RFID wrist strap production monitoring method and system based on data analysis

Abstract

The invention discloses an RFID wrist strap production monitoring method and system based on data analysis, and relates to the technical field of product production monitoring, and the method comprises the steps: generating an expansion effect factor of a base material according to environment parameters of a processing workshop, monitoring the working state and motion characteristics of surface mounting equipment in real time, carrying out the dynamic modeling of the surface mounting equipment through employing a Markov model, and carrying out the monitoring of the surface mounting equipment. According to the method, the change of working parameters of the mounting equipment is predicted, change abnormal factors are output to the mounting equipment, whether the modified mounting pressure needs to be dynamically adjusted in advance or not is judged by combining and analyzing the expansion effect factors and the change abnormal factors, and a corresponding control strategy is generated for the mounting equipment according to a judgment result. The monitoring system can actively adjust the mounting pressure in combination with the expansion effect factor and the change abnormal factor, effectively reduce the feeding point error when the influence factor changes, and generate a corresponding control strategy for the mounting equipment according to the multi-factor decision result, so that the mounting equipment can maintain the optimal mounting precision under the complex environment condition.

Description

Technical Field

[0001] The present invention relates to the technical field of product production monitoring, and in particular to a method and system for monitoring the production of an RFID wristband based on data analysis. Background Art

[0002] An RFID wristband is a smart wearable device that integrates radio frequency identification (RFID) technology. It is usually used in scenarios such as identity recognition, information tracking, and management control. It communicates with the reading device by embedding an RFID tag to achieve fast, contactless data transmission. An RFID wristband usually consists of two parts: an embedded chip (tag) and a matching reading device (such as an RFID reader).

[0003] The prior art has the following defects:

[0004] During the chip mounting process of the RFID wristband, when applied to ultra-high frequency (UHF) or microwave frequency bands (such as 2.4GHz), the size of the antenna is extremely small (for example, the side length of the microstrip antenna is only 10mm), and the position of the feeding point directly affects the resonant frequency and impedance matching. The feeding point error is essentially the deviation between the physical structure of the antenna and the design parameters, which is manifested as the deviation between the actual contact position of the chip and the antenna feeding point and the theoretical optimal position. This deviation will change the antenna resonant frequency and impedance matching characteristics. The wristband antenna is usually designed with a microstrip or flexible antenna. A feed point offset of 0.1mm can cause a frequency offset of more than 50MHz. This offset will cause the tag resonant frequency to deviate from the reader operating frequency band (such as UHF 860-960MHz), resulting in signal attenuation or even failure.

[0005] Based on this, the present invention proposes an RFID wristband production monitoring method and system based on data analysis, which can actively adjust the mounting pressure according to the results of multi-factor decision-making, effectively reduce the feeding point error when the influencing factors change, and generate corresponding control strategies for the mounting equipment, so that it can maintain the best mounting accuracy under complex environmental conditions. Summary of the invention

[0006] The purpose of the present invention is to provide a RFID wristband production monitoring method and system based on data analysis to solve the shortcomings of the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a RFID wristband production monitoring method based on data analysis, the monitoring method comprising the following steps:

[0008] The monitoring system obtains the substrate and chip information of the currently produced RFID wristband tags, and substitutes the substrate and chip information into the database to match the initial mounting pressure;

[0009] In each placement process, after analyzing the substrate thickness deviation and chip size deviation, the initial placement pressure is corrected based on the optimization algorithm to obtain the corrected placement pressure;

[0010] Generate the expansion effect factor of the substrate according to the environmental parameters of the processing workshop;

[0011] Monitor the working status and motion characteristics of the placement equipment in real time, use the Markov model to dynamically model the placement equipment, predict the changes in the working parameters of the placement equipment, and output the abnormal change factor for the placement equipment;

[0012] Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected placement pressure in advance, and a corresponding control strategy is generated for the placement equipment based on the judgment result.

[0013] In a preferred embodiment, the expansion effect factor and the abnormal change factor are analyzed to dynamically adjust the corrected placement pressure, including the following steps:

[0014] After obtaining the expansion effect factor and the abnormal change factor, the dynamic adjustment index is calculated and the expression is: , where is the dynamic adjustment index, is the expansion effect factor, is the abnormal change factor, , is the weight coefficient, and ;

[0015] The modified placement pressure is adjusted by dynamically adjusting the index, and the adjustment algorithm expression is:

[0016] , where To adjust the mounting pressure, To correct the mounting pressure, It is a dynamic adjustment index.

[0017] In a preferred embodiment, outputting a variation abnormality factor for a placement device includes the following steps:

[0018] The nozzle blockage degree and the vibration amplitude of the mounting head are normalized so that the value range of the nozzle blockage degree and the vibration amplitude of the mounting head is mapped to [0,1], and the nozzle blockage degree and the vibration amplitude of the mounting head after the normalization process are summed to obtain the proportional coefficient;

[0019] The proportionality factor is transferred to the state The probability of combining to generate the abnormal change factor is expressed as: , where is the abnormal change factor, is the proportionality coefficient, To transfer to the state probability.

