Industrial big data governance method and system

Abstract

The present invention discloses an industrial big data management method and system, mainly targeting the field of semiconductor manufacturing. The method first collects process parameters and equipment operation data by installing multiple sensors on the equipment. After preprocessing, time series analysis and wavelet transform are used to extract the fluctuation characteristics of the photolithography data. The photoresist thickness fluctuation model is constructed by combining historical data and process principles. Based on this model, an improved PID control algorithm is used to formulate and implement a fluctuation suppression strategy to stabilize the photoresist thickness. At the same time, the relevant data is classified and stored for query and call. The present invention effectively solves the data management and process control problems in semiconductor manufacturing, improves data quality, ensures process stability, improves production efficiency and product quality, and promotes the intelligent development of the semiconductor manufacturing industry.

Description

Technical Field

[0001] The present invention relates to fields related to data processing, and specifically to an industrial big data governance method and system. Background Art

[0002] As the semiconductor manufacturing industry is deeply integrated into the wave of industrial big data development, its data governance faces numerous severe challenges, which are hindering the high-quality development of the industry. In the field of industrial data processing in semiconductor manufacturing, with the continuous advancement of technology, chip manufacturing processes are becoming increasingly complex and sophisticated, placing extremely high demands on data processing and management during the production process. As a core link in semiconductor manufacturing, the stability of photolithography processes directly determines the quality and performance of chips, and precise control of photoresist thickness is the key to the photolithography process.

[0003] In the traditional semiconductor manufacturing process, in the photolithography process, key parameters that affect the thickness of the photoresist, such as the temperature of the photolithography equipment near the photoresist coating area and the pressure applied to the photoresist during coating, are not effectively monitored in real time and data collection is not carried out. This makes it impossible to accurately grasp the changing pattern of the photoresist thickness and difficult to timely and accurately regulate the production process. In terms of controlling the thickness of the photoresist, existing technical means cannot meet the high-precision process requirements. Since the thickness of the photoresist is affected by a combination of factors such as exposure energy, temperature, the characteristics of the photoresist itself, equipment aging, and ambient humidity, traditional control strategies are difficult to fully consider these complex factors. For example, when faced with performance changes caused by equipment aging, traditional methods are unable to adjust the control parameters in a timely manner, resulting in large fluctuations in the thickness of the photoresist, which in turn affects the yield rate of the chip. To address the above problems, an industrial big data governance method and system are designed. Summary of the Invention

[0004] The purpose of the present invention is to provide an industrial big data governance method and system to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an industrial big data governance method, comprising the following steps:

[0006] Step 1: Data collection: By installing temperature sensors, pressure sensors, and photolithography precision sensors on semiconductor manufacturing equipment, process parameter data during the manufacturing process can be collected in real time.

[0007] Step 2: Data preprocessing: Perform preliminary cleaning on the collected data, remove obvious erroneous values, fill in a small number of missing values, standardize the data, unify the data format, and obtain preprocessed data;

[0008] Step 3: Fluctuation feature extraction: Receive pre-processed data and extract the fluctuation period and amplitude characteristics of the exposure time and photoresist thickness changes in the lithography process. Use wavelet transform technology to decompose the data into different frequency components to identify high-frequency noise fluctuations and low-frequency trend fluctuations.

[0009] Step 4: Establish a fluctuation model: Based on the extracted fluctuation characteristics and combined with the principles of semiconductor manufacturing process, a process fluctuation prediction model is established. Specifically, based on historical data and physical models, a mathematical model is constructed between the fluctuation of photoresist thickness and exposure energy and temperature in the photolithography process. This model simulates the fluctuation of process parameters under different conditions.

[0010] Step 5: Formulate and implement a fluctuation suppression strategy: Based on the fluctuation prediction model, formulate a corresponding fluctuation suppression strategy. That is, when a large fluctuation in photoresist thickness is predicted, the exposure energy and temperature control parameters are automatically adjusted. The feedback control system intervenes in the manufacturing process in real time to suppress fluctuations in process parameters and ensure the stability of the semiconductor manufacturing process.

