Ceramic packaging substrate surface roughness control system and method

Abstract

The invention belongs to the technical field of process control, and discloses a surface roughness control system and method for a ceramic packaging substrate; the method comprises: dividing the substrate into regions according to a preset roughness standard; measuring the substrate surface to obtain substrate elevation data; analyzing the substrate elevation data to obtain the regional surface roughness of each region; optimizing the polishing parameters of each region according to the roughness standard and the regional surface roughness of each region using a natural inspiration optimization algorithm, and marking them as optimization parameters; sorting each region according to the optimization parameters of each region; controlling the polishing equipment to adjust to the corresponding optimization parameters in turn according to the regional sequence; the invention can effectively improve the polishing efficiency and reduce resource waste; effectively respond to the needs of different regions and different substrates, and improve the surface quality and consistency of the substrate, thereby enhancing the reliability and performance of electronic devices.

Description

Technical Field

[0001] The present invention relates to the technical field of process control, and more specifically, to a surface roughness control system and method for a ceramic packaging substrate. Background Art

[0002] As electronic products develop towards miniaturization, high integration and high performance, ceramic packaging substrates play an increasingly important role in the field of electronic packaging. Ceramic materials have excellent insulation, thermal conductivity and mechanical strength, and are very suitable for the packaging of high-power and high-frequency electronic devices. However, the roughness of the surface of the ceramic substrate will have an important impact on the performance of the device. Excessive surface roughness will increase contact resistance, reduce thermal conductivity, and may cause problems such as circuit short circuits. Excessive surface roughness will affect the coating process and packaging reliability. Therefore, precise control of the surface roughness of ceramic substrates is one of the key technologies in electronic packaging manufacturing. Traditional surface roughness control mainly relies on the accumulation of process parameter experience and manual detection feedback, which makes it difficult to achieve refined and automated control.

[0003] Of course, there are also intelligent surface roughness control methods. For example, the patent with publication number CN115555923A discloses a system and method for intelligent identification and real-time control of the physical characteristics of the surface of a lead-based anode; it includes a rotating positioning platform, a fine control module, an optical recognition module, a software analysis module, and an intelligent decision-making module; the lead-based anode is clamped and fixed by the rotating positioning platform and rotated to a certain position, and the fine control module is used to realize the processing of the surface roughness and geometric structure of the anode. The optical recognition module completes the collection of the physical characteristic data of the entire anode surface, and feeds the data back to the software analysis module for precise quantitative analysis. The intelligent decision-making module controls the linkage of the control module and the recognition module in a closed-loop manner based on the software analysis results until the physical indicators such as the surface roughness and deformation of the anode are met, so as to realize the integration of positioning, control, recognition, analysis and decision-making, improve the intelligence level and processing capacity of the anode surface treatment, and reduce the generation of heavy metal pollutants from the source.

[0004] In the above technology, the surface roughness is controlled through an intelligent feedback control system; that is, real-time roughness quality feedback is given to the anode surface after grinding, the abnormal roughness areas or points on the anode surface are accurately identified and located, and the abnormal roughness areas or points are repeatedly ground, and the cycle is repeated to achieve intelligent control and optimization of the anode surface roughness; but when grinding the anode surface, the particle size and pressure of the grinding wheel are pre-set, and cannot be dynamically optimized and adjusted according to the real-time surface roughness of the anode, which may cause the effect of multiple grinding to be unable to adapt to various surface roughness problems, and it is difficult to deal with complex surface defects or occasions requiring different precision; and fixed particle size and pressure will also cause surface damage or excessive wear, affecting the final surface quality; in addition, repeated grinding will lead to lower processing efficiency, and more iterations are required to achieve the ideal surface quality, resulting in waste of materials and energy, especially when the grinding effect is not good, more time and resources are required to correct the problem.

[0005] In view of this, the present invention proposes a ceramic package substrate surface roughness control system and method to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purpose, the present invention provides the following technical solution: a method for controlling the surface roughness of a ceramic packaging substrate, comprising:

[0007] Divide the substrate into regions according to the preset roughness standard;

[0008] Measuring the substrate surface to obtain substrate elevation data;

[0009] Analyze the substrate elevation data to obtain the regional surface roughness of each area;

[0010] According to the roughness standard of each area and the surface roughness of the area, the polishing parameters of each area are optimized using a nature-inspired optimization algorithm and marked as optimized parameters;

[0011] Sort each region according to its optimization parameters;

[0012] The polishing equipment is controlled to operate with corresponding optimized parameters in sequence according to the zone order.

[0013] Further, the step of optimizing the polishing parameters of each area includes:

[0014] Step 1: Construct N parameter sets, set different numerical labels for different parameter sets, and mark them as set labels, where N is an integer greater than 1;

[0015] Step 2: Preset the population size Z and the number threshold T;

[0016] Step 3: Initialize the population. The positions of the warm-sensitive animals in the initial population are defined in a one-dimensional search space. The positions of the warm-sensitive animals correspond to the set labels one by one. The number of iterations t of the initial population is 0.

[0017] Step 4: Determine the fitness function;

[0018] Step 5: Define the temperature and food intake of each thermosensitive animal location;

[0019] Step 6: Determine whether a thermosensitive animal has entered the cave to escape the heat, and update the position of the thermosensitive animal that has entered the cave;

[0020] Step 7: The thermosensitive animals that have not entered the cave enter the foraging phase, and the positions of the thermosensitive animals are updated;

[0021] Step 8: If the number of iterations t is less than the number threshold T, return to step 5 and set the number of iterations ; If the number of iterations t is greater than or equal to the number threshold T, go to step 9;

[0022] Step 9: Calculate the fitness corresponding to the position of each warm-sensitive animal, sort all fitnesses from large to small, obtain the position of the warm-sensitive animal corresponding to the fitness ranked first, and obtain the parameter set corresponding to the corresponding set label based on the obtained position of the warm-sensitive animal.

[0023] Further, the step 1 includes: obtaining a parameter range, the parameter range including a polishing time range, a polishing pressure range, a polishing speed range, a polishing liquid concentration range, an abrasive particle size range and a temperature range; randomly selecting a value from each range within the parameter range to construct a parameter set, and constructing a total of N parameter sets, wherein each parameter set is different;

[0024] In step 3, the population is initialized , It is the Zth warm-sensitive animal;

[0025] The range of the set label and the one-dimensional search space are ; The expression for the position of each warm-sensitive animal is: ; In the formula, is the initial position of the i-th warm-sensitive animal, for A random number between .

[0026] Furthermore, in step 4, the fitness function is expressed as: ; In the formula, For fitness, is the similarity; the method for obtaining the similarity is as follows: obtain the parameter set corresponding to the set label corresponding to the position of the temperature-sensitive animal, use the roughness standard and the regional surface roughness as test data, use the parameter set and the test data as analysis data, input the analysis data into the trained surface prediction model, predict the corresponding surface roughness, and mark it as predicted roughness; subtract the roughness standard in the analysis data from the predicted roughness and take the absolute value to obtain the roughness difference, and use the inverse of the roughness difference as the similarity;

[0027] In step 5, the expression of the temperature at the location of the temperature-sensitive animal is: ; In the formula, is the temperature at the location of the i-th temperature-sensitive animal;

[0028] A food intake model of temperature-sensitive animals is pre-built, and the food intake model of temperature-sensitive animals is a normal distribution model;

[0029] The expression for food intake is: ; In the formula, is the food intake of the i-th warm-sensitive animal, is the normalization constant, exp is the exponential function, is the standard deviation of food intake, The optimal temperature is the temperature at which the food intake of thermosensitive animals is the highest.

