Specific Implementation Methods One to Four in that step five executes an adaptive optimization method for air blowing delay based on particle swarm optimization algorithm to update the mounting air blowing delay; the specific process is as follows:

[0160] Step 51: Initialize the particle count variable n = 1;

[0161] Step 52: Determine if n is less than N. If yes, proceed to Step 53; otherwise, proceed to Step 6.

[0162] Step 53: Update the position and velocity of the nth particle; the specific process is as follows:

[0163] Step 531: Determine if n is equal to j. If yes, proceed to step 54; otherwise, proceed to step 532.

[0164] Note: Considering the uncertainties in the placement process, the optimal air-blowing delay remains unchanged for each round in order to more fully and thoroughly evaluate the placement effect of the air-blowing delay;

[0165] Step 532: Update particle velocity (vb) n =w*vb n +c1*rand(1)*(po n -pb n )+c2*rand(1)*(OT b -pb n );

[0166] Step 533: Update particle position pb n =pb n +vb n ;

[0167] Step 54: Update the blowing delay of the nth blowing test point to pb. n ;

[0168] Step 55: Update the particle count variable n = n + 1, and return to Step 52.

[0169] The other steps and parameters are the same as those in one of the specific implementation methods one to four.

[0170] Specific Implementation Method Six: This implementation method differs from Specific Implementation Methods One through Five in that: in step six, the next circuit board is mounted based on the updated surface mount process parameters from steps four and five, and the mounted circuit board undergoes automatic optical inspection; the specific process is as follows:

[0171] Based on the updated surface mount process parameters in steps four and five, the mounting of the k-th circuit board is completed; automatic optical inspection is performed on the mounted circuit board; and the results of the automatic optical inspection are obtained.

[0172] The results of automated optical inspection include: total number of mounting defects D 0 The offset of each patch point ∈ X ,∈ Y ,∈ θ Source of defects;

[0173] The sources of defects are categorized as follows: defects caused by excessive mounting height, defects caused by insufficient air blowing delay, and defects caused by excessive mounting offset.

[0174] The other steps and parameters are the same as those in one of the specific implementation methods one to five.

[0175] Specific Implementation Method Seven: This implementation method differs from Specific Implementation Methods One through Six in that: in step seven, the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm are updated according to the automatic optical detection results in step six; the specific process is as follows:

[0176] Step 71: Based on the automatic optical inspection results, determine whether the defect is caused by excessive mounting height. If so, update the upper limit of the mounting height δZ search range and assign it the value SS. δZ (2) = δZ; if not, update the lower limit of the search range for the mounting height δZ and assign SS. δZ (1) = δZ;

[0177] Step 72: Calculate the defect rate R = D 0 / D l Update the number of mounting defects D on the previous circuit board. l =D 0 ;

[0178] Step 73: Update the historical mounting height array δz′=[δz′δZ];

[0179] Step 74: Update the historical defect rate array r = [r R].

[0180] The other steps and parameters are the same as those in one of the specific implementation methods one to six.

[0181] Specific Implementation Method Eight: This implementation method differs from Specific Implementation Methods One through Six in that: in step eight, the parameters of the blowing delay adaptive optimization method based on the particle swarm optimization algorithm are updated according to the automatic optical detection results in step six; the specific process is as follows:

[0182] Step 81: Initialize the particle count variable n = 1;

[0183] Step 82: Determine if n is less than or equal to N. If yes, proceed to step 83; otherwise, proceed to step 87.

[0184] Step 83: Based on the automatic optical inspection results, determine whether the nth air blowing test point has a placement defect caused by excessively short air blowing delay or excessive placement offset. If so, set ψ n =1; otherwise, set ψ n =0;

[0185] Step 84: Calculate the fitness value of the particle. n =pb n *ψ n *V+pb n ;

[0186] Step 85: Determine fit n Is it less than fo? n If we update the optimal fitness value fo of particle n... n =fit n Update the individual optimal position po of particle n. n =pb n Otherwise, it will not be updated;

[0187] Step 86: Update the particle count variable n = n + 1, and return to step 82;

[0188] Step 87: Obtain the optimal fitness value OF and the particle number j that has reached the optimal fitness value, [OF,j] = min(fit);

[0189] Step 88: Determine if OF is greater than V. If so, it means that all tested air-blowing delays will lead to defects. Set the optimal air-blowing delay to the safe value OT. b =T ba If not, then set the optimal blowing delay OT. b =pb j ;

[0190] Steps 8 and 9: Reduce the blowing delay T b Search range SS Tb .