[0011] Step 6. Data storage: Store the preprocessed data in step 2 and the fluctuation suppression strategy data in step 5 in a distributed database, and classify and store them according to different process links, equipment types, and time series for quick query and call.

[0012] Preferably, the process parameter data in step 1 specifically include the following: photolithography link: install a temperature sensor near the photoresist coating area of ​​the photolithography equipment to monitor the ambient temperature of the photoresist during the photolithography process; add a pressure sensor to the photoresist coating equipment to collect pressure data applied to the photoresist during coating; use a high-precision photolithography precision sensor to record the accuracy deviation between the photolithography pattern and the actual photoresist forming in real time; related link: install a flow sensor on the etching equipment to collect etching gas flow data, and the etching gas flow affects the etching rate.

[0013] Preferably, the specific steps of extracting the fluctuation characteristics are as follows:

[0014] Step S31, data preparation: collect the exposure time and photoresist thickness data of the photolithography step in step 2 for a period of time, so as to obtain continuous Exposure time series at each time point and the photoresist thickness sequence ;

[0015] Step S32: Time series analysis algorithm:

[0016] Periodic analysis: Use Fourier transform to extract periodic features. The Fourier transform formula is: ,in is the original time series, i.e., the exposure time and photoresist thickness , It is the transformed frequency domain representation, which is obtained by exposing the time series Perform Fourier transform to obtain its amplitude at different frequencies. In the frequency domain, the period corresponding to the frequency component with larger amplitude is the fluctuation period of exposure time. The same principle is applied to the photoresist thickness sequence. , find the fluctuation period;

[0017] Amplitude feature extraction: Extract the fluctuation amplitude feature and calculate the range of the time series, that is, for the exposure time series , calculate the range Indicates the maximum fluctuation range of exposure time within the observation period; for the photoresist thickness sequence , very poor Indicates the fluctuation amplitude of photoresist thickness;

[0018] Step 33: Decompose the frequency components by wavelet transform: Use discrete wavelet transform to process the exposure time and photoresist thickness data. Discrete wavelet transform selects appropriate wavelet basis function ,in is the scale parameter, is the translation parameter, for the exposure time series Perform discrete wavelet transform , we get the wavelet coefficients at different scales and positions. At high-frequency scales, the wavelet coefficients reflect the fast-changing part in the exposure time series, that is, the high-frequency noise fluctuation. Similarly, for the photoresist thickness series Perform discrete wavelet transform , separate its high-frequency noise fluctuations and low-frequency trend fluctuations, that is, fully grasp the fluctuation characteristics of the lithography process parameter data, and provide an accurate data basis for the subsequent establishment of the fluctuation model.

[0019] Preferably, the specific steps of establishing the fluctuation model are as follows:

[0020] Step 41: Data collation and preparation: Collect historical data from the photolithography process, including the photoresist thickness measurements during the production of different batches of products, the corresponding exposure energy settings, and the ambient temperature or temperature data of key parts of the equipment. Set data and record the photoresist thickness as , the exposure energy is recorded as , the temperature is recorded as ;

[0021] Step 42: Modeling is performed using a multiple linear regression algorithm. Based on the principle of photolithography, the thickness of the photoresist is affected by the combined influence of exposure energy and temperature. The multiple linear regression algorithm is used for modeling, specifically: ,in is the dependent variable, i.e., the photoresist thickness, and , is the intercept, and is the regression coefficient, is a random error term, where 、 and Estimates based on historical data;

[0022] Step 43, parameter estimation: use the least square method to estimate the regression coefficient so that the photoresist thickness and the model prediction value The sum of squared errors between Minimum; through About 、 and Find the partial derivatives and set them to zero to obtain a system of equations:

[0023]

[0024] Solving the above system of equations, we get 、 and Estimated value of 、 and The final photoresist thickness fluctuation prediction model is ,in is the predicted value, is the exposure energy, is temperature;

[0025] Step 44: Model Testing and Optimization: Using the Coefficient of Determination The photoresist thickness fluctuation prediction model is evaluated, where ,in yes The average value of The closer it is to 1, the better the model fits the data.