[0030] Furthermore, in step 6, the method for determining whether the temperature-sensitive animals enter the cave to escape the heat includes:

[0031] like , then the warm-sensitive animals enter the cave to escape the heat; if , then the warm-sensitive animals did not enter the cave to escape the heat, and the warm-sensitive animals entered the foraging stage; It is an indicator of behavioral changes in warm-sensitive animals;

[0032] Methods for updating the location of thermosensitive animals entering a cave include:

[0033] The expression for the cave is: ; In the formula, For the cave location, is the position of the warm-sensitive animal with the highest fitness during the population iteration process, The position of the warm-sensitive animal with the highest fitness in the last population iteration;

[0034] Each thermosensitive animal that enters the cave is given a random number , for A random number between

[0035] like , then the calculation method for the position of the thermosensitive animals entering the cave includes:

[0036] ;

[0037] ;

[0038] In the formula, is the position of the i-th thermosensitive animal entering the cave, is the current position of the i-th warm-sensitive animal, is a decreasing curve;

[0039] like , then the calculation method for the position of the thermosensitive animals entering the cave includes:

[0040] ;

[0041] In the formula, is the position of a random warm-blooded animal in the population.

[0042] Furthermore, in step 7, the method for updating the position of the temperature-sensitive animal entering the foraging stage includes:

[0043] The expression for food position is: ; In the formula, for food location;

[0044] The expression for food size is: ; In the formula, For food size, For food factors, , is the fitness corresponding to the position of the i-th temperature-sensitive animal, is the fitness corresponding to the food position, for A random number between

[0045] like , the warm-sensing animals tear the food into pieces; the expression of the position of the food after tearing the food into pieces is: ; In the formula, The position of food after it has been torn into pieces;

[0046] Methods for calculating the location of thermosensitive animals entering the foraging phase include:

[0047] ;

[0048] In the formula, is the position of the i-th warm-sensitive animal after entering the foraging phase, for A random number between

[0049] like , then the thermosensitive animals move towards the food and eat. The calculation method of the thermosensitive animals entering the foraging stage includes:

[0050] .

[0051] Further, the polishing parameters include polishing time, polishing pressure, polishing speed, polishing liquid concentration, abrasive particle size and temperature; the polishing pressure is the pressure applied by the polishing pad to the substrate; the polishing speed is the rotation speed of the polishing pad or the substrate; the abrasive particle size is the particle size of the abrasive used in the polishing process;

[0052] The substrate elevation data is the height value of each point on the substrate surface;

[0053] The method for acquiring the elevation data of the substrate is: using a laser scanner, irradiating the laser beam on the substrate surface for scanning, marking n points in each area of ​​the substrate surface, where n is an integer greater than 1, and measuring the height values ​​of the n points.

[0054] Furthermore, the method for obtaining the regional surface roughness of each region includes:

[0055] The base plate elevation data of each area is smoothed and marked as smoothed data; the smoothing method is: the sliding window size is preset to m, ; For the height values ​​of the n points corresponding to each area, a sliding window is used to construct an analysis set in sequence. Each area corresponds to M analysis sets. ; If m is an odd number, the median of each analysis set is obtained; if m is an even number, two medians of each analysis set are obtained, and the mean of the two medians is used as the median of the corresponding analysis set; the median of each analysis set is used as smoothed data;

[0056] According to the construction order of the analysis set corresponding to each region, the difference between two adjacent medians corresponding to each region is calculated in turn and marked as median difference, where the first median and the last median are adjacent medians; the absolute values ​​of the median differences corresponding to each region are added in turn and then divided by M to obtain the regional surface roughness of each region.

[0057] Furthermore, the step of sorting each region includes:

[0058] Step S1: obtaining the unit adjustment speed of each parameter in the polishing parameters;

[0059] Step S2: multiply each parameter in the optimization parameters corresponding to each area by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameters corresponding to each area; select the adjustment time with the largest value from the adjustment times corresponding to each area as the total adjustment time of the corresponding area;

[0060] Step S3: sort the total adjustment time of each area from small to large, mark the area corresponding to the front total adjustment time as the polishing area according to the positive order, and add it to the pre-constructed sorting table; mark the area not added to the sorting table as the area to be polished;

[0061] Step S4: subtract the corresponding parameter in the optimization parameter corresponding to the polishing area from the corresponding parameter in the optimization parameter corresponding to each area to be polished and take the absolute value to obtain the adjustment amount of each parameter in the optimization parameter corresponding to each area to be polished, and multiply by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameter corresponding to each area to be polished; select the adjustment time with the largest value from the adjustment time corresponding to each area to be polished as the total adjustment time of the corresponding area to be polished;

[0062] Step S5: Mark the polishing area in the sorting table as a completed area; sort the total adjustment time of each area to be polished from small to large, mark the area to be polished corresponding to the total adjustment time at the front as a polishing area according to the positive order, and add it to the pre-constructed sorting table;

[0063] Step S6: looping steps S4 to S5 until all regions are added to the pre-built sorting table, then the loop ends and proceeds to step S7;

[0064] Step S7: sort all the regions in the sorting table according to the order in which they are added to the sorting table.

[0065] Furthermore, the method further comprises: performing parameter compensation on the optimization parameters of each region;

[0066] Obtaining parameter compensation, which includes the compensation of each parameter in the polishing parameters; adding the corresponding compensation in the parameter compensation to the optimization parameters of each area, to obtain the compensated optimization parameters of each area;

[0067] Methods for obtaining parameter compensation include:

[0068] After the last substrate polishing is completed, the substrate surface is measured again, and the obtained substrate elevation data is analyzed to obtain the regional surface roughness of each area and mark it as polishing roughness; the roughness standard of each area during the last substrate polishing is obtained and marked as a historical standard; the optimization parameters of each area during the last substrate polishing are obtained and marked as historical parameters; according to the historical standard and polishing roughness of each area, the nature-inspired optimization algorithm is used again to obtain the optimization parameters corresponding to each area and mark them as actual parameters; each parameter in the actual parameters corresponding to each area is subtracted from the corresponding parameter in the historical parameters corresponding to each area to obtain the compensation amount of each parameter in the polishing parameters corresponding to each area.

[0069] A ceramic package substrate surface roughness control system, used to implement the ceramic package substrate surface roughness control method, comprising:

[0070] A region division module divides the substrate into regions according to a preset roughness standard;

[0071] A surface measurement module is used to measure the surface of the substrate and obtain the elevation data of the substrate;

[0072] A data analysis module is used to analyze the elevation data of the substrate to obtain the regional surface roughness of each area;

[0073] The process optimization module uses a nature-inspired optimization algorithm to optimize the polishing parameters of each area according to the roughness standard and regional surface roughness of each area, and marks them as optimized parameters;

[0074] The sequential arrangement module sorts each region according to its optimization parameters;

[0075] The process adjustment module is used to control the polishing equipment to adjust to the corresponding optimized parameters according to the zone sequence.

[0076] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the method for controlling the surface roughness of a ceramic package substrate is implemented when the processor executes the computer program.

[0077] A computer-readable storage medium stores a computer program, and when the computer program is executed, the method for controlling the surface roughness of a ceramic package substrate is implemented.