[0020] In a preferred embodiment, the expansion effect factor of the substrate is generated according to the environmental parameters of the processing workshop;

[0021] When the substrate is conveyed to the mounting equipment through the conveying equipment, the workshop environment temperature is recorded in real time through the temperature sensor, and the workshop environment humidity is recorded in real time through the humidity sensor;

[0022] The real-time recorded workshop ambient temperature is compared with the preset temperature threshold, and the period when the workshop ambient temperature exceeds the temperature threshold is recorded as the temperature warning period; the real-time recorded workshop ambient humidity is compared with the preset humidity threshold, and the period when the workshop ambient humidity exceeds the humidity threshold is recorded as the humidity warning period;

[0023] The temperature warning period and the humidity warning period are accumulated and integrated to generate the expansion effect factor of the substrate. The expression is: ; In the formula, is the expansion effect factor, For the substrate at time The volume change at a given moment, It is the temperature warning period. This is the humidity warning period.

[0024] In a preferred embodiment, after analyzing the substrate thickness deviation and the chip size deviation, the initial mounting pressure is corrected based on an optimization algorithm, including the following steps:

[0025] The initial mounting pressure is corrected according to the thickness deviation and size deviation, and the initial mounting pressure is corrected based on the optimization algorithm. The optimization algorithm expression is: , where To correct the mounting pressure, is the initial mounting pressure, is the thickness deviation, is the size deviation, , is the adjustment coefficient, and the adjustment coefficient is greater than 0.

[0026] In a preferred embodiment, a laser thickness gauge or an inductive thickness sensor is used to measure the current substrate thickness, and the thickness deviation is obtained by subtracting the standard substrate thickness from the current substrate thickness. An industrial camera is used to obtain the size of the current chip, and the size deviation is obtained by subtracting the standard chip size from the current chip size.

[0027] In a preferred embodiment, the monitoring system substitutes the substrate and chip information into a database to match the initial mounting pressure, including the following steps:

[0028] Obtain substrate type, read substrate thickness, record substrate flexibility characteristics, and record chip size;

[0029] Search historical production data to find matching pressure range constraints based on substrate type, die size, and package type;

[0030] If there is historical mounting data with the same material, size, and process, extract its pressure setting value as the initial mounting pressure;

[0031] If there is no completely matching data, a recommendation algorithm based on similarity calculation is used to estimate the initial mounting pressure, and the initial mounting pressure needs to meet the pressure range constraint.

[0032] In a preferred embodiment, the initial placement pressure is estimated using a recommendation algorithm based on similarity calculation, including the following steps:

[0033] Extract the feature vector set {V_1, V_2, ..., V_N} of all historical production batches in the database, where N represents the number of historical production batches and V_i represents the production feature vector of the i-th batch;

[0034] The cosine similarity is used to calculate the similarity score between the current batch feature vector V_cur and the historical data {V_1,...,V_N};

[0035] The M groups of data with similarity scores greater than the score threshold are used as reference data sets, and the similarity scores of all batch production feature vectors in the reference data sets are summed to obtain the total similarity score value, and the weight of each batch production is obtained by dividing the similarity score by the total similarity score value;

[0036] The historical mounting pressure of each batch of production in the reference data set is extracted and marked as: {P_1, P_2, ..., P_M}, and the weighted average method is used to calculate the initial mounting pressure.

[0037] In a preferred embodiment, cosine similarity is used to calculate the similarity between the current batch feature vector V_cur and the historical data {V_1, ..., V_N}, and the expression is: , where Score the similarity between the current batch production feature vector and the i-th batch production feature vector, represents the dot product of the current batch production feature vector and the i-th batch production feature vector, Produce the feature vector norm for the current batch, Generate the feature vector norm for the i-th batch;

[0038] The weighted average method is used to calculate the initial mounting pressure, and the expression is: , where is the initial mounting pressure, is the number of batches in the reference dataset, is the weight of the i-th batch production, is the historical mounting pressure of the i-th batch of production.

[0039] The present invention also provides an RFID wristband production monitoring system based on data analysis, including an initial matching module, a correction module, an analysis module, and a control module;

[0040] Initial matching module: obtain the substrate and chip information of the currently produced RFID wristband label, and substitute the substrate and chip information into the database to match the initial mounting pressure;

[0041] Correction module: In each placement process, after analyzing the substrate thickness deviation and chip size deviation, the initial placement pressure is corrected based on the optimization algorithm to obtain the corrected placement pressure;

[0042] Analysis module: Generates the expansion effect factor of the substrate according to the environmental parameters of the processing workshop, monitors the working status and motion characteristics of the placement equipment in real time, uses the Markov model to dynamically model the placement equipment, predicts the changes in the working parameters of the placement equipment, and outputs the abnormal change factor for the placement equipment;

[0043] Control module: Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected placement pressure in advance, and a corresponding control strategy is generated for the placement equipment based on the judgment result.