[0026] Preferably, the specific working logic of the fluctuation suppression strategy formulation and execution is as follows:

[0027] Threshold setting: Based on historical data and the quality standards of the photolithography process, the normal fluctuation range of the photoresist thickness is determined, and the average normal photoresist thickness is set to , the photoresist thickness fluctuation range is ,in A reasonable fluctuation threshold set according to process accuracy requirements;

[0028] Prediction and judgment: Use the established photoresist thickness fluctuation prediction model , input the current exposure energy and temperature data in real time, predict the photoresist thickness, when the predicted value Extraordinary fluctuation range When , it is determined that the photoresist thickness will fluctuate greatly;

[0029] Strategy formulation: Use PID control algorithm to formulate control strategy, specifically, its output By the proportional term , integral item and the differential term , the formula is: , is the error at the current moment, that is, the average value of normal photoresist thickness and predicted value The difference , 、 and They are proportional coefficient, integral coefficient and differential coefficient respectively, among which the proportional term According to the size of the current error, the control amount is adjusted immediately; the integral term Used to eliminate the steady-state error of the system; differential term Adjust the control amount in advance according to the changing trend of the error; the above coefficients are specifically improved according to the characteristics of the photolithography process and the response of the equipment. The improvement is based on the deviation degree and change rate of the predicted photoresist thickness from the normal range, and the coefficients are dynamically adjusted;

[0030] Considering multi-factor interference compensation: the thickness of the photoresist is not only affected by exposure energy and temperature, but also by the characteristics of the photoresist itself, equipment aging, and environmental humidity. Introducing interference compensation terms into the PID algorithm, a multi-factor interference model is established; collecting data on the impact of different factors on the thickness of the photoresist, and analyzing the relationship between each factor and the change in the photoresist thickness. Specifically, let the interference factor set be , the corresponding compensation coefficient is , then the control quantity after compensation is ,in is the control quantity after dynamic adjustment of coefficient;

[0031] Strategy execution: Control quantity calculated according to PID algorithm , automatically adjust the exposure energy control device of the exposure equipment and the control parameters of the temperature regulation equipment.

[0032] Preferably, the 、 and Specific improvements are made based on the characteristics of the lithography process and the response of the equipment. The specific implementation steps are as follows:

[0033] Step A, real-time data acquisition: During the photolithography process, the photoresist thickness data is collected in real time through the sensor. The photoresist thickness is measured as , the normal photoresist thickness range is , and obtain the exposure energy at the corresponding moment and temperature ;

[0034] Step B, calculate the rate of change: calculate the photoresist thickness error Rate of change ,in is the time interval for data collection, and the error change rate, i.e., the change trend of the photoresist thickness error, is obtained by calculating the error values ​​at adjacent moments;

[0035] Step C: Construct coefficient adjustment function:

[0036] Determine the function form: Use multiple linear regression to construct the coefficient adjustment function, where the proportional coefficient The adjustment function is , let the function form be ,in 、 and is the coefficient to be determined; similarly, the integral coefficient The adjustment function is set to , differential coefficient The adjustment function is set to ;

[0037] Coefficient training: Through the preset experimental data, the coefficients in the above function are determined, and several sets of experimental data under different photolithography process conditions are prepared. Each set of data contains the measured value of the photoresist thickness and the corresponding error. , error change rate and the best performing 、 and value, use the least square method to solve the coefficient; for Solve the adjustment function coefficients, assuming that the experimental data are group, target , respectively 、 and Taking the partial derivatives and setting them equal to zero, we obtain the system of equations:

[0038]

[0039] Solving this system of equations, we get 、 and The value of The adjustment function of and Adjust the coefficients in the function 、 、 and 、 、 ;

[0040] Step D, dynamic coefficient update: calculate the error , error change rate , substitute into the determined coefficient adjustment function to calculate the adjusted coefficient to be used at the current moment 、 and , substitute the adjusted coefficient into the PID control algorithm formula , get the control quantity after dynamic adjustment coefficient , calculate the control quantity .