[0078] The technical effects and advantages of the surface roughness control system and method of the ceramic packaging substrate of the present invention are as follows:

[0079] 1. Adopting data-driven optimization ideas, the elevation data is obtained by scanning and measuring the substrate surface, and the elevation data is analyzed to obtain the surface roughness of each area; the polishing parameters of different areas are optimized through the temperature-sensitive animal optimization algorithm, and accurate polishing of each area is achieved, so that the surface roughness of each area after polishing meets the roughness standard, thereby realizing dynamic adjustment of polishing parameters; it can effectively improve polishing efficiency, reduce resource waste, and improve the degree of automation of the polishing process; and effectively respond to the needs of different regions and different substrates, to improve the surface quality and consistency of the substrate, thereby enhancing the reliability and performance of electronic devices.

[0080] 2. By re-testing the substrate surface data after each polishing, the nature-inspired optimization algorithm is used to calculate the optimal parameters actually required for each area, and the parameter deviation is obtained by comparing with the historical parameters, and then the optimized parameters of each area are digitally compensated; dynamic identification of equipment performance changes and numerical correction guidance are given to compensate for parameter deviations caused by equipment wear and reduced accuracy; the accuracy and consistency of the polishing process are further improved, the rework rate and production costs are reduced, thereby improving product quality stability and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 Schematic diagram of a surface roughness control system for a ceramic package substrate according to Embodiment 1 of the present invention;

[0082] Figure 2 This is an algorithm flow chart for optimizing polishing parameters using a temperature-sensitive animal optimization algorithm according to Example 1 of the present invention;

[0083] Figure 3 Schematic diagram of a surface roughness control system for a ceramic package substrate according to Embodiment 2 of the present invention;

[0084] Figure 4 This is a flow chart of a method for controlling the surface roughness of a ceramic packaging substrate according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0085] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0086] Example 1

[0087] See also Figure 1 As shown, this embodiment discloses a surface roughness control system for a ceramic package substrate, including a region division module, a surface measurement module, a data analysis module, a process optimization module, a sequence arrangement module, and a process adjustment module; each module is connected by wire and / or wireless means to achieve data transmission between modules;

[0088] The area division module divides the substrate into areas according to the preset roughness standard.

[0089] The substrate is a ceramic packaging substrate; the roughness standard is preset by technical personnel in this field according to the actual standard requirements of the substrate; the functional requirements of different positions in the substrate are different, so the corresponding roughness standards are also different. For example, the pad area in the substrate needs to ensure uniform adhesion of the solder and good electrical connection, so the roughness standard is usually less than 1μm; in order to optimize the heat dissipation performance, the roughness standard of the heat dissipation area is usually less than 2μm; regional division of the substrate ensures that each part of the substrate meets its functional requirements after the subsequent polishing operation is completed, thereby improving the overall performance and reliability of the substrate.

[0090] The surface measurement module is used to measure the surface of the substrate and obtain the elevation data of the substrate.

[0091] The substrate elevation data is the height value of each point on the substrate surface, which is used to represent the microscopic undulations and unevenness of the substrate surface;

[0092] The method for acquiring the elevation data of the substrate is as follows: a laser scanner installed next to the polishing pad in the polishing equipment is used, and the focus of the laser scanner is adjusted to the polishing platform, and the polishing platform is the platform for polishing the substrate; the laser beam is irradiated on the surface of the substrate for scanning, and n points are marked in each area of ​​the substrate surface, where n is an integer greater than 1, and the height values ​​of the n points are measured; the polishing equipment is a device for polishing the surface of the substrate.

[0093] The data analysis module is used to analyze the substrate elevation data and obtain the regional surface roughness of each area.

[0094] Methods for obtaining the regional surface roughness of each region include:

[0095] The base plate elevation data of each area is smoothed and marked as smoothed data; the smoothing method is: the sliding window size is preset to m, ; For the height values ​​of the n points corresponding to each area, a sliding window is used to construct an analysis set in sequence. Each area corresponds to M analysis sets. ; If m is an odd number, the median of each analysis set is obtained; if m is an even number, two medians of each analysis set are obtained, and the mean of the two medians is used as the median of the corresponding analysis set; the median of each analysis set is used as smoothed data;

[0096] According to the construction order of the analysis set corresponding to each region, the difference between two adjacent medians corresponding to each region is calculated in turn and marked as median difference, where the first median and the last median are adjacent medians; the absolute values ​​of the median differences corresponding to each region are added in turn and then divided by M to obtain the regional surface roughness of each region.

[0097] The expression of regional surface roughness is: ;

[0098] In the formula, is the regional surface roughness, is the j-th median value, .

[0099] For example, the substrate elevation data includes 1, 2, 3, 4 and 5, and the sliding window size is 2, then the analysis set is , , , , there are 4 analysis sets in total. Since 2 is an even number, the median values ​​of each analysis set are 1.5, 2.5, 3.5, and 4.5 respectively; the difference between every two adjacent medians is 1, 1, 1, and 3 respectively; therefore, the regional surface roughness is .

[0100] It should be noted that the purpose of smoothing the substrate elevation data is to remove the larger shape errors or contour errors (such as tilt, curvature, etc.) on the substrate surface caused by errors in the measuring equipment, substrate installation or environmental factors in actual measurement, so that the calculation of surface roughness is more accurate; and it can reduce large-scale noise and interference, highlight the tiny irregularities on the substrate surface, so as to more accurately evaluate the actual quality of the surface and improve the stability and reliability of the calculation results.

[0101] The process optimization module optimizes the polishing parameters of each area using a nature-inspired optimization algorithm according to the roughness standard of each area and the regional surface roughness, and marks them as optimized parameters.

[0102] Polishing parameters include polishing time, polishing pressure, polishing speed, polishing liquid concentration, abrasive particle size and temperature; among them, polishing pressure is the pressure applied by the polishing pad to the substrate; polishing speed is the rotation speed of the polishing pad or substrate; abrasive particle size is the particle size of the abrasive used in the polishing process; different polishing parameters will affect the polishing effect and polishing efficiency, so it is necessary to optimize the polishing parameters to improve surface quality and processing efficiency, reduce costs, and ensure that each substrate meets the specification requirements and reduce quality fluctuations.

[0103] Naturally inspired optimization algorithms include genetic algorithms, clonal selection algorithms, remora optimization algorithms, thermosensitive animal optimization algorithms, etc.; this embodiment takes the thermosensitive animal optimization algorithm as an example to introduce the polishing parameter optimization steps, wherein the polishing parameter optimization steps for each area are consistent.

[0104] See also Figure 2 The algorithm flow for optimizing polishing parameters using the temperature-sensitive animal optimization algorithm includes:

[0105] Step 1: Construct N parameter sets, set different numerical labels for different parameter sets, and mark them as set labels, where N is an integer greater than 1;

[0106] Step 2: Preset the population size Z and the number threshold T;

[0107] Step 3: Initialize the population. The positions of the warm-sensitive animals in the initial population are defined in a one-dimensional search space. The positions of the warm-sensitive animals correspond to the set labels one by one. The number of iterations t of the initial population is 0.

[0108] Step 4: Determine the fitness function;

[0109] Step 5: Define the temperature and food intake of each thermosensitive animal location;

[0110] Step 6: Determine whether a thermosensitive animal has entered the cave to escape the heat, and update the position of the thermosensitive animal that has entered the cave;

[0111] Step 7: The thermosensitive animals that have not entered the cave enter the foraging phase, and the positions of the thermosensitive animals are updated;

[0112] Step 8: If the number of iterations t is less than the number threshold T, return to step 5 and set the number of iterations , that is, add one to the value of the number of iterations and then assign the number of iterations; if the number of iterations t is greater than or equal to the number threshold T, go to step 9;

[0113] Step 9: Calculate the fitness corresponding to the position of each warm-sensitive animal, sort all fitnesses from large to small, obtain the position of the warm-sensitive animal corresponding to the fitness ranked first, and obtain the parameter set corresponding to the corresponding set label based on the obtained position of the warm-sensitive animal.

