Surface mount process parameter adaptive optimization method based on automatic optical inspection feedback

Abstract

The application relates to a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback and belongs to the electrical technology and electrical engineering fields.The application aims to solve the problems that the existing surface mounting process parameter optimization method depends on manual adjustment, cannot effectively dynamically and adaptively regulate and control various parameters, cannot guarantee the regulation and control effect, and leads to the reduction of mounting precision and efficiency, the increase of mounting defects and manual repair work.The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, optimizes mounting height based on a Bayesian optimization method, optimizes air blowing delay based on a particle swarm optimization method, automatically compensates mounting deviation based on a moving average method, and completely provides the adaptive optimization method based on automatic optical detection feedback; various parameter adjustments realize the reduction of defects in the mounting process, the improvement of mounting precision and efficiency.

Description

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] Figure 2 The placement machine 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 awaits pickup by the pick-up rod. Simultaneously, a conveyor belt transports the circuit board to the designated position for placement.

[0004] exist Figure 3 Inside the placement carrier, 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. To accommodate various component types, matching nozzles must be selected; nozzle replacement is achieved through an automatic nozzle changer.

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

[0006] In conventional surface mount production, technicians can adjust the surface mount process parameters of the placement machine according to the detection results of the automatic 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 blowing delay. These parameters have a significant impact on the placement yield, placement accuracy and placement efficiency of components.

[0007] In order to analyze the causes of various placement defects, it is necessary to first understand the inherent deviations in the mechanical structure. There are machining and assembly errors in the machine itself, so it is necessary to compensate for the X-axis placement coordinates, Y-axis placement coordinates and θ-axis rotation angle of the placement point. With the continuous operation of the placement machine for a long time, dynamic changes such as thermal deformation occur, so it is necessary to make real-time online compensation, which is currently impossible for manual compensation.

[0008] On the micron scale, any assembly deviation will significantly affect the contact between the surface of the circuit board and the end of the suction rod. Figure 5 The assembly deviation angle of the suction rod is shown, where α and β represent the deviation angles of the Z-axis and θ-axis motors in three-dimensional space, respectively. These angle deviations cause the end of the suction rod to no longer remain stationary when the θ-axis motor rotates 360 degrees, but instead forms an approximately elliptical trajectory in a non-horizontal plane. This not only indicates the presence of linear and angular deviations, but also means that there is a height difference Δz between the highest and lowest points of the suction rod. Therefore, different placement height compensations are needed for different placement angles. Past placement height adjustments rely on a large number of trial and error by hand. In addition, the height of different components and the thickness of the circuit board differ, and the circuit board may be warped, making it difficult for traditional adjustment methods to adapt to these changes.

[0009] When the end of the suction rod reaches the specified placement position and height, the air blowing process begins. This action stops the suction force of the vacuum nozzle, causing the component to fall and complete placement. The duration of the air blowing process directly affects the reliability of placement. Although extending the air blowing time can improve the placement result, too long will reduce the overall circuit board placement efficiency. Conversely, if the air blowing time is too short, components that have been placed in place may be sucked up again due to residual vacuum suction, resulting in placement deviation or even placement defects. As the air blowing delay approaches the critical threshold, increasing the delay time can improve the accuracy of placement, which can be used as an indicator for adjusting the air blowing delay. Usually a longer air blowing time is set first, and then the air blowing time is gradually shortened. In the past, the air blowing delay was set manually, and this method of adjustment was rough and it was difficult to find the shortest air blowing delay that ensured placement results.

[0010] By using industrial internet of things technology, the chip mounter and the automatic optical inspection equipment can be seamlessly operated as a unified system, thereby significantly improving the mounting quality and efficiency. Ethernet for control automation technology (EtherCAT) is designed for real-time control applications in the automation field and is used to integrate the two machines. This integration can directly transfer the registration information of components in the chip mounter into the automatic optical inspection equipment, simplifying and accelerating the data preparation process. By associating and storing the mounting precision and defect type of each mounting point with its corresponding process parameters, it is expected to independently adjust the parameters of each mounting point and achieve personalized adjustment under various production factors, thereby improving the mounting precision and yield.

[0011] The main defect of the present research is that the existing surface mounting process parameter optimization method relies on manual adjustment, cannot effectively dynamically and adaptively adjust each parameter, and cannot guarantee the adjustment effect, thereby leading to reduced mounting precision and efficiency, increased mounting defects and manual repair work. SUMMARY

[0012] The purpose of the present application is to solve the problem that the existing surface mounting process parameter optimization method relies on manual adjustment, cannot effectively dynamically and adaptively adjust each parameter, and cannot guarantee the adjustment effect, thereby leading to reduced mounting precision and efficiency, increased mounting defects and manual repair work, and proposes a surface mounting process parameter adaptive optimization method based on automatic optical inspection feedback.

[0013] The specific process of the surface mounting process parameter adaptive optimization method based on automatic optical inspection feedback is as follows:

[0014] Step one, obtain the data of the circuit board to be mounted, initialize the surface mounting process parameters and the search range of the surface mounting process parameters, initialize the mounting height adaptive optimization method parameters based on the Bayesian optimization algorithm, the mounting alignment adaptive optimization method parameters based on the moving average method, and the blow delay adaptive optimization method parameters based on the particle swarm optimization algorithm; the specific process is as follows:

[0015] Step one, obtain the data file of the circuit board to be mounted, which includes the serial number, X-axis coordinate, Y-axis coordinate and θ-axis angle of each mounting point;

[0016] Step two, set the total number K of the circuit board to be mounted;

[0017] Step three, initialize the surface mounting process parameters;

[0018] The surface mounting process parameters include the X-axis offset compensation value δX, the Y-axis offset compensation value δY, the θ-axis deflection compensation value δθ, the mounting height compensation value δZ and the mounting blow delay T of each mounting point b ;

[0019] Step 1.4, initialize the search range of surface mounting process parameters, the search range of 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], the search range of T b is SS Tb = [5, 50];

[0020] SS δX is in units of μm, SS δY is in units of μm, SS δθ is in units of °, SS δZ is in units of μm, SS Tb is in units of ms;

[0021] Step 1.5, for the mounting height adaptive optimization method based on the Bayesian optimization algorithm, initialize the exploration-exploitation 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);

[0022] The OB is the total number of mounting height observation points in the Bayesian optimization algorithm, SZ is the observation point array, and linspace(SS δZ , OB) means that an array of OB elements is generated at equal intervals within the range of SS δZ ;

[0023] Step 1.6, for the mounting alignment adaptive optimization method based on the moving average method, initialize the size of the moving average window W = 20; the X-axis offset cumulative value ε X = 0, the Y-axis offset cumulative value ε Y = 0, and the θ-axis offset cumulative value ε θ = 0;

[0024] Step 1.7, for the blow delay adaptive optimization method based on the particle swarm optimization algorithm, initialize the particle number N = 5, randomly select N mounting points for blow delay test, the particle count variable n = 1, the individual learning factor of each particle c1 = 1, the social learning factor of each particle c2 = 1, and the inertia factor w = 0.5; the position pb n and the speed vb n of each particle are randomly initialized within the search space SS Tb ;

[0025] Step two, based on the initialized surface mount process parameters, complete the mounting of the first circuit board, perform automatic optical detection on the mounted circuit board, and update the mounting height adaptive optimization method parameters based on the Bayesian optimization algorithm, update the mounting alignment adaptive optimization method parameters based on the moving average method, update the X-axis offset compensation value, Y-axis offset compensation value and theta-axis deflection compensation value, and update the air blowing delay adaptive optimization method parameters based on the particle swarm optimization algorithm.