[0041] Preferably, an industrial big data management system is implemented according to an industrial big data management method, and the system includes:

[0042] The data acquisition module collects process parameters and equipment operating status data in the manufacturing process in real time through sensors, and cleans the collected data of errors, fills in missing values, and standardizes the data format to obtain pre-processed data;

[0043] The semiconductor process fluctuation suppression module uses a time series algorithm to process process parameter data, using wavelet transforms to distinguish high-frequency noise from low-frequency trend fluctuations. It then constructs a prediction model based on the fluctuation characteristics and process principles, namely a mathematical model of photoresist thickness, exposure energy, and temperature. Based on the model predictions, it automatically adjusts the exposure energy parameters to stabilize the process.

[0044] The data storage module is used to store the pre-processed data in the acquisition module and the fluctuation suppression strategy data in the semiconductor process fluctuation suppression module in a distributed database, and classify and store them according to different process links, equipment types and time series.

[0045] Compared with existing technologies, the present invention offers the following advantages: It precisely captures fluctuation characteristics and optimizes the photolithography process. By utilizing a time series analysis algorithm and wavelet transform technology to extract fluctuation characteristics, it can accurately determine the fluctuation period, amplitude, and frequency components of key data such as exposure time and photoresist thickness in the photolithography process, and clearly distinguish between high-frequency noise fluctuations and low-frequency trend fluctuations. This facilitates a deeper understanding of the dynamic changes in the photolithography process and provides precise data support for establishing accurate fluctuation models, thereby enabling precise control and optimization of the photolithography process and improving the stability and accuracy of the photolithography process.

[0046] Establishing a reliable fluctuation model to predict process fluctuations: A mathematical model linking photoresist thickness fluctuations with exposure energy and temperature, constructed based on historical data and physical models, fully considers the principles of semiconductor manufacturing processes and effectively simulates fluctuations in process parameters under different conditions. Through rigorous data organization, multivariate linear regression modeling, precise parameter estimation, and scientific model validation and optimization, the model's reliability and accuracy are ensured. This model can predict photoresist thickness fluctuations in advance, providing a reliable basis for fluctuation suppression strategies, enabling production processes to proactively address potential process fluctuations and ensuring the stability of semiconductor manufacturing processes.

[0047] Implementing an effective fluctuation suppression strategy to ensure product quality: This fluctuation suppression strategy, developed based on a fluctuation prediction model and combined with a PID control algorithm and optimized for photolithography process characteristics, rapidly responds to abnormal fluctuations in photoresist thickness. Dynamic coefficient adjustment adapts to varying process conditions, while also accounting for multi-factor interference compensation to comprehensively address the complex situation where photoresist thickness is affected by multiple factors. Automatically adjusting exposure energy and temperature control parameters, and providing real-time intervention in the manufacturing process, effectively suppresses photoresist thickness fluctuations, ensuring that photoresist thickness remains within the normal range. This ensures chip manufacturing quality and performance, and improves product yield. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the method flow structure of the present invention;

[0049] Figure 2 Schematic diagram of the process of extracting the fluctuation characteristics of the present invention;

[0050] Figure 3 This is a schematic flow chart of the steps for establishing a fluctuation model according to the present invention;

[0051] Figure 4 A schematic diagram of the workflow for formulating and executing the fluctuation suppression strategy of the present invention;

[0052] Figure 5 Schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0054] Example 1

[0055] See also Figure 1 , the present invention provides a technical solution: an industrial big data governance method, comprising the following steps:

[0056] Step 1: Data collection: By installing temperature sensors, pressure sensors, and photolithography precision sensors on semiconductor manufacturing equipment, process parameter data during the manufacturing process can be collected in real time.