[0114] The above step 1 includes: obtaining a parameter range, the parameter range includes a polishing time range, a polishing pressure range, a polishing speed range, a polishing liquid concentration range, an abrasive particle size range and a temperature range; the parameter range is obtained by a technician in this field according to the technical specifications of the polishing equipment; a value is randomly selected from each range within the parameter range to construct a parameter set, and a total of N parameter sets are constructed, wherein each parameter set is different.

[0115] In the above step 2, the population size Z is determined by technical personnel in this field. During the historical polishing parameter optimization process, under multiple groups of different test data conditions, multiple different population sizes are set for the same test data, and multiple temperature-sensitive animal optimization algorithms are performed. After the same number of iterations, the corresponding set label is obtained, and the population size with the same set label as the actual set label is used as the population size corresponding to the group of test data, and the test data includes roughness standard and regional surface roughness; the actual set label is a digital label of the parameter set that best matches the group of test data, and the actual set label is obtained by technical personnel in this field based on actual experience; and the population size corresponding to each group of test data is obtained in this way, and the average of multiple population sizes is used as the preset population size Z.

[0116] The number threshold T is determined by a technician in this field. During the historical polishing parameter optimization process, under multiple groups of different test data conditions, the same test data is subjected to multiple temperature-sensitive animal optimization algorithms to obtain multiple set labels, wherein the number of iterations of the temperature-sensitive animal optimization algorithm is different each time, and the population size is the same and is Z; the number of iterations that is closest to the actual set label is used as the number of iterations corresponding to the group of test data; and the number of iterations corresponding to each group of test data is obtained by analogy, and the average of multiple iterations is used as the number threshold T.

[0117] It should be understood that the population size Z determines the breadth of the search. A larger population size can explore more solution spaces, while the number threshold T determines the termination condition of the algorithm and can control the speed at which the algorithm converges.

[0118] In step 3 above, the population is initialized , It is the Zth warm-sensitive animal.

[0119] The range of the set label and the one-dimensional search space are ; The expression for the position of each warm-sensitive animal is: ; In the formula, is the initial position of the i-th warm-sensitive animal, for A random number between .

[0120] In the above step 4, the expression of the fitness function is: ; In the formula, For fitness, is the similarity; the method for obtaining the similarity is: obtain the parameter set corresponding to the set label corresponding to the position of the temperature-sensitive animal, use the roughness standard and the regional surface roughness as test data, use the parameter set and the test data as analysis data, input the analysis data into the trained surface prediction model, predict the corresponding surface roughness, and mark it as predicted roughness; subtract the roughness standard in the analysis data from the predicted roughness and take the absolute value to obtain the roughness difference, and use the inverse of the roughness difference as the similarity.

[0121] The training process of the surface prediction model includes:

[0122] Collect b groups of analysis data in advance, set corresponding surface roughness for each of the b groups of analysis data, where b is an integer greater than 1, and convert the analysis data and the corresponding surface roughness into a corresponding set of feature vectors; the surface roughness corresponding to the analysis data is collected by a technician in the field during the operation of a historical polishing device, and the substrate corresponding to the test data in the analysis data is polished by a polishing device whose parameters are adjusted to the parameter set in the corresponding analysis data, and then the surface roughness corresponding to each substrate is collected; and the corresponding surface roughness is sequentially set for the b groups of analysis data;

[0123] Each set of feature vectors is used as the input of the surface prediction model. The surface prediction model uses a set of predicted surface roughness corresponding to each set of analysis data as output, and uses the actual surface roughness corresponding to each set of analysis data as the prediction target. The actual surface roughness is the surface roughness corresponding to the analysis data collected in advance. Minimizing the sum of the prediction errors of all analysis data is used as the training target. The calculation formula of the prediction error is: ,in is the prediction error, K is the group number of the eigenvector corresponding to the analysis data, is the predicted surface roughness corresponding to the Kth group of analysis data, is the actual surface roughness corresponding to the Kth group of analysis data; the surface prediction model is trained until the sum of the prediction errors reaches convergence and the training is stopped.

[0124] The above-mentioned surface prediction model is specifically a deep neural network model; it includes an input layer, a hidden layer and an output layer; each hidden layer includes multiple neurons, each neuron is connected to the neurons in the next layer, and the connection contains weights, which determine the importance and influence of data transmission in the neural network; an activation function is applied to each neuron between the hidden layer and the output layer, and the activation function introduces nonlinearity, allowing the network to learn more complex patterns and features.

[0125] In step 5 above, the expression for the temperature at the location of the thermosensitive animal is: ; In the formula, is the temperature at the location of the i-th temperature-sensing animal.

[0126] A food intake model for temperature-sensitive animals is constructed in advance. The food intake model for temperature-sensitive animals is a normal distribution model used to predict the food intake of temperature-sensitive animals at different temperatures. The food intake model for temperature-sensitive animals is constructed by technicians in this field by collecting food intake data of temperature-sensitive animals at different temperatures in advance.

[0127] The expression for food intake is: ; In the formula, is the food intake of the i-th warm-sensitive animal, is a normalization constant used to adjust the feeding model of thermosensitive animals to ensure that the sum of the probabilities of all values ​​is 1. exp is an exponential function. is the standard deviation of food intake, is the optimal temperature, which is the temperature corresponding to the maximum food intake of thermosensitive animals. and optimal temperature All are obtained based on the food intake model of thermosensitive animals.

[0128] In step 6 above, the method for determining whether the thermosensitive animal enters the cave to escape the heat includes:

[0129] like , then the thermosensitive animals enter the cave to escape the heat, which means that the temperature where the thermosensitive animals are located is too high; It is an indicator of behavioral changes in temperature-sensitive animals. ;

[0130] like , it means that the warm-sensing animals did not enter the cave to escape the heat, which means that the temperature where the warm-sensing animals are located is suitable for them to eat, and the warm-sensing animals enter the foraging stage.

[0131] Methods for updating the location of thermosensitive animals entering a cave include:

[0132] The expression for the cave is: ; In the formula, For the cave location, is the position of the warm-sensitive animal with the highest fitness during the population iteration process, is the position of the warm-sensitive animal with the highest fitness in the last population iteration; and The difference is that is the position of the warm-sensitive animal with the highest fitness during multiple iterations of the population, and Only the locations of warm-sensitive animals with the highest fitness in the last population iteration.

[0133] Each thermosensitive animal that enters the cave is given a random number , for A random number between

[0134] like , indicating that the cave entered by the corresponding warm-sensitive animal has no competition from other warm-sensitive animals. The calculation method of the position of the warm-sensitive animal entering the cave includes:

[0135] ;

[0136] ;

[0137] In the formula, is the position of the i-th thermosensitive animal entering the cave, is the current position of the i-th warm-sensitive animal, It is a decreasing curve.

[0138] like , indicating that the cave entered by the corresponding warm-sensitive animal has competition from other warm-sensitive animals. The calculation method of the position of the warm-sensitive animal entering the cave includes:

[0139] ;

[0140] In the formula, is the position of a random warm-blooded animal in the population.

[0141] In the above step 7, the method for updating the position of the temperature-sensing animal entering the foraging stage includes:

[0142] When eating, warm-sensing animals will choose whether to tear the food into pieces based on the size of the food; if the food is of appropriate size, the warm-sensing animals will directly ingest the food; if the food is too large, the warm-sensing animals will tear the food into pieces before ingesting it.