[0026] Step three, determine whether the circuit board count variable k is less than or equal to the total number K of circuit boards.

[0027] If yes, determine the automatic optical detection result, if the automatic optical detection result is a mounting defect caused by too high mounting height, execute step four, if the automatic optical detection result is a mounting defect caused by too short air blowing delay or too large mounting offset, execute step five.

[0028] If no, it means that all circuit board mounting is completed, and the production is stopped.

[0029] Step four, execute the mounting height adaptive optimization method based on the Bayesian optimization algorithm, and update the mounting height compensation value.

[0030] Step five, execute the air blowing delay adaptive optimization method based on the particle swarm optimization algorithm, and update the mounting air blowing delay.

[0031] Step six, based on the updated surface mount process parameters in steps four and five, perform the mounting of the next circuit board, and perform automatic optical detection on the mounted circuit board.

[0032] Step seven, update the parameters of the mounting height adaptive optimization method according to the automatic optical detection result in step six.

[0033] Step eight, update the parameters of the air blowing delay adaptive optimization method according to the automatic optical detection result in step six.

[0034] Step nine, execute the mounting alignment adaptive optimization method based on the moving average method according to the mounting point offset detected in step six, and update the X-axis offset compensation value, Y-axis offset compensation value and theta-axis deflection compensation value.

[0035] The beneficial effects of the present application are:

[0036] The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, proposes a complete sequential optimization framework, regulates and controls surface mounting process parameters from high to low according to importance, realizes integrated operation of a chip mounter and automatic optical detection equipment, and further improves circuit board mounting quality and mounting efficiency.In the optimization framework, the mounting height is optimized based on a Bayesian optimization method, the air blowing delay is optimized based on a particle swarm optimization method, the mounting offset is automatically compensated based on a moving average method, and the adaptive optimization method based on automatic optical detection feedback is completely given.Parameters are adjusted, so that defects in the mounting process are reduced, and mounting precision and efficiency are significantly improved.

[0037] Specifically, the application has the following two innovations:

[0038] The application associates mounting quality with surface mounting process parameters, performs accurate calculation and adaptive adjustment, realizes integrated operation of a chip mounter and automatic optical detection equipment, and improves circuit board mounting quality and efficiency.

[0039] The application proposes a complete sequential optimization framework, and in the framework, introduces an adaptive optimization method guided by automatic optical detection feedback, adjusts various parameters, and is adapted based on Bayesian optimization and particle swarm optimization methods respectively, so that defects in the mounting process are reduced, and mounting precision and efficiency are improved. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, proposes a complete sequential optimization framework, regulates and controls surface mounting process parameters from high to low according to importance, realizes integrated operation of a chip mounter and automatic optical detection equipment, and further improves circuit board mounting quality and mounting efficiency.In the optimization framework, the mounting height is optimized based on a Bayesian optimization method, the air blowing delay is optimized based on a particle swarm optimization method, the mounting offset is automatically compensated based on a moving average method, and the adaptive optimization method based on automatic optical detection feedback is completely given.Parameters are adjusted, so that defects in the mounting process are reduced, and mounting precision and efficiency are significantly improved.

[0041] Figure 2 The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, proposes a complete sequential optimization framework, regulates and controls surface mounting process parameters from high to low according to importance, realizes integrated operation of a chip mounter and automatic optical detection equipment, and further improves circuit board mounting quality and mounting efficiency.In the optimization framework, the mounting height is optimized based on a Bayesian optimization method, the air blowing delay is optimized based on a particle swarm optimization method, the mounting offset is automatically compensated based on a moving average method, and the adaptive optimization method based on automatic optical detection feedback is completely given.Parameters are adjusted, so that defects in the mounting process are reduced, and mounting precision and efficiency are significantly improved.

[0042] Figure 3 The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, proposes a complete sequential optimization framework, regulates and controls surface mounting process parameters from high to low according to importance, realizes integrated operation of a chip mounter and automatic optical detection equipment, and further improves circuit board mounting quality and mounting efficiency.In the optimization framework, the mounting height is optimized based on a Bayesian optimization method, the air blowing delay is optimized based on a particle swarm optimization method, the mounting offset is automatically compensated based on a moving average method, and the adaptive optimization method based on automatic optical detection feedback is completely given.Parameters are adjusted, so that defects in the mounting process are reduced, and mounting precision and efficiency are significantly improved.

[0043] Figure 4 The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, proposes a complete sequential optimization framework, regulates and controls surface mounting process parameters from high to low according to importance, realizes integrated operation of a chip mounter and automatic optical detection equipment, and further improves circuit board mounting quality and mounting efficiency.In the optimization framework, the mounting height is optimized based on a Bayesian optimization method, the air blowing delay is optimized based on a particle swarm optimization method, the mounting offset is automatically compensated based on a moving average method, and the adaptive optimization method based on automatic optical detection feedback is completely given.Parameters are adjusted, so that defects in the mounting process are reduced, and mounting precision and efficiency are significantly improved.

[0044] Figure 5 The application discloses a surface mounting process parameter adaptive optimization method based on automatic optical detection feedback, proposes a complete sequential optimization framework, regulates and controls surface mounting process parameters from high to low according to importance, realizes integrated operation of a chip mounter and automatic optical detection equipment, and further improves circuit board mounting quality and mounting efficiency.In the optimization framework, the mounting height is optimized based on a Bayesian optimization method, the air blowing delay is optimized based on a particle swarm optimization method, the mounting offset is automatically compensated based on a moving average method, and the adaptive optimization method based on automatic optical detection feedback is completely given.Parameters are adjusted, so that defects in the mounting process are reduced, and mounting precision and efficiency are significantly improved. DETAILED DESCRIPTION

[0045] Specific embodiment one: the specific process of the surface mounting process parameter adaptive optimization method based on automatic optical detection feedback in the embodiment is as follows:

[0046] Step one, obtain the data of the circuit board to be mounted, initialize the surface mounting process parameters and the search range of the surface mounting process parameters, initialize the mounting height adaptive optimization method parameters based on the Bayesian optimization algorithm, the mounting alignment adaptive optimization method parameters based on the moving average method, and the blow delay adaptive optimization method parameters based on the particle swarm optimization algorithm; The specific process is:

[0047] Step one, obtain the data file of the circuit board to be mounted, which includes the serial number, X-axis coordinate, Y-axis coordinate, and θ-axis angle of each mounting point (the X-axis coordinate, Y-axis coordinate, and θ-axis of the existing double-drive gantry motion platform are introduced in the lowermost part of the first page of the X-axis coordinate, Y-axis coordinate, and θ-axis) ; Such as table 1 circuit board data file;

[0048] Step two, set the total number K of the circuit board to be mounted;

[0049] Step three, initialize the surface mounting process parameters;

[0050] The surface mounting process parameters include the X-axis offset compensation value δX, the Y-axis offset compensation value δY, the θ-axis deflection compensation value δθ, the mounting height compensation value δZ, and the mounting blow delay T of each mounting point b ; Such as table 2 surface mounting process parameter file;

[0051] Step four, initialize the search range of the surface mounting process parameters, the search range of 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], and the search range of T b is SS Tb = [5, 50];

[0052] The unit of SS δX is μm, the unit of SS δY is μm, the unit of SS δθ is °, the unit of SS δZ is μm, and the unit of SS Tb is ms;

[0053] Step five, for the mounting height adaptive optimization method based on the Bayesian optimization algorithm, initialize the exploration-exploitation parameter ξ = 3, the initial mounting height compensation value δZ = 0, and initialize the mounting height observation point number OB = 30: SZ = linspace(SS δZ , OB);

[0054] Note: The parameters set in step one five will be applied to the adaptive optimization method of the mounting height based on the Bayesian optimization algorithm in step four;

[0055] OB is the total number of mounting height observation points in the Bayesian optimization algorithm, SZ is the observation point array, linspace(SS δZ , OB) means generating an array of OB elements at equal intervals in the range of SS δZ ;

[0056] Step one six, for the adaptive optimization method of mounting alignment based on the moving average method, initialize the size of the moving average window W = 20; X-axis offset cumulative value ε X = 0, Y-axis offset cumulative value ε Y = 0, θ-axis offset cumulative value ε θ = 0;

[0057] Note: The parameters set in step one six will be applied to the adaptive optimization method of mounting alignment based on the moving average method in step seven;

[0058] Step one seven, for the adaptive optimization method of air blowing delay based on the particle swarm optimization algorithm, initialize the particle number N = 5, randomly select N mounting points (referred to as "air blowing test points") for air blowing delay test, particle count variable n = 1, individual learning factor of each particle c1 = 1, social learning factor of each particle c2 = 1, inertia factor w = 0.5, randomly initialize the position pb Tb and velocity vb n of each particle in the search space SS n ;

[0059] Note: The parameters set in step one seven will be applied to the adaptive optimization method of air blowing delay based on the particle swarm optimization algorithm in step five.

[0060] Step two, based on the initialized surface mounting process parameters, complete the mounting of the first circuit board, perform automatic optical detection on the mounted circuit board, and update the adaptive optimization method parameters of the mounting height based on the Bayesian optimization algorithm, update the adaptive optimization method parameters of the mounting alignment based on the moving average method, update the X-axis offset compensation value, Y-axis offset compensation value and θ-axis deflection compensation value, and update the adaptive optimization method parameters of the air blowing delay based on the particle swarm optimization algorithm;

[0061] Step three, judge whether the circuit board count variable k is less than or equal to the total number K of circuit boards;

[0062] If yes, judge the automatic optical inspection result, if the automatic optical inspection result is caused by the mounting height being too high, execute step four, if the automatic optical inspection result is caused by the air blowing time being too short or the mounting offset being too large, execute step five;

[0063] If no, it means that all the circuit board mounting is completed, and the production is stopped;

[0064] Step four, execute the mounting height self-adaptive optimization method based on the Bayesian optimization algorithm, and update the mounting height compensation value;

[0065] Step five, execute the air blowing time self-adaptive optimization method based on the particle swarm optimization algorithm, and update the mounting air blowing time;

[0066] Step six, based on the updated surface mounting process parameters in steps four and five, perform the mounting of the next circuit board, and perform the automatic optical inspection on the mounted circuit board;

[0067] Step seven, update the parameters of the mounting height self-adaptive optimization method according to the automatic optical inspection result in step six;

[0068] Step eight, update the parameters of the air blowing time self-adaptive optimization method according to the automatic optical inspection result in step six;

[0069] Step nine, according to the offset of each mounting point detected in step six, execute the mounting alignment self-adaptive optimization method based on the moving average method, and update the X-axis offset compensation value, the Y-axis offset compensation value, and the θ-axis deflection compensation value.

[0070] Specific implementation method two: the difference between this implementation method and the specific implementation method one is that in the step two, based on the initialized surface mounting process parameters, the mounting of the first circuit board is completed, the automatic optical inspection is performed on the mounted circuit board, and the parameters of the mounting height self-adaptive optimization method based on the Bayesian optimization algorithm, the mounting alignment self-adaptive optimization method based on the moving average method, the X-axis offset compensation value, the Y-axis offset compensation value, and the θ-axis deflection compensation value, and the air blowing time self-adaptive optimization method based on the particle swarm optimization algorithm are updated according to the automatic optical inspection result;

[0071] The specific process is as follows:

[0072] Step two one, initialize the surface mounting process parameters, including: the initial mounting height compensation value δZ = 0, the X-axis offset compensation value δX = 0, the Y-axis offset compensation value δY = 0, and the θ-axis deflection compensation value δθ = 0, and the air blowing time of each air blowing test point is set to pb n , and the air blowing time of each mounting point (other mounting points are the mounting points remaining after removing the air blowing test points from all the mounting points) is set to a safety value T ba = SSTb (2); Initialize the historical mounting height array δz′=δZ;

[0073] The SS Tb (2) Delay time for mounting air blowing T b The upper limit of the search range; n∈[1:N];

[0074] 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;

[0075] 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.

[0076] 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.