[0057] The process parameter data specifically include the following: Photolithography link: Install a temperature sensor near the photoresist coating area of ​​the photolithography equipment to monitor the ambient temperature of the photoresist during the photolithography process. Temperature changes affect the reaction of the photoresist and thus change the thickness; Add a pressure sensor to the photoresist coating equipment to collect pressure data applied to the photoresist during coating. The pressure has a direct effect on the initial coating thickness of the photoresist; Through the high-precision photolithography precision sensor, the accuracy deviation between the photolithography pattern and the actual forming of the photoresist is recorded in real time, indirectly reflecting the impact of the uniformity of the photoresist thickness on the photolithography precision; Related links: Install a flow sensor on the etching equipment to collect etching gas flow data. The etching gas flow affects the etching rate and is indirectly related to the change of photoresist thickness during the etching process.

[0058] Step 2: Data preprocessing: Perform preliminary cleaning on the collected data, remove obvious erroneous values, fill in a small number of missing values, standardize the data, unify the data format, and obtain preprocessed data;

[0059] Step 3: Fluctuation Feature Extraction: Receive pre-processed data, extract the fluctuation period and amplitude characteristics of the exposure time and photoresist thickness changes in the lithography process, decompose the data into different frequency components through wavelet transform technology, and identify high-frequency noise fluctuations and low-frequency trend fluctuations; please refer to Figure 2 , where the specific execution steps are as follows:

[0060] Step S31, data preparation: collect the exposure time and photoresist thickness data of the photolithography step in step 2 for a period of time, so as to obtain continuous Exposure time series at each time point and the photoresist thickness sequence ;

[0061] Step S32: Time series analysis algorithm:

[0062] Periodic analysis: Use Fourier transform to extract periodic features. The Fourier transform formula is: ,in is the original time series, i.e., the exposure time and photoresist thickness , It is the transformed frequency domain representation, which is obtained by exposing the time series Perform Fourier transform to obtain its amplitude at different frequencies. In the frequency domain, the period corresponding to the frequency component with larger amplitude is the fluctuation period of exposure time. The same principle is applied to the photoresist thickness sequence. , find the fluctuation period;

[0063] For example: After calculation, the frequency The corresponding amplitude is larger, so its corresponding period This is an important fluctuation period of exposure time. The same method can be applied to the photoresist thickness sequence , find its fluctuation period.

[0064] Amplitude feature extraction: Extract the fluctuation amplitude feature and calculate the range of the time series, that is, for the exposure time series , calculate the range Indicates the maximum fluctuation range of exposure time within the observation period; for the photoresist thickness sequence , very poor Indicates the fluctuation amplitude of photoresist thickness;

[0065] Step 33: Decompose the frequency components by wavelet transform: Use discrete wavelet transform to process the exposure time and photoresist thickness data. Discrete wavelet transform selects appropriate wavelet basis function ,in is the scale parameter, is the translation parameter, for the exposure time series Perform discrete wavelet transform , we get the wavelet coefficients at different scales and positions. At high-frequency scales, the wavelet coefficients reflect the fast-changing part in the exposure time series, that is, the high-frequency noise fluctuation. Similarly, for the photoresist thickness series Perform discrete wavelet transform , separate its high-frequency noise fluctuations and low-frequency trend fluctuations, that is, fully grasp the fluctuation characteristics of the lithography process parameter data, and provide an accurate data basis for the subsequent establishment of the fluctuation model; for example, at scale At high frequency scales, the wavelet coefficients with larger absolute values ​​and more dispersed distribution correspond to the high-frequency noise in the exposure time. At low frequency scales, the wavelet coefficients reflect the trend fluctuation of the data.