[0143] The expression for food position is: ; In the formula, For food location.

[0144] The expression for food size is: ; In the formula, For food size, For food factors, , is the fitness corresponding to the position of the i-th temperature-sensitive animal, is the fitness corresponding to the food position, for A random number between .

[0145] like , indicating that the food is too big and warm-sensing animals need to tear the food into pieces; the expression of the food position after tearing the food into pieces is: ; In the formula, The position of food after it has been shredded.

[0146] Methods for calculating the location of thermosensitive animals entering the foraging phase include:

[0147] ;

[0148] In the formula, is the position of the i-th warm-sensitive animal after entering the foraging phase, for A random number between .

[0149] like , then the thermosensitive animals move directly to the food and eat. The calculation method of the position of thermosensitive animals entering the foraging stage includes:

[0150] ;

[0151] It should be noted that the reason for using the temperature-sensitive animal optimization algorithm to optimize the polishing parameters is:

[0152] Strong global search capability: The thermosensitive animal algorithm simulates the behaviors of thermosensitive animals such as foraging, summer heat avoidance, and competition in different environments, and can perform global search in complex multi-dimensional search spaces. By simulating the foraging and burrowing behaviors of thermosensitive animals, the algorithm can effectively avoid falling into local optimality, thereby increasing the probability of finding the global optimal solution.

[0153] Strong adaptability: The thermosensitive animal algorithm can adapt to different optimization problems and the characteristics of polishing equipment by adjusting the population size Z and the number threshold T; the parameter optimization process can dynamically adjust the population size and the number of iterations according to historical data to ensure the applicability and efficiency of the algorithm;

[0154] High flexibility: The temperature-sensing animal optimization algorithm can flexibly define the fitness function, temperature and food intake model during the optimization process; enabling the algorithm to perform personalized optimization according to specific polishing requirements and goals, thereby improving polishing effect and efficiency;

[0155] Fast convergence speed: By reasonably setting the fitness function and cave update mechanism, the thermosensitive animal optimization algorithm can accelerate convergence and shorten the optimization time; in practical applications, it means that the optimal polishing parameters can be found faster and production efficiency can be improved;

[0156] Taking into account local search capabilities: The thermosensitive animal optimization algorithm can dynamically adjust the behavior of thermosensitive animals during the foraging stage (such as eating directly or tearing food into pieces) to achieve refined processing of local searches; this search strategy that takes into account both local and global aspects improves the performance of the algorithm in complex multi-peak optimization problems.

[0157] The sequential arrangement module sorts each region according to its optimization parameters.

[0158] The steps to sort each region include:

[0159] Step S1: obtaining the unit adjustment speed of each parameter in the polishing parameters;

[0160] The unit adjustment speed is determined by a technician in this field, who determines the corresponding adjustment time when the polishing parameters of the first polishing equipment are adjusted multiple times during the operation of the historical polishing equipment; the adjustment amount of each parameter in the polishing parameters is divided by the corresponding adjustment time to obtain the parameter change speed; the average of the multiple parameter adjustment speeds corresponding to each parameter is used as the unit adjustment speed of the corresponding parameter;

[0161] Step S2: multiply each parameter in the optimization parameters corresponding to each area by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameters corresponding to each area; select the adjustment time with the largest value from the adjustment times corresponding to each area as the total adjustment time of the corresponding area;

[0162] Step S3: sort the total adjustment time of each area from small to large, mark the area corresponding to the front total adjustment time as the polishing area according to the positive order, and add it to the pre-constructed sorting table; mark the area not added to the sorting table as the area to be polished;

[0163] Step S4: subtract the corresponding parameter in the optimization parameter corresponding to the polishing area from the corresponding parameter in the optimization parameter corresponding to each area to be polished and take the absolute value to obtain the adjustment amount of each parameter in the optimization parameter corresponding to each area to be polished, and multiply by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameter corresponding to each area to be polished; select the adjustment time with the largest value from the adjustment time corresponding to each area to be polished as the total adjustment time of the corresponding area to be polished;

[0164] Step S5: Mark the polishing area in the sorting table as a completed area; sort the total adjustment time of each area to be polished from small to large, mark the area to be polished corresponding to the total adjustment time at the front as a polishing area according to the positive order, and add it to the pre-constructed sorting table;

[0165] Step S6: looping steps S4 to S5 until all regions are added to the pre-built sorting table, then the loop ends and proceeds to step S7;

[0166] Step S7: sort all the regions in the sorting table according to the order in which they are added to the sorting table.

[0167] It should be understood that the purpose of ranking each area is to:

[0168] Minimize total adjustment time: By calculating the total adjustment time of each area and sorting them from small to large, the area with the shortest adjustment time can be processed first, thereby reducing the equipment's idle time and adjustment waiting time, and improving the overall polishing efficiency;

[0169] Optimize resource utilization: The sorted polishing sequence can maximize the utilization of equipment and reduce the waste of resources caused by frequent adjustment of different parameters; it helps to better distribute the workload of polishing equipment and avoid unnecessary energy consumption and time waste;

[0170] Reduce error accumulation: During the polishing process, if the areas are not sorted properly, it may cause cumulative errors that affect the polishing quality of subsequent areas. Through reasonable sorting, the errors that may be introduced during parameter adjustment can be reduced to ensure that the polishing quality of each area is stable and reliable.

[0171] Improve process stability: Reasonable area sorting helps maintain the stability of the polishing process; by prioritizing polishing areas with small polishing parameter changes and short adjustment time, the polishing equipment settings can be gradually adjusted to avoid large parameter jumps during the polishing process and ensure smooth operation of the polishing process.

[0172] The process adjustment module is used to control the polishing equipment to adjust to the corresponding optimized parameters according to the zone sequence.

[0173] This embodiment adopts a data-driven optimization approach, obtains elevation data by scanning and measuring the surface of the substrate, and analyzes the elevation data to obtain the surface roughness of each area; optimizes the polishing parameters of different areas through a temperature-sensitive animal optimization algorithm, and achieves precise polishing of each area, so that the surface roughness of each area after polishing meets the roughness standard, thereby realizing dynamic adjustment of polishing parameters; can effectively improve polishing efficiency, reduce resource waste, and improve the degree of automation of the polishing process; and effectively respond to the needs of different regions and different substrates, to achieve improved substrate surface quality and consistency, thereby enhancing the reliability and performance of electronic devices.

[0174] Example 2

[0175] See also Figure 3As shown, this embodiment further improves the design on the basis of Embodiment 1. As the running time of the polishing equipment increases, the various components of the polishing equipment, such as the polishing pad, the transmission system, etc., will experience a certain degree of performance degradation, such as wear, reduced precision, etc.; resulting in that under the same set parameters, the actual performance of the polishing equipment will be weaker than that in the initial use stage, that is, the phenomenon of parameter deviation occurs; as a result, after the substrate is polished, the corresponding surface roughness does not reach the roughness standard, thereby affecting product quality and production efficiency, and increasing rework rate and production cost; therefore, this embodiment provides a surface roughness control system for a ceramic packaging substrate, which also includes a process compensation module.

[0176] The process compensation module is used to compensate the optimization parameters of each area.

[0177] Obtain parameter compensation, which includes compensation for each parameter in the polishing parameters; add the corresponding compensation in the parameter compensation to the optimization parameters of each area, and obtain the compensated optimization parameters of each area.