[0044] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0045] 1. The present invention generates the expansion effect factor of the substrate according to the environmental parameters of the processing workshop, monitors the working state and motion characteristics of the mounting equipment in real time, and uses the Markov model to dynamically model the mounting equipment, predicts the changes in the working parameters of the mounting equipment, and outputs the abnormal change factor for the mounting equipment. Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected mounting pressure in advance, and a corresponding control strategy is generated for the mounting equipment based on the judgment result. The monitoring system combines the expansion effect factor and the abnormal change factor, so that the system can actively adjust the mounting pressure, effectively reduce the feed point offset when the influencing factors change, and generate a corresponding control strategy for the mounting equipment based on the multi-factor decision results, so that it can maintain the best mounting accuracy under complex environmental conditions.

[0046] 2. The present invention obtains the substrate and chip information of the currently produced RFID wristband label, substitutes the substrate and chip information into the database to match the initial mounting pressure, analyzes the substrate thickness deviation and chip size deviation in each mounting process, and corrects the initial mounting pressure based on the optimization algorithm to obtain the corrected mounting pressure. Since in actual applications, the thickness and size of the substrate and chip of the same batch may vary, the monitoring can adjust the mounting equipment to the optimal mounting pressure according to the substrate thickness and chip in each mounting process, improve the mounting accuracy, and ensure the quality of the RFID wristband label.

[0047] 3. The generation of the expansion effect factor in the present invention helps manufacturers to gain a deeper understanding of the relationship between environmental factors and substrate performance. Based on this factor, the company can further optimize the production process, such as adjusting parameters such as mounting pressure according to different environmental conditions, so that the production process can better adapt to environmental changes and improve production efficiency and product quality stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0049] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0051] Example 1: Please refer to Figure 1 As shown, the RFID wristband production monitoring method based on data analysis described in this embodiment includes the following steps:

[0052] The monitoring system obtains the substrate and chip information of the currently produced RFID wristband tags, and substitutes the substrate and chip information into the database to match the initial mounting pressure;

[0053] Obtain the substrate type (such as PI, PET, FR4, etc.). Read the substrate thickness (such as ±0.02mm tolerance range). Record the substrate flexibility characteristics (rigid or flexible substrate, which determines the mounting method). Record the chip size (length, width, thickness), which affects the mounting pressure calculation. Collect the pad area size and alignment requirements (such as ±0.1mm mounting accuracy requirements). Obtain the chip packaging material characteristics (such as silicon, ceramic, polymer, which affects the force bearing capacity).

[0054] According to the substrate type, chip size, and package type, historical production data is retrieved to find the most matching pressure range constraint. If there is historical placement data with the same material, size, and process, its pressure setting value is extracted as the initial placement pressure. If there is no completely matching data, the recommended algorithm based on similarity calculation is used to estimate the initial placement pressure. The initial placement pressure needs to meet the pressure range constraint (P min ~P max） , ensuring that it can be mounted firmly without damaging the chip or substrate.

[0055] In this application, a recommended algorithm based on similarity calculation is used to estimate the initial placement pressure, including the following steps:

[0056] Extract the feature vector set {V_1, V_2, ..., V_N} of all historical production batches in the database, where N represents the number of historical production batches and V_i represents the production feature vector of the i-th batch;

[0057] The cosine similarity is used to calculate the similarity between the current batch feature vector V_cur and the historical data {V_1,...,V_N}. The expression is: , where Score the similarity between the current batch production feature vector and the i-th batch production feature vector, represents the dot product of the current batch production feature vector and the i-th batch production feature vector, Produce the feature vector norm for the current batch, The norm of the feature vector for the i-th batch production, a higher similarity score indicates a higher similarity;

[0058] The M groups of data with similarity scores greater than the score threshold are used as reference data sets, and the similarity scores of all batch production feature vectors in the reference data sets are summed to obtain the total similarity score value, and the weight of each batch production is obtained by dividing the similarity score by the total similarity score value;

[0059] The historical mounting pressure of each batch of production is extracted from the reference data set and marked as: {P_1, P_2, ..., P_M}, and the weighted average method is used to calculate the initial mounting pressure, which is expressed as:

[0060] , where is the initial mounting pressure, is the number of batches in the reference dataset, is the weight of the i-th batch production, is the historical mounting pressure of the i-th batch of production.

[0061] In the production process of RFID wristbands, the process parameters, substrate characteristics, chip specifications, equipment status and other factors of each production batch may be different. Therefore, we need to structure the data of each production batch and generate the corresponding batch production feature vector;

[0062] The batch production feature vector is a multidimensional vector containing feature data of multiple categories:

[0063] (1) Material_Features

[0064] Substrate type (e.g. PI, PET, FR4);

[0065] Substrate thickness (e.g. 0.1mm, 0.2mm);

[0066] Coefficient of thermal expansion (CTE) (e.g. 20ppm / °C);

[0067] Surface roughness (such as Ra0.01μm~0.05μm);

[0068] Substrate cutting tolerance (e.g. 0.03mm to 0.05mm);

[0069] (2) Chip Features

[0070] Chip size (length, width, thickness, such as 2mm×2mm×0.3mm);