[0066] Step 4. Establish a fluctuation model: Based on the extracted fluctuation characteristics and combined with the principles of semiconductor manufacturing process, a process fluctuation prediction model is established. Specifically, based on historical data and physical models, a mathematical model is constructed between the fluctuation of photoresist thickness and exposure energy and temperature in the photolithography process. This model simulates the fluctuation of process parameters under different conditions; please refer to Figure 3 , where the specific steps for establishing the volatility model are as follows:

[0067] Step 41: Data collation and preparation: Collect historical data from the photolithography process, including the photoresist thickness measurements during the production of different batches of products, the corresponding exposure energy settings, and the ambient temperature or temperature data of key parts of the equipment. Set data and record the photoresist thickness as , the exposure energy is recorded as , the temperature is recorded as ;

[0068] Step 42: Modeling is performed using a multiple linear regression algorithm. Based on the principle of photolithography, the thickness of the photoresist is affected by the combined influence of exposure energy and temperature. The multiple linear regression algorithm is used for modeling, specifically: ,in is the dependent variable, i.e., the photoresist thickness, and , is the intercept, and is the regression coefficient, is a random error term, where 、 and Estimates based on historical data;

[0069] Step 43, parameter estimation: use the least square method to estimate the regression coefficient so that the photoresist thickness and the model prediction value The sum of squared errors between Minimum; through About 、 and Find the partial derivatives and set them to zero to obtain a system of equations:

[0070]

[0071] Solving the above system of equations, we get 、 and Estimated value of 、 and The final photoresist thickness fluctuation prediction model is ,in is the predicted value, is the exposure energy, is temperature;

[0072] Step 44: Model Testing and Optimization: Using the Coefficient of Determination The photoresist thickness fluctuation prediction model is evaluated, where ,in yes The average value of The closer it is to 1, the better the model fits the data.

[0073] Step 5. Formulate and implement fluctuation suppression strategies: Based on the fluctuation prediction model, formulate corresponding fluctuation suppression strategies. That is, when it is predicted that the photoresist thickness will fluctuate significantly, automatically adjust the exposure energy and temperature control parameters. Through the feedback control system, intervene in the manufacturing process in real time to suppress the fluctuation of process parameters and ensure the stability of the semiconductor manufacturing process. Figure 4 , the specific working logic is as follows:

[0074] Threshold setting: Based on historical data and the quality standards of the photolithography process, the normal fluctuation range of the photoresist thickness is determined, and the average normal photoresist thickness is set to , the photoresist thickness fluctuation range is ,in A reasonable fluctuation threshold is set according to the process accuracy requirements. For example, after a large amount of historical data statistical analysis, the average thickness of the photoresist in a certain photolithography process is 500nm. Combined with the process requirements, the threshold is set. , the normal fluctuation range is ;

[0075] Prediction and judgment: Use the established photoresist thickness fluctuation prediction model , input the current exposure energy and temperature data in real time, predict the photoresist thickness, when the predicted value Extraordinary fluctuation range When , it is determined that the photoresist thickness will fluctuate greatly;

[0076] Strategy formulation: Use PID control algorithm to formulate control strategy, specifically, its output By the proportional term , integral item and the differential term , the formula is: , is the error at the current moment, that is, the average value of normal photoresist thickness and predicted value The difference , 、 and They are proportional coefficient, integral coefficient and differential coefficient respectively, among which the proportional term According to the size of the current error, the control amount is adjusted immediately; the integral term Used to eliminate the steady-state error of the system; differential term Adjust the control amount in advance according to the changing trend of the error; the above coefficients are specifically improved according to the characteristics of the photolithography process and the response of the equipment. The improvement is based on the deviation degree and change rate of the predicted photoresist thickness from the normal range, and the coefficients are dynamically adjusted;

[0077] What needs to be specified is:

[0078] Proportional term effect: For example, if the predicted value Above the upper limit , the proportional term will adjust the exposure energy and temperature control parameters in the direction of reducing the photoresist thickness, and the adjustment range is proportional to the error size;

[0079] Function of the integral term: As the integral term accumulates over time, it will continue to accumulate errors, prompting continuous adjustment of the control parameter until the error is zero. For example, in the photolithography process, if there are some minor interference factors that continuously affect the thickness of the photoresist, the integral term can gradually correct the control parameters to return the photoresist thickness to the normal range;

[0080] Function of the differential term: When the photoresist thickness is predicted to deviate rapidly from the normal range, the differential term will quickly increase or decrease the control amount to suppress this trend. If it is a positive and large value, it means that the photoresist thickness is increasing rapidly. The differential term will increase the adjustment of exposure energy and temperature control parameters, slowing down the rate of increase of photoresist thickness.