[0178] Methods for obtaining parameter compensation include:

[0179] After the last substrate polishing is completed, the substrate surface is measured again, and the obtained substrate elevation data is analyzed to obtain the regional surface roughness of each area and mark it as polishing roughness; the roughness standard of each area during the last substrate polishing is obtained and marked as a historical standard; the optimization parameters of each area during the last substrate polishing are obtained and marked as historical parameters; according to the historical standard and polishing roughness of each area, the nature-inspired optimization algorithm is used again to obtain the optimization parameters corresponding to each area and mark them as actual parameters; each parameter in the actual parameters corresponding to each area is subtracted from the corresponding parameter in the historical parameters corresponding to each area to obtain the compensation amount of each parameter in the polishing parameters corresponding to each area.

[0180] This embodiment re-detects the substrate surface data after each polishing is completed, uses a nature-inspired optimization algorithm to calculate the optimal parameters actually required for each area, compares them with historical parameters to obtain parameter deviations, and then digitally compensates for the optimized parameters of each area; dynamically identifies the change in equipment performance and gives numerical correction guidance to compensate for parameter deviations caused by equipment wear and reduced accuracy; further improves the accuracy and consistency of the polishing process, reduces rework rate and production costs, and thus improves product quality stability and production efficiency.

[0181] Example 3

[0182] See also Figure 4 As shown, the part not described in detail in this embodiment can be found in the description of Embodiment 1 and Embodiment 2. A method for controlling the surface roughness of a ceramic package substrate is provided, and the method includes:

[0183] Divide the substrate into regions according to the preset roughness standard;

[0184] Measuring the substrate surface to obtain substrate elevation data;

[0185] Analyze the substrate elevation data to obtain the regional surface roughness of each area;

[0186] According to the roughness standard of each area and the surface roughness of the area, the polishing parameters of each area are optimized using a nature-inspired optimization algorithm and marked as optimized parameters;

[0187] Sort each region according to its optimization parameters;

[0188] The polishing equipment is controlled to operate with corresponding optimized parameters in sequence according to the zone order.

[0189] Further, the step of optimizing the polishing parameters includes:

[0190] Step 1: Construct N parameter sets, set different numerical labels for different parameter sets, and mark them as set labels, where N is an integer greater than 1;

[0191] Step 2: Preset the population size Z and the number threshold T;

[0192] Step 3: Initialize the population. The positions of the warm-sensitive animals in the initial population are defined in a one-dimensional search space. The positions of the warm-sensitive animals correspond to the set labels one by one. The number of iterations t of the initial population is 0.

[0193] Step 4: Determine the fitness function;

[0194] Step 5: Define the temperature and food intake of each thermosensitive animal location;

[0195] Step 6: Determine whether a thermosensitive animal has entered the cave to escape the heat, and update the position of the thermosensitive animal that has entered the cave;

[0196] Step 7: The thermosensitive animals that have not entered the cave enter the foraging phase, and the positions of the thermosensitive animals are updated;

[0197] Step 8: If the number of iterations t is less than the number threshold T, return to step 5 and set the number of iterations ; If the number of iterations t is greater than or equal to the number threshold T, go to step 9;

[0198] Step 9: Calculate the fitness corresponding to the position of each warm-sensitive animal, sort all fitnesses from large to small, obtain the position of the warm-sensitive animal corresponding to the fitness ranked first, and obtain the parameter set corresponding to the corresponding set label based on the obtained position of the warm-sensitive animal.

[0199] Further, the step 1 includes: obtaining a parameter range, the parameter range including a polishing time range, a polishing pressure range, a polishing speed range, a polishing liquid concentration range, an abrasive particle size range and a temperature range; randomly selecting a value from each range within the parameter range to construct a parameter set, and constructing a total of N parameter sets, wherein each parameter set is different;

[0200] In step 3, the population is initialized , It is the Zth warm-sensitive animal;

[0201] The range of the set label and the one-dimensional search space are ; The expression for the position of each warm-sensitive animal is: ; In the formula, is the initial position of the i-th warm-sensitive animal, for A random number between .

[0202] Furthermore, in step 4, the fitness function is expressed as: ; In the formula, For fitness, is the similarity; the method for obtaining the similarity is as follows: obtain the parameter set corresponding to the set label corresponding to the position of the temperature-sensitive animal, use the roughness standard and the regional surface roughness as test data, use the parameter set and the test data as analysis data, input the analysis data into the trained surface prediction model, predict the corresponding surface roughness, and mark it as predicted roughness; subtract the roughness standard in the analysis data from the predicted roughness and take the absolute value to obtain the roughness difference, and use the inverse of the roughness difference as the similarity;

[0203] In step 5, the expression of the temperature at the location of the temperature-sensitive animal is: ; In the formula, is the temperature at the location of the i-th temperature-sensitive animal;

[0204] A food intake model of temperature-sensitive animals is pre-built, and the food intake model of temperature-sensitive animals is a normal distribution model;

[0205] The expression for food intake is: ; In the formula, is the food intake of the i-th warm-sensitive animal, is the normalization constant, exp is the exponential function, is the standard deviation of food intake, The optimal temperature is the temperature at which the food intake of thermosensitive animals is the highest.

[0206] Furthermore, in step 6, the method for determining whether the temperature-sensitive animals enter the cave to escape the heat includes:

[0207] like , then the warm-sensitive animals enter the cave to escape the heat; if , then the warm-sensitive animals did not enter the cave to escape the heat, and the warm-sensitive animals entered the foraging stage;

[0208] Methods for updating the location of thermosensitive animals entering a cave include:

[0209] The expression for the cave is: ; In the formula, For the cave location, is the position of the warm-sensitive animal with the highest fitness during the population iteration process, The position of the warm-sensitive animal with the highest fitness in the last population iteration;

[0210] Each thermosensitive animal that enters the cave is given a random number , for A random number between

[0211] like , then the calculation method for the position of the thermosensitive animals entering the cave includes:

[0212] ;

[0213] ;

[0214] In the formula, is the position of the i-th thermosensitive animal entering the cave, is the current position of the i-th warm-sensitive animal, is a decreasing curve;

[0215] like , then the calculation method for the position of the thermosensitive animals entering the cave includes:

[0216] ;

[0217] In the formula, is the position of a random warm-blooded animal in the population.

[0218] Furthermore, in step 7, the method for updating the position of the temperature-sensitive animal entering the foraging stage includes:

[0219] The expression for food position is: ; In the formula, for food location;

[0220] The expression for food size is: ; In the formula, For food size, For food factors, , is the fitness corresponding to the position of the i-th temperature-sensitive animal, is the fitness corresponding to the food position, for A random number between

[0221] like , the warm-sensing animals tear the food into pieces; the expression of the position of the food after tearing the food into pieces is: ; In the formula, The position of food after it has been torn into pieces;

[0222] Methods for calculating the location of thermosensitive animals entering the foraging phase include:

[0223] ;

[0224] In the formula, is the position of the i-th warm-sensitive animal after entering the foraging phase, for A random number between

[0225] like , then the thermosensitive animals move towards the food and eat. The calculation method of the thermosensitive animals entering the foraging stage includes:

[0226] .

[0227] Further, the polishing parameters include polishing time, polishing pressure, polishing speed, polishing liquid concentration, abrasive particle size and temperature; the polishing pressure is the pressure applied by the polishing pad to the substrate; the polishing speed is the rotation speed of the polishing pad or the substrate; the abrasive particle size is the particle size of the abrasive used in the polishing process;

[0228] The substrate elevation data is the height value of each point on the substrate surface;

[0229] The method for acquiring the elevation data of the substrate is: using a laser scanner, irradiating the laser beam on the substrate surface for scanning, marking n points in each area of ​​the substrate surface, where n is an integer greater than 1, and measuring the height values ​​of the n points.