[0071] Pad area (e.g. 0.8mm²);

[0072] Package type (bare die, ceramic package, polymer package);

[0073] Placement point structure (single point vs. multi-point placement);

[0074] (3) Equipment parameters (Equipment_Features)

[0075] Device model (such as JUKIRX-7, ASMSiplaceX);

[0076] Nozzle diameter (e.g. 0.3mm, 0.5mm);

[0077] Mounting accuracy requirements (such as ±0.05mm, ±0.1mm);

[0078] Vacuum adsorption capacity of the nozzle (e.g. 30kPa, 50kPa);

[0079] Mounting pressure range (e.g. 1N to 10N);

[0080] (4) Workshop environment parameters (Environment_Features)

[0081] Current temperature (e.g. 25°C);

[0082] Current humidity (e.g. 50%);

[0083] Equipment vibration amplitude (e.g. 5μm, 10μm);

[0084] Workshop air flow disturbance index (such as low, medium, high);

[0085] (5) Production process parameters (Process_Features)

[0086] Mounting speed (e.g. 3000CPH, 6000CPH);

[0087] Pressure control mode (constant pressure mode / adaptive mode);

[0088] Type of visual system (2D vs. 3D);

[0089] Visual system calibration accuracy (e.g. ±0.05mm);

[0090] (6) Mounting historical data (Historical_Features)

[0091] The placement pressure of the same substrate + chip last time;

[0092] The yield rate of the same substrate + chip last time;

[0093] The average pressure value of the past 10 placements;

[0094] Defect rate of the past 10 placements (such as chip shift, poor soldering);

[0095] Collect batch production data and construct feature vectors

[0096] (1) Collect data. Data sources include:

[0097] MES (Manufacturing Execution System): data such as substrate, chip, equipment model, and mounting process.

[0098] Equipment sensors: mounting pressure, nozzle pressure, equipment vibration, temperature and humidity data, etc.

[0099] Visual inspection system: data such as mounting accuracy, pad alignment deviation, etc.

[0100] Historical production database: historical mounting data under similar process parameters.

[0101] In each placement process, after analyzing the substrate thickness deviation and chip size deviation, the initial placement pressure is corrected based on the optimization algorithm to obtain the corrected placement pressure;

[0102] During the mounting process of RFID wristbands, slight deviations in substrate thickness and chip size will affect the mounting quality. Therefore, we need to dynamically adjust the mounting pressure before each mounting to ensure that the chip adheres stably to the substrate while avoiding damage to the chip or deformation of the substrate due to excessive pressure.

[0103] The core goal of optimizing and correcting the mounting pressure is to adapt to the fluctuations in substrate thickness and chip size and ensure mounting reliability. By dynamically adjusting the pressure through the optimization algorithm, the mounting accuracy can be improved and the defect rate can be reduced.

[0104] Use a laser thickness gauge or an inductive thickness sensor to measure the current substrate thickness, and subtract the standard substrate thickness from the current substrate thickness to obtain the thickness deviation. When the thickness deviation is greater than 0, it indicates that the current substrate thickness has increased and the mounting pressure needs to be increased. When the thickness deviation is less than 0, it indicates that the current substrate thickness has decreased and the mounting pressure needs to be reduced. When the thickness deviation is equal to 0, it indicates that the current substrate thickness is equal to the standard substrate thickness, and the initial mounting pressure is maintained;

[0105] Use a high-precision industrial camera (such as 3D structured light scanning) to obtain the size of the current chip, including the height. Subtract the standard chip size from the current chip size to obtain the size deviation. If the size deviation is greater than 0, it means that the current chip size has increased and the mounting pressure needs to be increased. If the size deviation is less than 0, it means that the current chip size has decreased and the mounting pressure needs to be reduced. If the size deviation is equal to 0, it means that the current chip size remains unchanged and the initial mounting pressure is maintained.

[0106] The initial mounting pressure is corrected according to the thickness deviation and size deviation, and the initial mounting pressure is corrected based on the optimization algorithm. The optimization algorithm expression is: , where To correct the mounting pressure, is the initial mounting pressure, is the thickness deviation, is the size deviation, , is the adjustment coefficient, and the adjustment coefficient , Greater than 0. In this application, the adjustment coefficient is determined by fitting experimental data.

[0107] Generate the expansion effect factor of the substrate according to the environmental parameters of the processing workshop;

[0108] When the substrate is conveyed to the mounting equipment through the conveying equipment, the workshop environment temperature is recorded in real time through the temperature sensor, and the workshop environment humidity is recorded in real time through the humidity sensor;

[0109] The real-time recorded workshop ambient temperature is compared with the preset temperature threshold, and the period when the workshop ambient temperature exceeds the temperature threshold is recorded as the temperature warning period; the real-time recorded workshop ambient humidity is compared with the preset humidity threshold, and the period when the workshop ambient humidity exceeds the humidity threshold is recorded as the humidity warning period;

[0110] The temperature warning period and the humidity warning period are accumulated and integrated to generate the expansion effect factor of the substrate. The expression is: ; In the formula, is the expansion effect factor, For the substrate at time The volume change at a given moment, It is the temperature warning period. This is the humidity warning period.