[0077] 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:

[0078] 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;

[0079] 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;

[0080] Step 232: Calculate the defect rate R = D 0 / P;

[0081] Initialize the historical defect rate array r = R;

[0082] Initialize the number of mounting defects D on the previous circuit board l =D0 ;

[0083] Where P is the total number of mounting points on the circuit board;

[0084] 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:

[0085] 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)];

[0086] 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;

[0087] ∈ 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 = εX / n B , Y-axis offset compensation value δY = ε Y / n B , θ-axis deflection compensation value δZ = ε θ / n B ;

[0092] Step two five, update the parameters of the blow delay adaptive optimization method based on particle swarm optimization algorithm according to the automatic optical inspection result; the specific process is:

[0093] Step two five one, initialize the particle count variable n = 1, and set the penalty weight value V = 10 90 ;

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

[0095] Step two five two, judge whether n is less than N, if yes, execute step two five three, otherwise execute step two five seven;

[0096] Step two five three, according to the automatic optical inspection result, judge whether the n th blow test point exists due to the blow delay is too short and leads to the mounting defects or the mounting offset is too large and leads to the mounting defects, if yes, set ψ n = 1; if no, ψ n = 0;

[0097] Wherein, ψ n is a binary variable, indicating whether the n th blow test point exists mounting defects;

[0098] Step two five four, calculate the fitness value fit n = pb n * ψ n * V + pb n ;

[0099] Wherein, * is multiplication;

[0100] Note: fitness refers to the degree of good or bad of a solution in the optimization process, which is usually represented by a numerical value, that is, fitness value;

[0101] Note: the smaller the fitness value given by step two six four is, the better, and the calculation function punishes the situation of defects due to too small blow delay through the penalty weight V;

[0102] Step two five five, update the optimal fitness value fo n = fit n , update the individual best position po n = pb n of the n th particle;

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

[0104] Step 257, get the optimal fitness value OF and the particle serial number j which reaches the optimal fitness value, [OF, j] = min(fit);

[0105] The min(·) means to find the minimum value of the array in the parentheses and the element serial number of the minimum value in the array;

[0106] Step 258, judge whether the OF is greater than V, if yes, it means that all the tested blow delay will cause defects, set the optimal blow delay OT as the safety value b = T ba , if not, set the optimal blow delay OT b = pb j ;

[0107] Wherein, Pb j is the position of the jth particle.

[0108] The other steps and parameters are the same as those in the first embodiment.

[0109] The third embodiment is different from the first or second embodiment in that in step three, it is judged whether the circuit board count variable k is less than or equal to the total number K of circuit boards;

[0110] If yes, judge the automatic optical detection result, if the automatic optical detection result is caused by the mounting height being too high, execute step four, if the automatic optical detection result is caused by the blow delay being too short or the mounting offset being too large, execute step five;

[0111] If not, it means that all the circuit boards are mounted, and the production is stopped;

[0112] The specific process is as follows:

[0113] Step three, update the mounted circuit board count variable k = k + 1;

[0114] Step three, judge whether the circuit board count variable k is less than or equal to the total number K of circuit boards;

[0115] If yes, judge the automatic optical detection result, if the automatic optical detection result is caused by the mounting height being too high, execute step four, if the automatic optical detection result is caused by the blow delay being too short or the mounting offset being too large, execute step five;

[0116] If not, execute step three;

[0117] Step three three, after all the circuit board mounting is completed, stop production.

[0118] Other steps and parameters are the same as embodiment one or two.

[0119] Embodiment four: the difference between this embodiment and one of embodiments one to three is that the mounting height adaptive optimization method based on Bayesian optimization algorithm is executed in step four to update the mounting height compensation value; the specific process is:

[0120] Step four one, according to the observation result, the Gaussian process regression model gprm=GPRM(δz', r) is fitted by using the square exponential kernel to fit the mounting height history value array δz' and the defect rate history value array r;

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

[0122] The observation result refers to the defect rate r when the mounting height is δz';

[0123] The kernel function is used to describe the correlation between data points; the square exponential kernel is the most commonly used kernel function, which assumes that the data points close to each other are highly correlated, and the correlation index decays with the increase of distance;

[0124] Note: Gaussian process regression model (Gaussian Process regression Model, GPRM), Gaussian process is a powerful non-parametric machine learning method, which is often used in regression and classification tasks; it can predict new data points by using the correlation (covariance function) between data points; Gaussian process model can not only predict the mean value of observation points, but also give the standard deviation of this prediction;

[0125] Step four two, according to the Gaussian process regression model gprm fitted in step four one, the prediction mean μ(ob) and standard deviation v(ob) at each observation point SZ(ob) are predicted;

[0126] Wherein, ob∈[1:OB] is the element index of the observation point array SZ;

[0127] Step four three, calculate rb=min(r), d=(rb-μ-ξ). / v;

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

[0129] Wherein, rb is the minimum defect rate in the search defect rate history value array, r is the defect rate history value array, d is an array (an improvement amount for guiding the next decision of the Bayesian optimization algorithm, reflecting the possible improvement degree at the current position compared with the current known optimal solution; ), ξ is an exploration-exploitation parameter, which is a parameter for balancing the exploration of new solutions (exploration) and the use of known optimal solutions (exploitation);

[0130] μ is a prediction mean array, and v is a standard deviation array;

[0131] Step four, calculate the expected improvement EI = d * Φ(d) + v * φ(d);

[0132] Wherein, 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 that the calculation is the cumulative probability of the random variable being less than or equal to each point in the 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 the array d under the standard normal distribution;

[0135] Step four, five: Weight the EI in the range of the optimal value; The specific process is:

[0136] Note: Step four, five aims to improve the selection probability of the observation point in the potential optimal value region, so as to more efficiently locate the optimal mounting height;

[0137] Step four, five, one: Initialize the observation point index array obs = [1:OB];

[0138] obs = [1:OB] is an array for indicating all observation points in the search mounting height process, wherein each element is 1:2:…:OB, ob is an index variable of the mounting height observation point, and ob ∈ [1:OB] is a numerical value, taking the value range of 1:2:…:OB;

[0139] Step four, five, two: Get the minimum index ob1 = find (SZ < SS δZ (1)) and the maximum index ob2 = find (SZ > SS δZ (2)) outside the range of the optimal value;

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

[0141] Wherein, 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 four six four, randomly select the index of the next observation point iδz = ceil(rand(l)*nδzs);

[0156] The rand(l) represents a real number randomly obtained in (0, 1); ceil(·) represents the upward rounding of the array in the parentheses;

[0157] Step four six five, update the mounting height compensation value δz = δz'(iδz).

[0158] The other steps and parameters are the same as one of the first to third embodiments.

[0159] Embodiment five: the difference between this embodiment and one of the first to fourth embodiments is that the step five performs a blowing delay adaptive optimization method based on a particle swarm optimization algorithm to update the mounting blowing delay; the specific process is:

[0160] Step five one, initialize the particle count variable n = 1;

[0161] Step five two, judge whether n is less than N, if yes, execute step five three, otherwise execute step six;

[0162] Step five three, update the position and speed of the nth particle; the specific process is:

[0163] Step five three one, judge whether n is equal to j, if yes, execute step five four, otherwise execute step five three two;

[0164] Note: Considering the uncertainty in the mounting process, the optimal blowing delay of each round remains unchanged in order to more fully and thoroughly evaluate the mounting effect of the blowing delay;

[0165] Step five three two, update the particle speed vb n = w*vb n + c1*rand(l)*(po n -pb n )+c2*rand(l)*(OT b -pb n );

[0166] Step five three three, update the particle position pb n = pb n + vb n ;

[0167] Step five four, update the blowing delay of the nth blowing test point to pb n ;

[0168] Step five five, update the particle count variable n = n + 1, and return to step five two.