[0230] Furthermore, the method for obtaining the regional surface roughness of each region includes:

[0231] The base plate elevation data of each area is smoothed and marked as smoothed data; the smoothing method is: the sliding window size is preset to m, ; For the height values ​​of the n points corresponding to each area, a sliding window is used to construct an analysis set in sequence. Each area corresponds to M analysis sets. ; If m is an odd number, the median of each analysis set is obtained; if m is an even number, two medians of each analysis set are obtained, and the mean of the two medians is used as the median of the corresponding analysis set; the median of each analysis set is used as smoothed data;

[0232] According to the construction order of the analysis set corresponding to each region, the difference between two adjacent medians corresponding to each region is calculated in turn and marked as median difference, where the first median and the last median are adjacent medians; the absolute values ​​of the median differences corresponding to each region are added in turn and then divided by M to obtain the regional surface roughness of each region.

[0233] Furthermore, the step of sorting each region includes:

[0234] Step S1: obtaining the unit adjustment speed of each parameter in the polishing parameters;

[0235] Step S2: multiply each parameter in the optimization parameters corresponding to each area by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameters corresponding to each area; select the adjustment time with the largest value from the adjustment times corresponding to each area as the total adjustment time of the corresponding area;

[0236] Step S3: sort the total adjustment time of each area from small to large, mark the area corresponding to the front total adjustment time as the polishing area according to the positive order, and add it to the pre-constructed sorting table; mark the area not added to the sorting table as the area to be polished;

[0237] Step S4: subtract the corresponding parameter in the optimization parameter corresponding to the polishing area from the corresponding parameter in the optimization parameter corresponding to each area to be polished and take the absolute value to obtain the adjustment amount of each parameter in the optimization parameter corresponding to each area to be polished, and multiply by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameter corresponding to each area to be polished; select the adjustment time with the largest value from the adjustment time corresponding to each area to be polished as the total adjustment time of the corresponding area to be polished;

[0238] Step S5: Mark the polishing area in the sorting table as a completed area; sort the total adjustment time of each area to be polished from small to large, mark the area to be polished corresponding to the total adjustment time at the front as a polishing area according to the positive order, and add it to the pre-constructed sorting table;

[0239] Step S6: looping steps S4 to S5 until all regions are added to the pre-built sorting table, then the loop ends and proceeds to step S7;

[0240] Step S7: sort all the regions in the sorting table according to the order in which they are added to the sorting table.

[0241] Furthermore, the method further comprises: performing parameter compensation on the optimization parameters of each region;

[0242] Obtaining parameter compensation, which includes the compensation of each parameter in the polishing parameters; adding the corresponding compensation in the parameter compensation to the optimization parameters of each area, to obtain the compensated optimization parameters of each area;

[0243] Methods for obtaining parameter compensation include:

[0244] After the last substrate polishing is completed, the substrate surface is measured again, and the obtained substrate elevation data is analyzed to obtain the regional surface roughness of each area and mark it as polishing roughness; the roughness standard of each area during the last substrate polishing is obtained and marked as a historical standard; the optimization parameters of each area during the last substrate polishing are obtained and marked as historical parameters; according to the historical standard and polishing roughness of each area, the nature-inspired optimization algorithm is used again to obtain the optimization parameters corresponding to each area and mark them as actual parameters; each parameter in the actual parameters corresponding to each area is subtracted from the corresponding parameter in the historical parameters corresponding to each area to obtain the compensation amount of each parameter in the polishing parameters corresponding to each area.

[0245] Example 4

[0246] The present application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories store computer-readable codes, and when the computer-readable codes are run by the one or more processors, the surface roughness control method of the ceramic package substrate as described above may be executed.

[0247] The method or system according to the embodiment of the present application can also be implemented with the aid of the architecture of the electronic device. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. A storage device in the electronic device, such as a ROM or a hard disk, can store the surface roughness control method of a ceramic package substrate provided in the present application. Furthermore, the electronic device may also include a user interface. Of course, the architecture of this embodiment is only exemplary. When implementing different devices, one or more components in the electronic device of this embodiment may be omitted according to actual needs.

[0248] Example 5

[0249] One embodiment of the present application discloses a computer-readable storage medium. Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by a processor, the surface roughness control method of the ceramic package substrate according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0250] In addition, according to the embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided in the present application, for example: a method for controlling the surface roughness of a ceramic package substrate. When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0251] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0252] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling the surface roughness of a ceramic packaging substrate, characterized in that: include: Divide the substrate into regions according to the preset roughness standard; Measuring the substrate surface to obtain substrate elevation data; Analyze the substrate elevation data to obtain the regional surface roughness of each area; According to the roughness standard of each area and the surface roughness of the area, the polishing parameters of each area are optimized using a nature-inspired optimization algorithm and marked as optimized parameters; Sort each region according to its optimization parameters; Control the polishing equipment to adjust to the corresponding optimized parameters according to the order of the areas; The steps to optimize the polishing parameters for each area include: Step 1: Construct N parameter sets, set different numerical labels for different parameter sets, and mark them as set labels, where N is an integer greater than 1; Step 2: Preset the population size Z and the number threshold T; Step 3: Initialize the population. The positions of the warm-sensitive animals in the initial population are defined in a one-dimensional search space. The positions of the warm-sensitive animals correspond to the set labels one by one. The number of iterations t of the initial population is 0. Step 4: Determine the fitness function; Step 5: Define the temperature and food intake of each thermosensitive animal location; Step 6: Determine whether a thermosensitive animal has entered the cave to escape the heat, and update the position of the thermosensitive animal that has entered the cave; Step 7: The thermosensitive animals that have not entered the cave enter the foraging phase, and the positions of the thermosensitive animals are updated; Step 8: If the number of iterations t is less than the number threshold T, return to step 5 and set the number of iterations ; If the number of iterations t is greater than or equal to the number threshold T, go to step 9; Step 9: Calculate the fitness corresponding to the position of each warm-sensing animal, sort all fitnesses from large to small, obtain the position of the warm-sensing animal corresponding to the fitness ranked first, and obtain the parameter set corresponding to the corresponding set label according to the obtained position of the warm-sensing animal; The step 1 comprises: obtaining a parameter range, the parameter range comprising a polishing time range, a polishing pressure range, a polishing speed range, a polishing liquid concentration range, an abrasive particle size range and a temperature range; randomly selecting a value from each range within the parameter range to construct a parameter set, and constructing a total of N parameter sets, wherein each parameter set is different; In step 3, the population is initialized , It is the Zth warm-sensitive animal; The range of the set label and the one-dimensional search space are ; The expression for the position of each warm-sensitive animal is: ; In the formula, is the initial position of the i-th warm-sensitive animal, for A random number between .

2. The method for controlling the surface roughness of a ceramic package substrate according to claim 1, characterized in that: In step 4, the fitness function is expressed as: ; In the formula, For fitness, is the similarity; the method for obtaining the similarity is as follows: obtain the parameter set corresponding to the set label corresponding to the position of the temperature-sensitive animal, use the roughness standard and the regional surface roughness as test data, use the parameter set and the test data as analysis data, input the analysis data into the trained surface prediction model, predict the corresponding surface roughness, and mark it as predicted roughness; subtract the roughness standard in the analysis data from the predicted roughness and take the absolute value to obtain the roughness difference, and use the inverse of the roughness difference as the similarity; In step 5, the expression of the temperature at the location of the temperature-sensitive animal is: ; In the formula, is the temperature at the location of the i-th temperature-sensitive animal; A food intake model of temperature-sensitive animals is pre-built, and the food intake model of temperature-sensitive animals is a normal distribution model; The expression for food intake is: ; In the formula, is the food intake of the i-th warm-sensitive animal, is the normalization constant, exp is the exponential function, is the standard deviation of food intake, The optimal temperature is the temperature at which the food intake of thermosensitive animals is the highest.