[0111] During the production process, the temperature and humidity environment of the workshop will have an impact on the substrate, especially causing the expansion of the substrate. By recording the workshop environment temperature and humidity in real time, comparing with the preset threshold to determine the warning period, and then using integral calculation to generate the expansion effect factor, it is possible to accurately quantify the comprehensive impact of temperature and humidity changes on the expansion of the substrate at different times. This allows the effect of environmental factors on the substrate during the production process to be expressed in specific values, providing accurate data support for subsequent production control and quality analysis.

[0112] This application records the temperature warning period and humidity warning period, which means that it can timely detect situations where environmental parameters exceed the normal range. When these warning periods occur, production personnel can take measures in advance, such as adjusting the air conditioning system in the workshop to control the temperature and humidity, or taking special treatment methods for the substrate to avoid quality problems such as excessive expansion of the substrate due to environmental factors, thereby playing the role of early warning and prevention, and reducing the production of defective products.

[0113] The generation of the expansion effect factor helps manufacturers gain a deeper understanding of the relationship between environmental factors and substrate performance. Based on this factor, companies can further optimize the production process, such as adjusting parameters such as mounting pressure according to different environmental conditions, making the production process more adaptable to environmental changes and improving production efficiency and product quality stability.

[0114] During the mounting process of RFID wristband production, increases in ambient temperature and humidity can cause the substrate to expand, mainly due to the thermal expansion and moisture expansion effects of the material.

[0115] When the temperature of a substrate (such as FR4 glass fiber board, polyimide PI film, PET plastic) rises, the internal molecular distance will increase, causing the overall size of the material to increase.

[0116] The coefficient of thermal expansion (CTE) of different substrates is shown in Table 1:

[0117] Table 1 Thermal expansion coefficients of different substrates Material Linear thermal expansion coefficient (ppm / ℃) FR4 (Fiberglass Board) 10~16ppm / ℃ PI (Polyimide) 20~40ppm / ℃ PET (Polyethylene terephthalate) 50~70ppm / ℃

[0118] FR4 substrate: Due to the glass fiber content, the thermal expansion is small, but it still produces micron-level expansion with temperature changes.

[0119] PI (polyimide) and PET (plastic film): Thermal expansion is more obvious. Temperature rise may cause expansion of 0.01mm level, affecting the placement accuracy.

[0120] RFID wristband substrates are usually multi-layer structures (such as PI / FR4 stacking).

[0121] Due to the different CTE of materials in different layers, they expand unevenly when the temperature rises, resulting in local deformation, warping or micro-displacement.

[0122] High-frequency RFID antennas require extremely high precision, but the expansion of the substrate may cause the feed point position to shift (e.g., ±0.05mm), affecting wireless performance.

[0123] Substrates such as FR4, PI, PET, etc. are mostly polymer materials, which will absorb moisture in the air, causing the volume to expand. When water molecules enter the material, they will increase the molecular distance, causing the material size to increase, which is called hygroscopic expansion. Humidity changes will affect the dielectric constant and dimensional stability of the material, especially in high-precision RFID production. The hygroscopic expansion coefficients (typical values) of different substrates are shown in Table 2:

[0124] Table 2 Expansion coefficient of different substrates Material Coefficient of moisture expansion (ppm / %RH) FR4 (Fiberglass Board) 10~20ppm / %RH PI (Polyimide) 50~100ppm / %RH PET (Polyethylene terephthalate) 150~250ppm / %RH

[0125] FR4 board has low moisture absorption rate and small swelling effect, but long-term high humidity will still cause dimensional drift. PI and PET substrates have higher moisture absorption rate. A 10%RH increase in humidity may cause 0.01mm level expansion, affecting mounting accuracy. In multi-layer RFID structures, the swelling degree of each layer is different, causing interlayer peeling or warping of the substrate. In extreme cases (humidity>85%RH), it will cause the antenna circuit to move slightly, affecting the RFID read and write distance and frequency matching.

[0126] Monitor the working status and motion characteristics of the placement equipment in real time, use the Markov model to dynamically model the placement equipment, predict the changes in the working parameters of the placement equipment, and output the abnormal change factor for the placement equipment;

[0127] The nozzle blockage degree and the vibration amplitude of the mounting head of the mounting equipment are collected. The calculation logic of the nozzle blockage degree is: obtain the vacuum drop value by subtracting the current vacuum degree of the nozzle from the normal vacuum degree of the nozzle, and obtain the nozzle blockage degree value by dividing the vacuum drop value by the normal vacuum degree of the nozzle. The vibration amplitude of the mounting head is directly obtained through the vibration sensor set on the mounting head. During the mounting process, if the vibration amplitude of the mounting head is too large, the mounting head may fail to stably press the chip, resulting in insufficient pressure. The greater the nozzle blockage degree, the pressure under the nozzle will decrease, the mounting force will be reduced, and thus the adsorption force will be reduced. If the adsorption force is insufficient, the chip may slide slightly during the mounting process, resulting in a decrease in the mounting pressure.