[0169] Other steps and parameters are the same as one of the first to fourth embodiments.

[0170] Sixth embodiment: The difference between this embodiment and the first to fifth embodiments is that: in the step six, the surface mounting process parameters updated based on the step four and the step five are used to mount the next circuit board, and the mounted circuit board is subjected to automatic optical inspection; the specific process is as follows:

[0171] The surface mounting process parameters updated based on the step four and the step five are used to complete the mounting of the kth circuit board, and the mounted circuit board is subjected to automatic optical inspection; the result of the automatic optical inspection is obtained.

[0172] The result of the automatic optical inspection includes: the total number of mounting defects D 0 , the offset of each mounting point ∈ X , ∈ Y , ∈ θ , and the defect source.

[0173] The defect source is divided into: defects caused by too high mounting height, mounting defects caused by too short air blowing delay, and mounting defects caused by too large mounting offset.

[0174] Other steps and parameters are the same as one of the first to fifth embodiments.

[0175] Seventh embodiment: The difference between this embodiment and the first to sixth embodiments is that: in the step seven, the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm are updated according to the automatic optical inspection result of the step six; the specific process is as follows:

[0176] Step seven one, according to the automatic optical inspection result, it is judged whether the defect is caused by too high mounting height, if yes, the upper limit of the mounting height δZ search range is updated, and the value SS δZ (2) = δZ is assigned; if not, the lower limit of the mounting height δZ search range is updated, and the value SS δZ (1) = δZ is assigned.

[0177] Step seven two, the defect rate R = D 0 / D l is calculated, the number of mounting defects D l of the last circuit board is updated, and the value D 0 is assigned.

[0178] Step seven three, the mounting height history value array δz' = [δz' δZ] is updated.

[0179] Step seven four, the defect rate history value array r = [r R] is updated.

[0180] Other steps and parameters are the same as one of the first to sixth embodiments.

[0181] Eighth Embodiment: Different from the first to sixth embodiments, the eighth embodiment is characterized in that: the step eight is to update the parameters of the air blowing delay adaptive optimization method based on the particle swarm optimization algorithm according to the automatic optical inspection result of the step six; and the specific process is as follows:

[0182] Step eight one, initialize the particle count variable n = 1;

[0183] Step eight two, judge whether n is less than or equal to N, if yes, execute step eight three, otherwise execute step eight seven;

[0184] Step eight three, according to the automatic optical inspection result, judge whether the n th air blowing test point exists the mounting defect caused by the air blowing delay being too short or the mounting defect caused by the mounting deviation being too large, if yes, set ψ n = 1; if no, set ψ n = 0;

[0185] Step eight four, calculate the fitness value fit n = pb n * ψ n * V + pb n ;

[0186] Step eight five, judge whether fit n is less than fo n , if yes, update the optimal fitness value fo n = fit n of the n th particle, and update the individual best position po n = pb n of the n th particle, otherwise do not update;

[0187] Step eight six, update the particle count variable n = n + 1, and return to step eight two;

[0188] Step eight seven, obtain the optimal fitness value OF and the particle serial number j that reaches the optimal fitness value, [OF, j] = min(fit);

[0189] Step eight eight, judge whether OF is greater than V, if yes, it means that all the tested air blowing delays will cause defects, set the optimal air blowing delay OT b = T ba , if no, set the optimal air blowing delay OT b = pb j ;

[0190] Step eight nine, narrow the search range SS b of the air blowing delay T Tb .

[0191] The other steps and parameters are the same as one of the first to eighth embodiments.

[0192] The ninth embodiment is different from the first to eighth embodiments in that the search range T b of the step eight-nine is reduced Tb ; the specific process is as follows:

[0193] Step eight-nine one, obtaining the test point index array ns = find (ψ == 1) of which the blow time is too short.

[0194] Wherein, ψ represents the array of whether there is a mounting defect for all blow test points.

[0195] Step eight-nine two, updating the lower limit SS b of the search range T Tb (1) = max ([pb ns SS Tb (1)]).

[0196] Wherein, pb ns represents the position of the ns particle, and SS Tb (1) represents the lower limit of T b .

[0197] Step one-four gives SS Tb , which is an array of two values, and the first element of which is updated, which represents the lower limit of the blow time search.

[0198] Step eight-nine three, judging whether OF is less than V, if yes, updating the upper limit SS b of the search range T Tb (2) = pb j ; if not, updating SS Tb (2) = SS Tb (2) + 1.

[0199] Wherein, SS Tb (2) represents the upper limit of T b .

[0200] Step one-four gives SS Tb , which is an array of two values, and the second element of which is updated, which represents the upper limit of the blow time search.

[0201] The other steps and parameters are the same as one of the first to eighth embodiments.

[0202] Specific implementation ten: the difference between this implementation and specific implementations one to eight is that: the mounting point offset amount detected by the automatic optical detection in step six is used to perform a mounting alignment adaptive optimization method based on a moving average method to update the X-axis offset compensation value, the Y-axis offset compensation value, and the θ-axis deflection compensation value; the specific process is as follows:

[0203] Step nine one, get the length n of the cache array B = length(B X );

[0204] Step nine two, judge whether n B is equal to W, if yes, update the X-axis offset cumulative value ε X = ε X -B X (1), the Y-axis offset cumulative value ε Y = ε Y -B Y (1), and the θ-axis offset cumulative value ε θ = ε θ -B θ (1), delete the first element B X (1) = [], B Y (1) = [], B θ (1) = []; otherwise, do not update;

[0205] wherein, B X (1) represents the first element in the X-axis mounting offset amount history data before compensation, B Y (1) represents the first element in the Y-axis mounting offset amount history data before compensation, and B θ (1) represents the first element in the θ-axis mounting offset amount history data before compensation, and the [] is an empty array. Assigning some elements in the array to [] means clearing these elements.

[0206] Step nine three, increase the offset amount of each mounting point to the cache array: B X = [B X (∈ X + δX)], B Y = [B Y (∈ Y + δY)], and B θ = [B θ (∈ θ + δθ)];

[0207] wherein, B X = [B X (∈ X + δX)] represents storing (∈ X + δX) into the array B XB Y = [B Y (∈ Y + δY) represents storing (∈ Y + δY) in the array B Y as the last element; B θ = [B θ (∈ θ + δθ) represents storing (∈ θ + δθ) in the array B θ as the last element;

[0208] Step nine four, updating the X-axis offset accumulated value ε X = ε X + ∈ X , the Y-axis offset accumulated value ε Y = ε Y + ∈ Y , the θ-axis offset accumulated value ε θ = ε θ + ∈ θ ;

[0209] Step nine five, updating the X-axis offset compensation value δX = ε X / n B , the Y-axis offset compensation value δY = ε Y / n B , the θ-axis deflection compensation value δZ = ε θ / n B ;

[0210] Step nine six: returning to step three one.