3. The method for controlling the surface roughness of a ceramic package substrate according to claim 2, characterized in that: In step 6, the method for determining whether the thermosensitive animals enter the cave to escape the heat includes: like , then the warm-sensitive animals enter the cave to escape the heat; if , then the warm-sensitive animals did not enter the cave to escape the heat, and the warm-sensitive animals entered the foraging stage; It is an indicator of behavioral changes in warm-sensitive animals; Methods for updating the location of thermosensitive animals entering a cave include: The expression for the cave is: ; In the formula, For the cave location, is the position of the warm-sensitive animal with the highest fitness during the population iteration process, The position of the warm-sensitive animal with the highest fitness in the last population iteration; Each thermosensitive animal that enters the cave is given a random number , for A random number between like , then the calculation method for the position of the thermosensitive animals entering the cave includes: ； ； In the formula, is the position of the i-th thermosensitive animal entering the cave, is the current position of the i-th warm-sensitive animal, is a decreasing curve; like , then the calculation method for the position of the thermosensitive animals entering the cave includes: ； In the formula, is the position of a random warm-blooded animal in the population.

4. The method for controlling the surface roughness of a ceramic package substrate according to claim 3, characterized in that: In step 7, the method for updating the position of the temperature-sensitive animal entering the foraging stage includes: The expression for food position is: ; In the formula, for food location; The expression for food size is: ; In the formula, For food size, For food factors, , is the fitness corresponding to the position of the i-th temperature-sensitive animal, is the fitness corresponding to the food position, for A random number between like , the warm-sensing animals tear the food into pieces; the expression of the position of the food after tearing the food into pieces is: ; In the formula, The position of food after it has been torn into pieces; Methods for calculating the location of thermosensitive animals entering the foraging phase include: ； In the formula, is the position of the i-th warm-sensitive animal after entering the foraging phase, for A random number between like , then the thermosensitive animals move towards the food and eat. The calculation method of the thermosensitive animals entering the foraging stage includes: 。 5. The method for controlling the surface roughness of a ceramic package substrate according to claim 4, characterized in that: The polishing parameters include polishing time, polishing pressure, polishing speed, polishing liquid concentration, abrasive particle size and temperature; the polishing pressure is the pressure applied by the polishing pad to the substrate; the polishing speed is the rotation speed of the polishing pad or the substrate; the abrasive particle size is the particle size of the abrasive used in the polishing process; The substrate elevation data is the height value of each point on the substrate surface; The method for acquiring the elevation data of the substrate is: using a laser scanner, irradiating the laser beam on the substrate surface for scanning, marking n points in each area of ​​the substrate surface, where n is an integer greater than 1, and measuring the height values ​​of the n points.

6. The method for controlling the surface roughness of a ceramic package substrate according to claim 5, characterized in that: The method for obtaining the regional surface roughness of each region comprises: The base plate elevation data of each area is smoothed and marked as smoothed data; the smoothing method is: the sliding window size is preset to m, ; For the height values ​​of the n points corresponding to each area, a sliding window is used to construct an analysis set in sequence. Each area corresponds to M analysis sets. ; If m is an odd number, the median of each analysis set is obtained; if m is an even number, two medians of each analysis set are obtained, and the mean of the two medians is used as the median of the corresponding analysis set; the median of each analysis set is used as smoothed data; According to the construction order of the analysis set corresponding to each region, the difference between two adjacent medians corresponding to each region is calculated in turn and marked as median difference, where the first median and the last median are adjacent medians; the absolute values ​​of the median differences corresponding to each region are added in turn and then divided by M to obtain the regional surface roughness of each region.

7. The method for controlling the surface roughness of a ceramic package substrate according to claim 6, characterized in that: The steps to sort each region include: Step S1: obtaining the unit adjustment speed of each parameter in the polishing parameters; Step S2: multiply each parameter in the optimization parameters corresponding to each area by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameters corresponding to each area; select the adjustment time with the largest value from the adjustment times corresponding to each area as the total adjustment time of the corresponding area; Step S3: sort the total adjustment time of each area from small to large, mark the area corresponding to the front total adjustment time as the polishing area according to the positive order, and add it to the pre-constructed sorting table; mark the area not added to the sorting table as the area to be polished; Step S4: subtract the corresponding parameter in the optimization parameter corresponding to the polishing area from the corresponding parameter in the optimization parameter corresponding to each area to be polished and take the absolute value to obtain the adjustment amount of each parameter in the optimization parameter corresponding to each area to be polished, and multiply by the corresponding unit adjustment speed to obtain the adjustment time of each parameter in the optimization parameter corresponding to each area to be polished; select the adjustment time with the largest value from the adjustment time corresponding to each area to be polished as the total adjustment time of the corresponding area to be polished; Step S5: Mark the polishing area in the sorting table as a completed area; sort the total adjustment time of each area to be polished from small to large, mark the area to be polished corresponding to the total adjustment time at the front as a polishing area according to the positive order, and add it to the pre-constructed sorting table; Step S6: looping steps S4 to S5 until all regions are added to the pre-built sorting table, then the loop ends and proceeds to step S7; Step S7: sort all the regions in the sorting table according to the order in which they are added to the sorting table.

8. The method for controlling the surface roughness of a ceramic package substrate according to claim 7, characterized in that: The method further comprises: performing parameter compensation on the optimization parameters of each region; Obtaining parameter compensation, which includes the compensation of each parameter in the polishing parameters; adding the corresponding compensation in the parameter compensation to the optimization parameters of each area, to obtain the compensated optimization parameters of each area; Methods for obtaining parameter compensation include: After the last substrate polishing is completed, the substrate surface is measured again, and the obtained substrate elevation data is analyzed to obtain the regional surface roughness of each area and mark it as polishing roughness; the roughness standard of each area during the last substrate polishing is obtained and marked as a historical standard; the optimization parameters of each area during the last substrate polishing are obtained and marked as historical parameters; according to the historical standard and polishing roughness of each area, the nature-inspired optimization algorithm is used again to obtain the optimization parameters corresponding to each area and mark them as actual parameters; each parameter in the actual parameters corresponding to each area is subtracted from the corresponding parameter in the historical parameters corresponding to each area to obtain the compensation amount of each parameter in the polishing parameters corresponding to each area.

9. A ceramic package substrate surface roughness control system, used to implement the ceramic package substrate surface roughness control method according to any one of claims 1 to 8, characterized in that: include: A region division module divides the substrate into regions according to a preset roughness standard; A surface measurement module is used to measure the surface of the substrate and obtain the elevation data of the substrate; A data analysis module is used to analyze the elevation data of the substrate to obtain the regional surface roughness of each area; The process optimization module uses a nature-inspired optimization algorithm to optimize the polishing parameters of each area according to the roughness standard and regional surface roughness of each area, and marks them as optimized parameters; The sequential arrangement module sorts each region according to its optimization parameters; The process adjustment module is used to control the polishing equipment to adjust to the corresponding optimized parameters according to the zone sequence.

Images