[0128] The nozzle blockage degree and the vibration amplitude of the placement head of the placement equipment are collected. When the acquired nozzle blockage degree and the vibration amplitude of the placement head change, the Markov model is used to dynamically model the placement equipment and define the following state variables of the placement equipment:

[0129] set up For placement equipment in time Status: , Indicates the degree of nozzle blockage. Represents the vibration amplitude of the placement head, and the state space is divided into:

[0130] S 1 : Normal state (stable mounting pressure, low blocking rate, low vibration amplitude);

[0131] S 2 : Mild abnormality (the predicted placement pressure is slightly affected and needs to be increased in advance to ensure placement stability, and the blockage rate or vibration amplitude is slightly higher);

[0132] S 3 : Moderate abnormality (predicted placement pressure is affected less, placement equipment is not supported, blockage rate and vibration amplitude are moderate);

[0133] S4: Severe abnormality (the predicted placement pressure is seriously reduced, the placement equipment does not support use, the blockage rate is high, and the vibration amplitude is high).

[0134] Statistics on the state changes of the placement equipment history, and calculate the state changes of the placement equipment when the nozzle blockage degree and the vibration amplitude of the placement head change. Transfer to state The probability of , and estimate the state transfer matrix P through historical data: , where Indicates the current status Transfer to the next state probability.

[0135] Calculate the state probability distribution at the next moment and let the current state probability be: , then the state probability at the next moment is: ,like , ( As the threshold is usually set to 50%), it is analyzed that when the nozzle blockage degree and the vibration amplitude of the placement head increase, it is predicted that the placement pressure will decrease. , it is judged that the placement equipment does not support operation and needs to be managed in advance. The management includes increasing the placement pressure in advance, ensuring the stability of the placement pressure, and controlling the placement equipment to stop running;

[0136] Predicting that the placement pressure will decrease and outputting the abnormal change factor for the placement equipment includes the following steps:

[0137] The nozzle blockage degree and the vibration amplitude of the mounting head are normalized so that the value range of the nozzle blockage degree and the vibration amplitude of the mounting head is mapped to [0,1]. The nozzle blockage degree and the vibration amplitude of the mounting head after the normalization process are summed to obtain the proportional coefficient. The larger the proportional coefficient, the larger the nozzle blockage degree and the vibration amplitude of the mounting head.

[0138] The proportionality factor is transferred to the state The probability of combining to generate the abnormal change factor is expressed as: , where is the abnormal change factor, is the proportionality coefficient, To transfer to the state The probability of abnormal change factor is larger, indicating that under comprehensive analysis, the nozzle blockage degree and the vibration amplitude of the placement head increase, resulting in the transfer to state The greater the probability.

[0139] Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the correction mounting pressure in advance. , ( As the threshold is usually set to 50%, it is analyzed that when the nozzle blockage degree and the vibration amplitude of the placement head increase, it is predicted that the placement pressure will decrease, and it is judged that the correction placement pressure needs to be dynamically adjusted in advance. No adjustment is required, and the corresponding control strategy is generated for the placement equipment based on the judgment result;

[0140] When the expansion effect factor is large, it means that the volume change of the substrate during the humidity warning period is large, that is, the substrate has undergone significant expansion due to changes in environmental humidity. Therefore, when the expansion effect factor of the substrate is large, the mounting pressure should be appropriately reduced to avoid damage to the chip due to excessive mounting pressure, and to adapt to the volume change of the substrate and reduce the risk of mounting misalignment.

[0141] When the abnormal change factor is large, it means that the nozzle blockage degree and the vibration amplitude of the placement head increase, resulting in the transfer to the state The probability of failure increases. At this time, it is predicted that the mounting pressure will decrease. In order to ensure stable mounting, the mounting pressure needs to be increased in advance.

[0142] In summary, after obtaining the expansion effect factor and the abnormal change factor, the dynamic adjustment index is calculated and the expression is: , where is the dynamic adjustment index, is the expansion effect factor, is the abnormal change factor, , is the weight coefficient, and ;

[0143] The modified placement pressure is adjusted by dynamically adjusting the index, and the adjustment algorithm expression is:

[0144] , where To adjust the mounting pressure, To correct the mounting pressure, is the dynamic adjustment index. Assuming that in the dynamic adjustment index, the expansion effect factor takes the value of 0, if the abnormal change factor increases, the dynamic adjustment index will be negative and will become smaller and smaller, indicating that the predicted placement pressure will decrease. When it is substituted into the adjustment algorithm, the adjustment placement pressure will increase. When the expansion effect factor increases, the placement pressure needs to be reduced. At this time, if the abnormal change factor increases, the predicted placement pressure will decrease. In this case, the dynamic adjustment index may take the value of 0, that is, no further adjustment is required, thereby improving control efficiency.