[0191] The other steps and parameters are the same as those in any of the specific implementation methods one to seven.

[0192] Specific Implementation Method Nine: This implementation method differs from Specific Implementation Methods One to Eight in that: in steps eight and nine, the blowing delay T is reduced. b Search range SS Tb The specific process is as follows:

[0193] Step 891: Obtain the array of test point indices ns = find(ψ==1);

[0194] Where ψ represents an array of all air-blowing test points to indicate whether there are mounting defects;

[0195] Step 892: Update the search range T b The lower SS Tb (1) = max([pb ns SS Tb (1)]);

[0196] Among them, pb ns Represented as the position of particle number ns, SS Tb (1) indicates T b The lower bound;

[0197] Step 14 gives SS Tb , is an array of two values. Here, we update the first element, which represents the lower bound of the blowing delay search.

[0198] Step 893: Determine if OF is less than V. If so, update the search range T. b Upper SS Tb (2) = pb j If not, update SS. Tb (2) = SS Tb (2)+1;

[0199] Among them, SS Tb (2) indicates T b The upper boundary.

[0200] Step 14 gives SS Tb is an array of two values. Here, we update the second element, which represents the upper bound of the blowing delay search.

[0201] The other steps and parameters are the same as those in one of the specific implementation methods one to eight.

[0202] Specific Implementation Method Ten: This implementation method differs from Specific Implementation Methods One through Eight in that: in step nine, based on the offset of each mounting point automatically detected by optical detection in step six, a mounting alignment adaptive optimization method based on the moving average method is executed to update the X-axis offset compensation value, Y-axis offset compensation value, and θ-axis deflection compensation value; the specific process is as follows:

[0203] Step 91: Obtain the length n of the cache array. B =length(B X );

[0204] Step 92, Determine nB If it equals W, update the accumulated X-axis offset ε. X =ε X -B X (1) Accumulated value of Y-axis offset ε Y =ε Y -B Y (1) Accumulated value of θ-axis offset ε θ =ε θ -B θ (1) Delete the first element B from the cache array. X (1) = [], B Y (1) = [], B θ (1) = []; otherwise, do not update.

[0205] Among them, B X (1) indicates the first in the historical data of the X-axis mounting offset when it is not compensated, B Y (1) indicates the first in the historical data of the Y-axis mounting offset when it is not compensated, B θ (1) indicates the first in the historical data of the θ-axis mounting offset when it is not compensated. The [] is an empty array. Assigning some elements in the array to [] indicates clearing these elements.

[0206] Step 93: Add the offset of each decoration point to the cache array: B X =[B X (∈ X +δX)]、B Y =[B Y (∈ Y +δY)]、B θ =[B θ (∈ θ +δθ)];

[0207] Among them, B X =[B X (∈ X +δX)] represents (∈ X +δX) is stored in array B X B is the last element; Y =[B Y (∈ Y +δY)] represents (∈ Y +δY) is stored in array B Y B is the last element; θ =[B θ (∈ θ +δθ)] represents (∈ θ +δθ) is stored in array B θ The middle element is the last element;

[0208] Step 94: Update the accumulated X-axis offset value ε X =ε X +∈ X Y-axis offset cumulative value ε Y =ε Y +∈ Y ε, the cumulative value of the θ-axis offset θ =ε θ +∈ θ ;

[0209] Step 95: Update the X-axis offset compensation value δX = ε X / n B Y-axis offset compensation value δY=ε Y / n B θ-axis deflection compensation value δZ=ε θ / n B ;

[0210] Step 96: Return to Step 31.

[0211] The other steps and parameters are the same as those in any of the specific implementation methods one to nine.