[0081] Considering multi-factor interference compensation: the thickness of the photoresist is not only affected by exposure energy and temperature, but also by the characteristics of the photoresist itself, equipment aging, and environmental humidity. Introducing interference compensation terms into the PID algorithm, a multi-factor interference model is established; collecting data on the impact of different factors on the thickness of the photoresist, and analyzing the relationship between each factor and the change in the photoresist thickness. Specifically, let the interference factor set be , the corresponding compensation coefficient is , then the control quantity after compensation is ,in is the control quantity after dynamic adjustment of the coefficient;

[0082] Strategy execution: Control quantity calculated according to PID algorithm , automatically adjusting the control parameters of the exposure equipment's exposure energy control device and temperature control device. For example, if the system calculates that the exposure energy needs to be reduced, it will send a command to the exposure equipment's energy regulation module to reduce the output exposure energy. For temperature control, if the temperature needs to be increased, the system will control the heating device to increase power, raising the temperature of the lithography environment or key parts of the equipment. Through real-time feedback control, the photoresist thickness is always maintained within the normal fluctuation range, ensuring the stability of the semiconductor manufacturing process.

[0083] What needs to be specified is 、 and Specific improvements are made based on the characteristics of the lithography process and the response of the equipment. The specific implementation steps are as follows:

[0084] Step A, real-time data acquisition: During the photolithography process, the photoresist thickness data is collected in real time through the sensor. The photoresist thickness is measured as , the normal photoresist thickness range is , and obtain the exposure energy at the corresponding moment and temperature ;

[0085] Step B, calculate the rate of change: calculate the photoresist thickness error Rate of change ,in is the time interval for data collection, and the error change rate, i.e., the change trend of the photoresist thickness error, is obtained by calculating the error values ​​at adjacent moments;

[0086] Step C: Construct coefficient adjustment function:

[0087] Determine the function form: Use multiple linear regression to construct the coefficient adjustment function, where the proportional coefficient The adjustment function is , let the function form be ,in 、 and is the coefficient to be determined; similarly, the integral coefficient The adjustment function is set to , differential coefficient The adjustment function is set to ;

[0088] Coefficient training: Through the preset experimental data, the coefficients in the above function are determined, and several sets of experimental data under different photolithography process conditions are prepared. Each set of data contains the measured value of the photoresist thickness and the corresponding error. , error change rate and the best performing 、 and value, use the least square method to solve the coefficient; for Solve the adjustment function coefficients, assuming that the experimental data are group, target , respectively 、 and Taking the partial derivatives and setting them equal to zero, we obtain the system of equations:

[0089]

[0090] Solving this system of equations, we get 、 and The value of The adjustment function of and Adjust the coefficients in the function 、 、 and 、 、 ;

[0091] Step D, dynamic coefficient update: calculate the error , error change rate , substitute into the determined coefficient adjustment function to calculate the adjusted coefficient to be used at the current moment 、 and , substitute the adjusted coefficient into the PID control algorithm formula , get the control quantity after dynamic adjustment coefficient , calculate the control quantity .

[0092] The system is based on the control quantity obtained by calculation , automatically adjusting exposure energy and temperature control parameters, such as controlling the energy output device and temperature regulation device of the exposure equipment, to achieve precise control of the photoresist thickness. During the subsequent photolithography process, the above steps of data collection, coefficient calculation, and control adjustment are continuously repeated, and the coefficients of the PID control algorithm are continuously and dynamically adjusted according to the actual photoresist thickness, ensuring that the photoresist thickness remains stable within the normal range.