[0145] A corresponding control strategy is generated for the placement equipment according to the judgment result. The control strategy is: a control instruction is generated according to adjusting the placement pressure to control the operation of the placement equipment.

[0146] It should be noted that when the output regulated mounting pressure is not within the mounting pressure range constraints, continuing processing will lead to an increase in the defective rate. At this time, the monitoring system sends a warning signal to the management personnel and controls the mounting equipment to stop running.

[0147] This application generates the expansion effect factor of the substrate based on the environmental parameters of the processing workshop, monitors the working state and motion characteristics of the mounting equipment in real time, uses the Markov model to dynamically model the mounting equipment, predicts the changes in the working parameters of the mounting equipment, and outputs the abnormal change factor for the mounting equipment. Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected mounting pressure in advance, and generates a corresponding control strategy for the mounting equipment based on the judgment result. The monitoring system combines the expansion effect factor and the abnormal change factor, so that the system can actively adjust the mounting pressure, effectively reduce the feeding point error when the influencing factors change, and generate a corresponding control strategy for the mounting equipment based on the multi-factor decision results, so that it can maintain the best mounting accuracy under complex environmental conditions.

[0148] This application obtains the substrate and chip information of the currently produced RFID wristband labels, and substitutes the substrate and chip information into the database to match the initial mounting pressure. During each mounting process, after analyzing the substrate thickness deviation and chip size deviation, the initial mounting pressure is corrected based on the optimization algorithm to obtain the corrected mounting pressure. Since in actual applications, the thickness and size of the substrate and chip of the same batch may vary, monitoring can adjust the mounting equipment to the optimal mounting pressure according to the substrate thickness and chip during each mounting process, improve mounting accuracy, and ensure the quality of RFID wristband labels.

[0149] Embodiment 2: The RFID wristband production monitoring system based on data analysis described in this embodiment includes an initial matching module, a correction module, an analysis module, and a control module;

[0150] Initial matching module: obtains the substrate and chip information of the currently produced RFID wristband label, and substitutes the substrate and chip information into the database to match the initial mounting pressure, and the initial mounting pressure is sent to the correction module;

[0151] Correction module: In each mounting process, after analyzing the substrate thickness deviation and chip size deviation, the initial mounting pressure is corrected based on the optimization algorithm to obtain the corrected mounting pressure, which is then sent to the analysis module.

[0152] Analysis module: Generates the expansion effect factor of the substrate according to the environmental parameters of the processing workshop, monitors the working status and motion characteristics of the placement equipment in real time, uses the Markov model to dynamically model the placement equipment, predicts the changes in the working parameters of the placement equipment, and outputs the abnormal change factor for the placement equipment. The abnormal change factor and the expansion effect factor are sent to the control module;

[0153] Control module: Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected placement pressure in advance, and a corresponding control strategy is generated for the placement equipment based on the judgment result.

[0154] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0155] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0156] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. The RFID wristband production monitoring method based on data analysis is characterized by: The monitoring method comprises the following steps: The monitoring system obtains the substrate and chip information of the currently produced RFID wristband tags, and substitutes the substrate and chip information into the database to match the initial mounting pressure; In each placement process, after analyzing the substrate thickness deviation and chip size deviation, the initial placement pressure is corrected based on the optimization algorithm to obtain the corrected placement pressure; Generate the expansion effect factor of the substrate according to the environmental parameters of the processing workshop; Monitor the working status and motion characteristics of the placement equipment in real time, use the Markov model to dynamically model the placement equipment, predict the changes in the working parameters of the placement equipment, and output the abnormal change factor for the placement equipment; Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected placement pressure in advance, and a corresponding control strategy is generated for the placement equipment based on the judgment result.

2. The RFID wristband production monitoring method based on data analysis according to claim 1 is characterized in that: Combined with the analysis of the expansion effect factor and the abnormal change factor, the correction mounting pressure is dynamically adjusted, including the following steps: After obtaining the expansion effect factor and the abnormal change factor, the dynamic adjustment index is calculated and the expression is: , where is the dynamic adjustment index, is the expansion effect factor, is the abnormal change factor, , is the weight coefficient, and ; The modified placement pressure is adjusted by dynamically adjusting the index, and the adjustment algorithm expression is: , where To adjust the mounting pressure, To correct the mounting pressure, It is a dynamic adjustment index.

3. The RFID wristband production monitoring method based on data analysis according to claim 2 is characterized in that: Outputting the variation anomaly factor for the placement equipment includes the following steps: The nozzle blockage degree and the vibration amplitude of the mounting head are normalized so that the value range of the nozzle blockage degree and the vibration amplitude of the mounting head is mapped to [0,1], and the nozzle blockage degree and the vibration amplitude of the mounting head after the normalization process are summed to obtain the proportional coefficient; The proportionality factor is transferred to the state The probability of combining to generate the abnormal change factor is expressed as: , where is the abnormal change factor, is the proportionality coefficient, To transfer to the state probability.