[0212] The beneficial effects of the present invention are verified using the following embodiments:

[0213] Example 1:

[0214] The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback in this embodiment is prepared according to the following steps:

[0215] The case study considers the process of optimizing surface mount technology (SMT) parameters in collaboration between a pick-and-place machine and an automated optical inspection system. The components on the circuit board used for the placement test are in 0201 packages, which are the smallest in size. When the pick-and-place machine is running at its highest speed, misalignment and defects are more likely to occur when assembling such components. Ten placement experiments were conducted, and the average value was calculated to evaluate the optimization effect.

[0216] Using the adaptive optimization of surface mount process parameters based on automatic optical detection feedback, the experimental results show that the method of this patent performs an average of 5.7 adjustments to the mounting height to eliminate defects, resulting in a total of 88 defects during this process; the traditional manual adjustment method performs an average of 6.5 adjustments to the mounting height to eliminate defects, resulting in a total of 192 defects during this process. In terms of improving mounting accuracy, the method of this invention increases the Cpk value (the higher the better) from 0.5 to 1.87. Without causing a decrease in mounting performance, the method of this invention shortens the mounting air blowing delay from 17.5ms to 15.9ms.

[0217] This invention may have other embodiments. Without departing from the spirit and essence of this invention, those skilled in the art can make various corresponding changes and modifications according to this invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

[0218] Table 1 Circuit Board Data Files

[0219] Table 2 Surface Mount Process Parameters

[0220] This invention may have other embodiments. Without departing from the spirit and essence of this invention, those skilled in the art can make various corresponding changes and modifications according to this invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

An adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback is characterized by: The specific process of the method is as follows: Step 1: Obtain the data of the circuit board to be mounted, initialize the surface mount process parameters and the search range of the surface mount process parameters, initialize the parameters of the mounting height adaptive optimization method based on Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on moving average method, and the parameters of the air blowing delay adaptive optimization method based on particle swarm optimization algorithm; the specific process is as follows: Step 11: Obtain the data file of the circuit board to be mounted. The data file of the circuit board to be mounted includes the serial number of each mounting point, the X-axis coordinate, the Y-axis coordinate, and the θ-axis angle. Step 1 & 2: Set the total number K of circuit boards to be mounted; Step 13: Initialize surface mount process parameters; The surface mount process parameters include the X-axis offset compensation value δX, Y-axis offset compensation value δY, θ-axis deflection compensation value δθ, mounting height compensation value δZ, and mounting air blowing delay T for each mounting point. b ; Step 1, Section 4: Initialize the search range for surface mount process parameters. The search range for δX is SS. δX =[-100,100], the search range of δY is SS δY =[-100,100], the search range of δθ is SS δθ =[-1,1], the search range of δZ is SS δZ =[-500,500]、T b The search scope is SS Tb =[5,50]; SS δX The unit is μm, SS δY The unit is μm, SS δθ The unit is °, SS δZ The units are μm and SS. Tb The unit is ms; Step 15: For the adaptive optimization method of mounting height based on Bayesian optimization algorithm, initialize the exploration-utilization parameter ξ = 3, the initial mounting height compensation value δZ = 0; initialize the number of mounting height observation points OB = 30: SZ = linspace(SS δZ ,OB); OB represents the total number of mounting height observation points in the Bayesian optimization algorithm, and SZ represents the observation point array. δZ ,OB) refers to SS δZ Within the range, generate an array of OB elements at equal intervals; Step 16: For the adaptive optimization method of mounting alignment based on the moving average method, initialize the size of the moving average window W = 20; the accumulated value of the X-axis offset ε X =0, Y-axis offset cumulative value ε Y =0, cumulative θ-axis offset ε θ =0; Step 17: For the adaptive optimization method of blowing delay based on particle swarm optimization algorithm, initialize the number of particles N=5, randomly select N mounting points for blowing delay testing, set the particle count variable n=1, the individual learning factor c1=1 for each particle, the social learning factor c2=1 for each particle, and the inertia factor w=0.