[0211] The other steps and parameters are the same as one of the specific embodiments one to nine.

[0212] The beneficial effects of the present application are verified by the following embodiments:

[0213] Embodiment one:

[0214] The present embodiment is based on the surface mount process parameter adaptive optimization method of automatic optical detection feedback, which is prepared according to the following steps:

[0215] The example considers the process of surface mount process parameter optimization of a chip mounter and automatic optical detection equipment, and the components on the circuit board for mounting test adopt 0201 packaging. The size of this packaging component is the smallest, and when the chip mounter runs at the highest speed, it is more prone to misplacement and defects. Ten mounting experiments are performed, and the average value is calculated to evaluate the optimization effect.

[0216] The surface mount process parameter adaptive optimization based on automatic optical detection feedback is used in the application, and experimental results show that the method of the patent adjusts the mounting height to eliminate defects for an average of 5.7 times in the process, and 88 defects are caused during this period; the traditional manual adjustment method adjusts the mounting height to eliminate defects for an average of 6.5 times in the process, and 192 defects are caused during this period; in the mounting precision improvement method, the Cpk value (the larger the better) is improved from 0.5 to 1.87 by using the method of the application; under the premise of not causing the decline of mounting performance, the mounting air blowing delay time is shortened from 17.5 ms to 15.9 ms by using the method of the application.

[0217] The application also has other various embodiments, and those skilled in the art can make various corresponding changes and modifications according to the application without departing from the spirit and essence of the application, and these corresponding changes and modifications should all belong to the protection scope of the claims attached to the application.

[0218] Table 1 circuit board data file

[0219] Serial number X-axis coordinate (mm) Y-axis coordinate (mm) θ-axis angle (°) 1 12 12 0 2 14.5 12 90 3 17.75 12 180 4 21.75 12 270 … … … …

[0220] Table 2 surface mount process parameters

[0221]

[0222] The application also has other various embodiments, and those skilled in the art can make various corresponding changes and modifications according to the application without departing from the spirit and essence of the application, and these corresponding changes and modifications should all belong to the protection scope of the claims attached to the application.

Claims

1. A surface mount process parameter adaptive optimization method based on automatic optical inspection feedback, characterized in that: The method specifically comprises the following steps: Step one, obtaining the data of the circuit board to be mounted, initializing the surface mounting process parameters and the search range of the surface mounting process parameters, initializing the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on the moving average method, and the parameters of the air blowing delay adaptive optimization method based on the particle swarm optimization algorithm; the specific process is as follows: Step one, obtaining the data file of the circuit board to be mounted, which includes the serial number, X-axis coordinate, Y-axis coordinate and θ-axis angle of each mounting point; Step two, setting the total number K of the circuit boards to be mounted; Step three, initializing the surface mounting process parameters; The surface mounting process parameters include 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 time T of each mounting point b ; Step one four, initialize the search range of surface mount process parameters, the search range of 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], the search range of T b is SS Tb = [5, 50]; SS δX in μm, SS δY in μm, SS δθ in °, SS δZ in μm, SS Tb in ms; Step one five, for the mounting height adaptive optimization method based on Bayesian optimization algorithm, initialize the exploration-exploitation parameter ξ = 3, the initial mounting height compensation value δZ = 0; initialize the mounting height observation point number OB = 30: SZ = linspace (SS δZ , OB); The OB is the total number of mounting height observation points in the Bayesian optimization algorithm, SZ is an observation point array, linspace(SS δZ , OB) means generating an array of OB elements at equal intervals within the range of SS δZ . Step one six, for the mobile average method based on the adaptive optimization method of assembly alignment, the size of the moving average window W = 20; X axis offset cumulative value ε X = 0, Y axis offset cumulative value ε Y = 0, θ axis offset cumulative value ε θ = 0; Step one seven, for the blowing delay adaptive optimization method based on particle swarm optimization algorithm, initialize the particle number N=5, randomly select N points for blowing delay test, particle count variable n=1, individual learning factor of each particle c1=1, social learning factor of each particle c2=1, inertia factor w=0.5, randomly initialize the position pb Tb and speed vb n of each particle in the search space SS n ; Step two, based on the initialized surface mounting process parameters, completing the mounting of the first circuit board, performing automatic optical detection on the mounted circuit board, and updating the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on the 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 the particle swarm optimization algorithm according to the automatic optical detection result; Step three, determining whether the circuit board count variable k is less than or equal to the total number K of the circuit boards; If yes, determining the automatic optical detection result, and if the automatic optical detection result is a mounting defect caused by excessively high mounting height, executing step four, and if the automatic optical detection result is a mounting defect caused by excessively short air blowing delay or excessively large mounting offset, executing step five; If no, it is indicated that the mounting of all the circuit boards is completed, and the production is stopped; Step four, executing the mounting height adaptive optimization method based on the Bayesian optimization algorithm, and updating the mounting height compensation value; Step five, executing the air blowing delay adaptive optimization method based on the particle swarm optimization algorithm, and updating the mounting air blowing delay; Step six, based on the updated surface mounting process parameters in steps four and five, mounting the next circuit board, and performing automatic optical detection on the mounted circuit board; Step seven, updating the parameters of the mounting height adaptive optimization method according to the automatic optical detection result in step six; Step eight, updating the parameters of the air blowing delay adaptive optimization method according to the automatic optical detection result in step six; Step nine, executing the mounting alignment adaptive optimization method based on the moving average method according to the mounting point offset detected in step six, and updating the X-axis offset compensation value, the Y-axis offset compensation value and the θ-axis deflection compensation value.