4. The RFID wristband production monitoring method based on data analysis according to claim 3 is characterized in that: Generate the expansion effect factor of the substrate according to the environmental parameters of the processing workshop; When the substrate is conveyed to the mounting equipment through the conveying equipment, the workshop environment temperature is recorded in real time through the temperature sensor, and the workshop environment humidity is recorded in real time through the humidity sensor; The real-time recorded workshop ambient temperature is compared with the preset temperature threshold, and the period when the workshop ambient temperature exceeds the temperature threshold is recorded as the temperature warning period; the real-time recorded workshop ambient humidity is compared with the preset humidity threshold, and the period when the workshop ambient humidity exceeds the humidity threshold is recorded as the humidity warning period; The temperature warning period and the humidity warning period are accumulated and integrated to generate the expansion effect factor of the substrate. The expression is: ; In the formula, is the expansion effect factor, For the substrate at time The volume change at a given moment, It is the temperature warning period. This is the humidity warning period.

5. The RFID wristband production monitoring method based on data analysis according to claim 4 is characterized in that: After analyzing the substrate thickness deviation and chip size deviation, the initial mounting pressure is corrected based on the optimization algorithm, including the following steps; The initial mounting pressure is corrected according to the thickness deviation and size deviation, and the initial mounting pressure is corrected based on the optimization algorithm. The optimization algorithm expression is: , where To correct the mounting pressure, is the initial mounting pressure, is the thickness deviation, is the size deviation, , is the adjustment coefficient, and the adjustment coefficient is greater than 0.

6. The RFID wristband production monitoring method based on data analysis according to claim 5 is characterized in that: A laser thickness gauge or an inductive thickness sensor is used to measure the current substrate thickness, and the thickness deviation is obtained by subtracting the standard substrate thickness from the current substrate thickness. An industrial camera is used to obtain the size of the current chip, and the size deviation is obtained by subtracting the standard chip size from the current chip size.

7. The RFID wristband production monitoring method based on data analysis according to claim 6 is characterized in that: The monitoring system enters the substrate and chip information into the database to match the initial placement pressure, including the following steps: Obtain substrate type, read substrate thickness, record substrate flexibility characteristics, and record chip size; Search historical production data to find matching pressure range constraints based on substrate type, die size, and package type; If there is historical mounting data with the same material, size, and process, extract its pressure setting value as the initial mounting pressure; If there is no completely matching data, a recommendation algorithm based on similarity calculation is used to estimate the initial mounting pressure, and the initial mounting pressure needs to meet the pressure range constraint.

8. The RFID wristband production monitoring method based on data analysis according to claim 7 is characterized in that: The initial placement pressure is estimated using a recommended algorithm based on similarity calculation, including the following steps: Extract the feature vector set {V_1, V_2, ..., V_N} of all historical production batches in the database, where N represents the number of historical production batches and V_i represents the production feature vector of the i-th batch; The cosine similarity is used to calculate the similarity score between the current batch feature vector V_cur and the historical data {V_1,...,V_N}; The M groups of data with similarity scores greater than the score threshold are used as reference data sets, and the similarity scores of all batch production feature vectors in the reference data sets are summed to obtain the total similarity score value, and the weight of each batch production is obtained by dividing the similarity score by the total similarity score value; The historical mounting pressure of each batch of production in the reference data set is extracted and marked as: {P_1, P_2, ..., P_M}, and the weighted average method is used to calculate the initial mounting pressure.

9. The RFID wristband production monitoring method based on data analysis according to claim 8 is characterized in that: The cosine similarity is used to calculate the similarity between the current batch feature vector V_cur and the historical data {V_1,...,V_N}. The expression is: , where Score the similarity between the current batch production feature vector and the i-th batch production feature vector, represents the dot product of the current batch production feature vector and the i-th batch production feature vector, Produce the feature vector norm for the current batch, Generate the feature vector norm for the i-th batch; The weighted average method is used to calculate the initial mounting pressure, and the expression is: , where is the initial mounting pressure, is the number of batches in the reference dataset, is the weight of the i-th batch production, is the historical mounting pressure of the i-th batch of production.

10. An RFID wristband production monitoring system based on data analysis, used to implement the monitoring method according to any one of claims 1 to 9, characterized in that: It includes an initial matching module, a correction module, an analysis module, and a control module; Initial matching module: obtain the substrate and chip information of the currently produced RFID wristband label, and substitute the substrate and chip information into the database to match the initial mounting pressure; Correction module: In each placement process, after analyzing the substrate thickness deviation and chip size deviation, the initial placement pressure is corrected based on the optimization algorithm to obtain the corrected placement pressure; Analysis module: Generates the expansion effect factor of the substrate according to the environmental parameters of the processing workshop, monitors the working status and motion characteristics of the placement equipment in real time, uses the Markov model to dynamically model the placement equipment, predicts the changes in the working parameters of the placement equipment, and outputs the abnormal change factor for the placement equipment; Control module: Combined with the analysis of the expansion effect factor and the abnormal change factor, it is determined whether it is necessary to dynamically adjust the corrected placement pressure in advance, and a corresponding control strategy is generated for the placement equipment based on the judgment result.

Images