5. In the search space SS... Tb Within, the positions pb of each particle are randomly initialized. n and speed VB n ; Step 2: Based on the initialized surface mount process parameters, complete the mounting of the first circuit board. Perform automatic optical inspection on the mounted circuit board, and update the parameters of the mounting height adaptive optimization method based on Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on moving average method, the X-axis offset compensation value, the Y-axis offset compensation value and the θ-axis deflection compensation value, and the parameters of the air blowing delay adaptive optimization method based on particle swarm optimization algorithm based on particle swarm optimization algorithm based on automatic optical inspection. Step 3: Determine whether the circuit board counting variable k is less than or equal to the total number of circuit boards K; If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five. If not, it means that all circuit boards have been mounted and production has stopped; Step 4: Execute the adaptive optimization method for mounting height based on Bayesian optimization algorithm to update the mounting height compensation value; Step 5: Execute the adaptive optimization method for air blowing delay based on particle swarm optimization algorithm to update the mounting air blowing delay; Step 6: Based on the updated surface mount process parameters from Steps 4 and 5, mount the next circuit board and perform automatic optical inspection on the mounted circuit board. Step 7: Update the parameters of the mounting height adaptive optimization method based on the automatic optical inspection results from Step 6; Step 8: Update the parameters of the blowing delay adaptive optimization method based on the automatic optical detection results from Step 6; Step 9: Based on the offset of each mounting point detected by automatic optical detection in Step 6, execute the mounting alignment adaptive optimization method based on the moving average method to update the X-axis offset compensation value, Y-axis offset compensation value and θ-axis deflection compensation value. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 1 is characterized in that: In step two, based on the initialized surface mount process parameters, the first circuit board is mounted. The mounted circuit board undergoes automatic optical inspection, and the parameters of the Bayesian optimization-based adaptive mounting height optimization method, the moving average-based adaptive mounting alignment optimization method, the X-axis offset compensation value, the Y-axis offset compensation value, and the θ-axis deflection compensation value are updated based on the automatic optical inspection results. The parameters of the particle swarm optimization-based adaptive air-blowing delay optimization method are also updated. The specific process is as follows: Step 2: Initialize surface mount process parameters, including: initial mount height compensation value δZ = 0, initial X-axis offset compensation value δX = 0, Y-axis offset compensation value δY = 0, θ-axis deflection compensation value δθ = 0, and set the air blowing delay at each air blowing test point to pb. n The blowing delay time for each other sticker point is set to a safe value T. ba =SS Tb (2); Initialize the historical mounting height array δz′=δZ; The SS Tb (2) Delay time for air blowing during mounting b The upper limit of the search range; n∈[1:N]; Step 22: Based on the initialized surface mount process parameters, complete the mounting of the first circuit board; set the count variable k=1 for mounted circuit boards, perform automatic optical inspection on the mounted circuit boards, and obtain the results of the automatic optical inspection; The results of automated optical inspection include the offset of each mounting point ∈ X ,∈ Y ,∈ θ Total number of mounting defects D 0 Defect sources are categorized as follows: defects caused by excessive mounting height, defects caused by insufficient air blowing delay, and defects caused by excessive mounting offset. Steps two and three: Update the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm according to the automatic optical inspection results; the specific process is as follows: Step 231: Based on the results of the automatic optical