2. The method of claim 1, wherein the method further comprises: In step two, based on the initialized surface mounting process parameters, the mounting of the first circuit board is completed, the mounted circuit board is subjected to automatic optical detection, and the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm, the parameters of the mounting alignment adaptive optimization method based on the 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 the particle swarm optimization algorithm are updated according to the automatic optical detection result; the specific process is as follows: Step two, initialize the surface mount process parameters, including: initial mounting height compensation value δZ = 0, initialize X-axis offset compensation value δX = 0, Y-axis offset compensation value δY = 0, θ-axis deflection compensation value δθ = 0, set the air blowing delay of each air blowing test point to pb n , set the air blowing delay of other mounting points to the safety value T ba = SS Tb (2); initialize the mounting height history value array δz' = δZ; The SS Tb (2) is attached to the delay T b The upper limit of the search range; n ∈ [1: N]; Step two two, based on the initialization of the surface mount process parameters, complete the first circuit board mounting; set the mounted circuit board count variable k = 1, the automatic optical inspection of the mounted circuit board, get the automatic optical inspection results; The result of the automatic optical inspection includes the offset of each mounting point X , the offset of each mounting point Y , the offset of each mounting point θ , the total number of mounting defects D 0 , and the defect source. The defect source is classified into the defect caused by the excessive mounting height, the mounting defect caused by the short air blowing delay, and the mounting defect caused by the excessive mounting offset. Step two three, according to the automatic optical inspection results update based on the Bayesian optimization algorithm of mounting height adaptive optimization method parameters; the specific process is: Step two three one, according to the automatic optical detection result, judge whether the defect is caused by the too high mounting height, if yes, then update the upper limit of the mounting height δZ search range, assign SS δZ (2) = δZ; if not, then update the lower limit of the mounting height δZ search range, assign SS δZ (1) = δZ; where SS δZ (2) is the upper limit of the search range for the mounting height δZ, and SS δZ (1) is the lower limit of the search range for the mounting height δZ. δZ (1) is the lower limit of the search range for the mounting height δZ. δZ (2) is the upper limit of the search range for the mounting height δZ. Step two three two, calculate the defect rate R = D 0 / P; Initialize the defect rate history value array r = R; Initialize the number of placement defects D of the previous board l = D 0 ; Wherein, P is the total number of mounting points on the circuit board; Step two four, according to the automatic optical inspection results update based on the moving average method of mounting alignment adaptive optimization method parameters, update the X axis offset compensation value, Y axis offset compensation value and theta axis deflection compensation value; the specific process is: Step two four one, increase the offset of each mounting point to the buffer array: B X = [B X (∈ X + δX)], B Y = [B Y (∈ Y + δY)], B θ = [B θ (∈ θ + δθ)]; wherein, δX represents an X-axis offset compensation value, δY represents a Y-axis offset compensation value, and δθ represents a θ-axis deflection compensation value; ∈ X , ∈ Y , ∈ θ represents an offset of each mounting point; B X represents a cache array storing X-axis mounting offset history data without compensation, B Y represents a cache array storing Y-axis mounting offset history data without compensation, B θ represents a cache array storing θ-axis mounting offset history data without compensation; Step two four two, update X axis offset accumulated value ε X = ε X + δx X + δx, Y axis offset accumulated value ε Y = ε Y + δy Y + δy, θ axis offset accumulated value ε θ = ε θ + δθ θ + δθ Step two four three, get the length of the cache array n B = length(B X ); Wherein, length() indicates the number of elements in the array in the parentheses; Step two four four, update 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 two five, according to the automatic optical inspection results update based on the particle swarm optimization algorithm of blowing delay adaptive optimization method parameters; the specific process is: Step two five one, initialize particle count variable n = 1, set penalty weight value V = 10 90 ; Step two five two, judge whether n is less than N, if yes, execute step two five three, otherwise execute step two five seven; Step two 253, according to the automatic optical inspection result, judging whether the n th blowing test point exists the mounting defect caused by the too short blowing delay or the mounting defect caused by the too large mounting deviation, if yes, setting ψ n = 1; if not, setting ψ n = 0; wherein ψ n is a binary variable, indicating whether a placement defect exists at the nth blow test point; Step two 254, calculate fitness value fit of particle n = pb n * ψ n * V + pb n ; Wherein, * is the multiplication sign; Step two 255, update the optimal fitness value fo of the n-th particle n = fit n , update the individual best position po of the n-th particle n = pb n ; Step two five six, update the particle count variable n = n + 1, return to step two five two; Step two five seven, get the optimal fitness value OF, and the particle serial number j that reaches the optimal fitness value, [OF, j] = min(fit); The min(·) refers to the minimum value of the array in the parentheses and the element serial number of the minimum value in the array; Step two five eight, judge if OF is greater than V, if yes, it means that all the tested blow delay will cause defects, set the optimal blow delay OT as the safety value b = T ba , if not, set the optimal blow delay OT b = pb j ; where Pb j is the position of the jth particle.

3. The method of claim 2, wherein the method further comprises: The step three is to judge whether the circuit board count variable k is less than or equal to the total number of circuit boards K; If yes, judge the automatic optical inspection result, if the automatic optical inspection result is due to the mounting height too high caused by mounting defects, execute step four, if the automatic optical inspection result is due to the blowing delay too short or due to the mounting offset too large caused by mounting defects, execute step five; If not, it means that all the circuit board mounting is completed, stop production; The specific process is: Step three one, update the mounting circuit board count variable k = k + 1; Step three two, judge whether the circuit board count variable k is less than or equal to the total number of circuit boards K; If yes, judge the automatic optical inspection result, if the automatic optical inspection result is due to the mounting height too high caused by mounting defects, execute step four, if the automatic optical inspection result is due to the blowing delay too short or due to the mounting offset too large caused by mounting defects, execute step five; If not, execute step three three; Step three three, all the circuit board mounting is completed, stop production.

4. The method of claim 3, wherein the method further comprises: The step four is to execute the mounting height adaptive optimization method based on the Bayesian optimization algorithm, update the mounting height compensation value; the specific process is: Step four one, according to the observation results, use the square exponential kernel to fit the mounting height history value array δz' and the defect rate history value array r to get the Gaussian process regression model gprm = GPRM(δz', r); Wherein, GPRM is the Gaussian process regression model, gprm is the Gaussian process regression model; The observation result refers to the defect rate r when the mounting height is δz'; Step four two, according to the model gprm of Gaussian process regression obtained in step four one, predict the prediction mean μ (ob) and the standard deviation v (ob) at each observation point SZ (ob) ; Wherein, ob∈[1:OB] is the element index in the observation point array SZ; Step four three, calculate rb=min(r), d=(rb-μ-ξ). / v; The. / 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; Wherein, rb is the smallest defect rate in the search defect rate history value array, r is the defect rate history value array, d is an array, and ξ is the exploration-exploitation parameter; μ is the prediction mean array, and v is the standard deviation array; Step four four, calculate the expected improvement EI=d*Φ(d)+v*φ(d); Wherein, 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; Step four five: weight the EI in the optimal value range; The specific process is: Step four five one: initialize the observation point index array obs=[1:OB]; Step four five two: get the minimum index ob1 = find (SZ < SS δZ (1)) and the maximum index ob2 = find (SZ > SS δZ (2)) outside the range of the optimal value. wherein SZ is an array of observation points, SS δZ (2) is an upper limit of the search range for the mounting height δZ, SS δZ (1) is a lower limit of the search range for the mounting height δZ; and δZ (2) is an upper limit of the search range for the mounting height δZ, SS δZ (1) is a lower limit of the search range for the mounting height δZ; and The find(·) means to obtain the element sequence number in the array after the bracket that meets the condition in the bracket; Step four five three: delete the index outside the optimal value range obs([ob1:ob2])=[]; The [] is an empty array; Step four five four: weight the EI in the optimal value range, EI(obs)=EI(obs)*V; Step four six, in the observation point corresponding to the maximum expected improvement, select the next mounting height value δZ; The specific process is: Step four six one, obtain the maximum value mei=max(EI) of the expected improvement; The max(·) means to obtain the maximum in the array in the bracket; Step four six two, obtain the index array iδzs of all elements equal to the maximum value mei of the expected improvement; Step four six three, obtain the number nδzs of elements in the array iδzs; Step four six four, randomly select the index iδz of the next observation point iδz=ceil(rand(1)*nδzs); The rand(1) means to randomly obtain a real number in (0,1); The ceil(·) means to round up the array in the bracket; Step four six five, update the mounting height compensation value δZ=δz′(iδz).