inspection, determine whether the defect is caused by excessive mounting height. If so, update the upper limit of the mounting height δZ search range and assign it the value SS. δZ (2) = δZ; if not, update the lower limit of the search range for the mounting height δZ and assign SS. δZ (1) = δZ; Among them, SS δZ (2) The search range for the mounting height δZ is SS. δZ The upper limit of SS δZ (1) The search range for the mounting height δZ is SS. δZ The lower limit; Step 232: Calculate the defect rate R = D 0 / P; Initialize the historical defect rate array r = R; Initialize the number of mounting defects D on the previous circuit board l =D 0 ; Where P is the total number of mounting points on the circuit board; Step 24: Update the parameters of the adaptive optimization method for mounting alignment based on the moving average method according to the automatic optical inspection results, and update the X-axis offset compensation value, Y-axis offset compensation value, and θ-axis deflection compensation value; the specific process is as follows: Step 241: Add the offset of each decoration point to the cache array: B X =[B X (∈ X +δX)]、B Y =[B Y (∈ Y +δY)]、B θ =[B θ (∈ θ +δθ)]; Where δX represents the X-axis offset compensation value, δY represents the Y-axis offset compensation value, and δθ represents the θ-axis deflection compensation value; ∈ X ,∈ Y ,∈ θ Indicates the offset of each mounting point; B X B represents a cache array storing historical data of the X-axis mounting offset when not compensated. Y B represents a cache array storing historical data of the Y-axis mounting offset when not compensated. θ This represents a cache array that stores historical data of the θ-axis mounting offset when it is not compensated. Step 242: Update the accumulated X-axis offset value ε X =ε X +∈ X +δX, Y axis offset cumulative value ε Y =ε Y +∈ Y +δY, θ axis offset cumulative value ε θ =ε θ +∈ θ +δθ; Steps 2, 4, and 3: Obtain the length n of the cache array. B =length(B X ); Here, length() means getting the number of elements in the array within the parentheses; Step 244: Update the X-axis offset compensation value δX = ε X / n B Y-axis offset compensation value δY=ε Y / n B θ-axis deflection compensation value δZ = ε θ / n B ; Step 25: Update the parameters of the blowing delay adaptive optimization method based on particle swarm optimization algorithm according to the automatic optical detection results; the specific process is as follows: Step 2.51: Initialize the particle count variable n = 1, and set the penalty weight value V = 10. 90 ; Step 252: Determine if n is less than N. If yes, proceed to Step 253; otherwise, proceed to Step 257. Step 253: Based on the automatic optical inspection results, determine whether the nth air blowing test point has a placement defect caused by excessively short air blowing delay or excessive placement offset. If so, set ψ n =1; if not, then ψ n =0; Where, ψ n is a binary variable, indicating whether there is a mounting defect at the nth air blowing test point; Step 254: Calculate the fitness value of the particle. n =pb n *ψ n *V+pb n ; Where * represents a multiplication sign; Step 2.55: Update the optimal fitness value fo of particle n. n =fit n Update the individual optimal position po of particle n. n =pb n ; Step 256: Update the particle count variable n = n + 1, and return to step 252; Step 257: Obtain the optimal fitness value OF and the particle number j that has reached the optimal fitness value, [OF,j] = min(fit); The min(·) function refers to finding the minimum value of the array within the parentheses and the index of the minimum value in the array. Step 258: Determine if OF is greater than V. If so, it means that all tested air-blowing delays will lead to defects. Set the optimal air-blowing delay to the safe value OT. b =T ba If not, then set the optimal blowing delay OT. b =pb j ; Among them, Pb j This represents the position of particle j. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 2 is characterized in that: In step three, it is determined whether the circuit board count variable k is less than or equal to the total number of circuit boards K; If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five. If not, it means that all circuit boards have been mounted and production has stopped; The specific process is as follows: Step 3:

1. Update the circuit board mounting count variable k = k + 1; Step 3.2: Determine whether the circuit board counting variable k is less than or equal to the total number of circuit boards K; If so, determine the automatic optical inspection result. If the automatic optical inspection result is due to the placement defect caused by the placement height being too high, proceed to step four. If the automatic optical inspection result is due to the placement defect caused by the air blowing delay being too short or the placement offset being too large, proceed to step five. If so, proceed to step three. Step 33: All circuit boards have been mounted; production has ceased. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 3 is characterized in that: In step four, a mounting height adaptive optimization method based on the Bayesian optimization algorithm is executed to update the mounting height compensation value; the specific process is as follows: Step 4: Based on the observation results, use the squared exponent to fit the historical value array of mounting height δz′ with the historical value array of defect rate r to obtain the Gaussian process regression model gprm=GPRM(δz′,r); Wherein, GPRM is a Gaussian process regression model, and gprm is a Gaussian process regression model; The observed result refers to the defect rate r when the mounting height is δz′. Step 42: Based on the Gaussian process regression model gprm obtained in Step 41, predict the predicted mean μ(ob) and standard deviation v(ob) at each observation point SZ(ob). Where ob∈[1:OB] is the element index in the observation point array SZ; Step 4.3: Calculate rb = min(r), d = (rb - μ - ξ) / v; The ". / " refers to dividing the elements in the array before and after the symbol one-to-one and storing the quotient as an array; Where rb is the minimum defect rate in the historical defect rate array, r is the historical defect rate array, d is an array, and ξ is the explore-exploit parameter; μ is the predicted mean array, and v is the standard deviation array; Step 4: Calculate the expected improvement EI = d*Φ(d) + v*φ(d); Where Φ(d) is the cumulative distribution function of the standard normal distribution, and φ(d) is the probability density function of the standard normal distribution; Steps four and five: Weight the EI values ​​within the range of the optimal value; the specific process is as follows: Step 451: Initialize the observation point index array obs = [1:OB]; Step 452: Obtain the smallest index outside the range of the optimal value: ob1 = find(SZ) <SS δZ (1) and the maximum index ob2 = find(SZ>SS) δZ (2)); Where SZ is the observation point array, SS δZ (2) The search range for the mounting height δZ is SS. δZ The upper limit of SS δZ (1) The search range for the mounting height δZ is SS. δZ The lower limit; The find(·) function retrieves the index of the element in the array within the parentheses that satisfies the condition within the parentheses. Step 453: Delete the indices outside the range of the optimal value: obs([ob1:ob2]) = []; The [] is an empty array; Step 454: Weight EI within the range of the optimal value, EI(obs) = EI(obs) * V; Step 46: Select the next mounting height value δZ from the observation points corresponding to the maximum expected improvement; the specific process is as follows: Step 461: Obtain the maximum value of the desired improvement, mei = max(EI); The max(·) function retrieves the maximum value in the array within the parentheses. Step 462: Obtain the index array iδzs = find(EI == mei) of all elements that are equal to the maximum value mei to be improved; Step 463: Obtain the number of elements in the array iδzs: nδzs = length(iδzs); Step 464: Randomly select the index of the next observation point iδz = ceil(rand(1)*nδzs); rand(1) represents randomly obtaining a real number within (0,1); ceil(·) represents rounding up the array within the parentheses; Step 465: Update the mounting height compensation value δZ=δz′(iδz). The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 4 is characterized in that: In step five, an adaptive optimization method for air blowing delay based on particle swarm optimization algorithm is executed to update the mounting air blowing delay; the specific process is as follows: Step 51: Initialize the particle count variable n = 1; Step 52: Determine if n is less than N. If yes, proceed to Step 53; otherwise, proceed to Step 6. Step 53: Update the position and velocity of the nth particle; the specific process is as follows: Step 531: Determine if n is equal to j. If yes, proceed to step 54; otherwise, proceed to step 532. Step 532: Update particle velocity (VB) n =w*vb n +c1*rand(1)*(po n -pb n )+c2*rand(1)*(OT b -pb n ); Step 533: Update particle position pb n =pb n +vb n ; Step 54: Update the blowing delay of the nth blowing test point to pb. n ; Step 55: Update the particle count variable n = n + 1, and return to Step 52. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 5 is characterized in that: In step six, based on the updated surface mount process parameters from steps four and five, the next circuit board is mounted, and the mounted circuit board undergoes automatic optical inspection; the specific process is as follows: Based on the updated surface mount process parameters in steps four and five, the mounting of the kth circuit board is completed; the mounted circuit board is then subjected to automatic optical inspection. Obtain the results of automated optical inspection; The results of automated optical inspection include: total number of mounting defects D 0 The offset of each patch point ∈ X ,∈ Y ,∈ θ Source of defects; The sources of defects are categorized as follows: defects caused by excessive mounting height, mounting defects caused by insufficient air blowing delay, and mounting defects caused by excessive mounting offset. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 6 is characterized in that: In step seven, the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm are updated according to the automatic optical detection results from step six; the specific process is as follows: Step 71: Based on the automatic optical inspection results, determine whether the defect is caused by excessive mounting height. If so, update the upper limit of the mounting height δZ search range and assign it the value SS. δZ (2) = δZ; if not, update the lower limit of the search range for the mounting height δZ and assign SS. δZ (1) = δZ; Step 72: Calculate the defect rate R = D 0 / D l Update the number of mounting defects D on the previous circuit board. l =D 0 ; Step 73: Update the historical mounting height array δz′=[δz′δZ]; Step 74: Update the historical defect rate array r = [r R]. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 7 is characterized in that: In step eight, the parameters of the blowing delay adaptive optimization method based on the particle swarm optimization algorithm are updated according to the automatic optical detection results in step six; the specific process is as follows: Step 81: Initialize the particle count variable n = 1; Step 82: Determine if n is less than or equal to N. If yes, proceed to step 83; otherwise, proceed to step 87. Step 83: Based on the automatic optical inspection results, determine whether the nth air blowing test point has a placement defect caused by insufficient air blowing delay or excessive placement offset. If so, set ψ n =1; otherwise, set ψ n =0; Step 84: Calculate the fitness value of the particle. n =pb n *ψ n *V+pb n ; Step 85: Determine fit n Is it less than fo? n If we update the optimal fitness value fo of particle n... n =fit n Update the individual optimal position po of particle n. n =pb n Otherwise, it will not be updated; Step 86: Update the particle count variable n = n + 1, and return to step 82; Step 87: Obtain the optimal fitness value OF and the particle number j that has reached the optimal fitness value, [OF,j] = min(fit); Step 88: Determine if OF is greater than V. If so, it means that all tested air-blowing delays will lead to defects. Set the optimal air-blowing delay to the safe value OT. b =T ba If not, then set the optimal blowing delay OT. b =pb j ; Steps 8 and 9: Reduce the blowing delay T b Search range SS Tb . The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 8 is characterized in that: In steps eight and nine, the blowing delay T is reduced. b Search range SS Tb The specific process is as follows: Step 891: Obtain the array of test point indices ns = find(ψ==1); Where ψ represents an array of all air-blowing test points to indicate whether there are mounting defects; Step 892: Update the search range T b The lower SS Tb (1) = max([pb ns SS Tb (1)]); Among them, pb ns Represented as the position of particle number ns, SS Tb (1) indicates T b The lower bound; Step 893: Determine if OF is less than V. If so, update the search range T. b Upper SS Tb (2) = pb j If not, update SS. Tb (2) = SS Tb (2)+1; Among them, SS Tb (2) indicates T b The upper boundary. The adaptive optimization method for surface mounting process parameters based on automatic optical inspection feedback as described in claim 9 is characterized in that: In step nine, based on the offset of each mounting point automatically detected by the optical system in step six, an adaptive optimization method for mounting alignment based on the moving average method is executed to update the X-axis offset compensation value, Y-axis offset compensation value, and θ-axis deflection compensation value; the specific process is as follows: Step 91: Obtain the length n of the cache array. B =length(B X ); Step 92, Determine n B If it equals W, update the accumulated X-axis offset ε. X =ε X -B X (1) Accumulated value of Y-axis offset ε Y =ε Y -B Y (1) Accumulated value of θ-axis offset ε θ =ε θ -B θ (1) Delete the first element B in the cache array. X (1) = [], B Y (1) = [], B θ (1) = []; otherwise, do not update. Among them, B X (1) indicates the first in the historical data of the X-axis mounting offset when it is not compensated, B Y (1) indicates the first in the historical data of the Y-axis mounting offset when it is not compensated, B θ (1) indicates the first in the historical data of the θ-axis mounting offset when it is not compensated, and the [] is an empty array; Step 93: Add the offset of each decoration point to the cache array: B X =[B X (∈ X +δX)]、B Y =[B Y (∈ Y +δY)]、B θ =[B θ (∈ θ +δθ)]; Among them, B X =[B X (∈ X +δX)] represents (∈ X +δX) is stored in array B X B is the last element; Y =[B Y (∈ Y +δY)] represents (∈ Y +δY) is stored in array B Y B is the last element; θ =[B θ (∈ θ +δθ)] represents (∈ θ +δθ) is stored in array B θ The middle element is the last element; Step 94: Update the accumulated X-axis offset value ε X =ε X +∈ X Y-axis offset cumulative value ε Y =ε Y +∈ Y ε, the cumulative value of the θ-axis offset θ =ε θ +∈ θ ; Step 95: Update the X-axis offset compensation value δX = ε X / n B Y-axis offset compensation value δY=ε Y / n B θ-axis deflection compensation value δZ = ε θ / n B ; Step 96: Return to Step 31.

Images