5. The method of claim 4, wherein the method further comprises: The blowing delay adaptive optimization method based on particle swarm optimization algorithm is executed in step five to update the mounting blowing delay; The specific process is: Step five one, initialize the particle count variable n=1; Step five two, judge whether n is less than N, if yes, execute step five three, otherwise execute step six; Step five three, update the position and velocity of the nth particle; The specific process is: Step five three one, judge whether n is equal to j, if yes, execute step five four, otherwise execute step five three two; Step five three two, update particle velocity vb n = w * vb n + c1 * rand(1) * (po n - pb n ) + c2 * rand(1) * (OT b - pb n ); Step five three, update particle position pb n = pb n + vb n ; Step five Four, update the blow delay of the nth blow test point to pb n ; Step five five, update the particle count variable n=n+1, and return to step five two.

6. The method of claim 5, wherein the method further comprises: The step six is based on the updated surface mounting process parameters in step four and step five, mounting the next circuit board, and performing automatic optical inspection on the mounted circuit board; the specific process is: Based on the updated surface mounting process parameters in step four and step five, complete the mounting of the kth circuit board; perform automatic optical inspection on the mounted circuit board; Get the result of automatic optical inspection; The result of the automatic optical inspection includes: total number of mounting defects D 0 , offset of each mounting point ∈ X , ∈ Y , ∈ θ , defect source The defect sources are: defects caused by too high mounting height, mounting defects caused by too short air blowing delay, and mounting defects caused by too large mounting offset.

7. The method of claim 6, wherein the method further comprises: The step seven is to update the parameters of the mounting height adaptive optimization method based on the Bayesian optimization algorithm according to the automatic optical inspection result in step six; the specific process is: Step seven one, according to the automatic optical detection result, judge whether the defect is caused by the too high mounting height, if yes, then update the upper limit of the mounting height δZ search range, assign SS δZ (2) = δZ; if not, update the lower limit of the mounting height δZ search range, assign SS δZ (1) = δZ; Step seven two, calculate the defect rate R = D 0 / D l ; update the number of mounting defects D of the previous circuit board l = D 0 ; Step seven three, update the mounting height history value array δz'= [δz' δZ]; Step seven four, update the defect rate history value array r=[r R].

8. The method of claim 7, wherein the method further comprises: The step eight is to update the parameters of the air blowing delay adaptive optimization method based on the particle swarm optimization algorithm according to the automatic optical inspection result in step six; the specific process is: Step eight one, initialize the particle count variable n=1; Step eight two, judge whether n is less than or equal to N, if yes, execute step eight three, otherwise execute step eight seven; Step eight three, according to the automatic optical detection result, judging whether the n th blowing test point exists the mounting defect caused by the too short blowing delay or the mounting defect caused by the too large mounting deviation, if yes, setting ψ n = 1; if not, setting ψ n = 0; Step eight four, calculate the fitness value fit of the particle n = pb n * ψ n * V + pb n ; Step eight five, judge fit n if less than fo n , if yes, update the optimal fitness value fo n = fit n of the n-th particle, update the individual best position po n = pb n of the n-th particle, otherwise do not update; Step eight six, update the particle count variable n=n+1, return to step eight two; Step eight seven, get the optimal fitness value OF, and the particle serial number j that reaches the optimal fitness value, [OF, j]=min(fit); Step eight, determine if OF is greater than V, if yes, it means that all the tested blow delay will cause defects, set the optimal blow delay OT as the safe value b = T ba , if no, set the optimal blow delay OT b = pb j ; Step eight nine, reduce the blow delay T b The search range SS Tb .

9. The method of claim 8, wherein the method further comprises: The step eight nine in the reduced blowing delay T b The search range SS Tb The specific process is: Step eight nine one, get the test point index array ns=find(ψ==1) of the air blowing delay that is too short; Wherein, ψ represents the array of whether there is mounting defect in all air blowing test points; Step eight nine two, update search range T b lower bound of SS Tb (1) = max([pb ns SS Tb (1)]); where pb ns denotes the position of the ns-th particle, SS Tb (1) denotes the lower bound of T b ; Step eight nine three, judge if OF is less than V, if yes, update search range T b upper bound of SS Tb (2) = pb j ; if not, update SS Tb (2) = SS Tb (2) + 1; where SS Tb (2) represents the upper bound of T b .

10. The method of claim 9, wherein the method further comprises: The step nine is to execute the mounting alignment adaptive optimization method based on the moving average method according to the mounting point offset of each mounting point detected in step six, and update the X axis offset compensation value, Y axis offset compensation value and θ axis deflection compensation value; the specific process is: Step nine one, get the length of the cache array n B = length(B X ); Step nine two, judge n B whether equal to W, if yes update X axis offset cumulative value ε X = ε X -B X (1), Y axis offset cumulative value ε Y = ε Y -B Y (1), θ axis offset cumulative value 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; where B X (1) indicates the first of the X-axis mounting offset history data stored without compensation, B Y (1) indicates the first of the Y-axis mounting offset history data stored without compensation, B θ (1) indicates the first of the θ-axis mounting offset history data stored without compensation, and the [] is an empty array. Step nine three, increase the offset of each mounting point to the cache array: B X = [B X (∈ X + δX)], B Y = [B Y (∈ Y + δY)], B θ = [B θ (∈ θ + δθ)]; where B X = [B X (∈ X + δx)] means that (∈ X + δx) is stored in the array B X as the last element; B Y = [B Y (∈ Y + δy)] means that (∈ Y + δy) is stored in the array B Y as the last element; B θ = [B θ (∈ θ + δθ)] means that (∈ θ + δθ) is stored in the array B θ as the last element; Step nine four, update the X-axis offset accumulated value ε X = ε X + ∈ X , the Y-axis offset accumulated value ε Y = ε Y + ∈ Y , the θ-axis offset accumulated value ε θ = ε θ + ∈ θ ; Step nine 95, update 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 nine six: return to step three one.

Images