[0093] Step 6. Data storage: Store the preprocessed data in step 2 and the fluctuation suppression strategy data in step 5 in a distributed database, and classify and store them according to different process links, equipment types, and time series for quick query and call.

[0094] Example 2

[0095] See also Figure 5 , an industrial big data governance system, implemented according to an industrial big data governance method, the system includes:

[0096] The data acquisition module collects process parameters and equipment operating status data in the manufacturing process in real time through sensors, and cleans the collected data of errors, fills in missing values, and standardizes the data format to obtain pre-processed data;

[0097] The semiconductor process fluctuation suppression module uses a time series algorithm to process process parameter data, using wavelet transforms to distinguish high-frequency noise from low-frequency trend fluctuations. It then constructs a prediction model based on the fluctuation characteristics and process principles, namely a mathematical model of photoresist thickness, exposure energy, and temperature. Based on the model predictions, it automatically adjusts the exposure energy parameters to stabilize the process.

[0098] The data storage module is used to store the pre-processed data in the acquisition module and the fluctuation suppression strategy data in the semiconductor process fluctuation suppression module in a distributed database, and classify and store them according to different process links, equipment types and time series.

[0099] This invention focuses on the field of industrial big data in semiconductor manufacturing, aims to solve the industry's difficulties in data governance and process control, and proposes an innovative industrial big data governance method and system.

[0100] The data processing process begins with data acquisition. By installing sensors for temperature, pressure, and lithography accuracy on various semiconductor manufacturing equipment, we collect real-time data on process parameters and equipment operating status during the manufacturing process, ensuring comprehensive and up-to-date data. The collected data undergoes preliminary cleaning, missing value filling, and standardization, resulting in a unified format to provide a high-quality data foundation for subsequent analysis.

[0101] For the photolithography process, we utilize time series analysis algorithms and wavelet transform techniques to extract the fluctuation period and amplitude characteristics of data such as exposure time and photoresist thickness, identify high-frequency noise and low-frequency trend fluctuations, and provide accurate data support for establishing a fluctuation model. Based on historical data and process principles, we use a multivariate linear regression algorithm to construct a mathematical model linking photoresist thickness with exposure energy and temperature. Parameters are estimated using the least squares method, and the coefficient of determination is used to evaluate and optimize the model, accurately simulating fluctuations in process parameters under different conditions.

[0102] To mitigate photoresist thickness fluctuations, a strategy was developed based on a fluctuation prediction model. A threshold for normal fluctuations was first set, and the model was used to predict photoresist thickness in real time. When the predicted value exceeded this range, an improved PID control algorithm was used to adjust exposure energy and temperature control parameters. This algorithm dynamically adjusts the coefficient based on the degree of deviation and rate of change of the predicted photoresist thickness from the normal range. It also considers and compensates for interference from multiple factors, including photoresist characteristics, equipment aging, and ambient humidity, ensuring stable photoresist thickness and safeguarding the stability of the semiconductor manufacturing process.

[0103] In addition, the present invention classifies and stores the pre-processed data and the fluctuation suppression strategy data in a distributed database, which is convenient for query and call.

[0104] Through a complete data governance process, the present invention effectively improves the data quality in the semiconductor manufacturing process, accurately controls key process parameters such as photoresist thickness, enhances process stability, improves product quality and production efficiency, and provides important technical support for the intelligent development of the semiconductor manufacturing industry. It has significant innovation and practicality.

[0105] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for industrial big data management, characterized in that: The following steps are involved: Step 1: Data collection: By installing temperature sensors, pressure sensors, and photolithography precision sensors on semiconductor manufacturing equipment, process parameter data during the manufacturing process can be collected in real time. Step 2: Data preprocessing: Perform preliminary cleaning on the collected data, remove obvious erroneous values, fill in a small number of missing values, standardize the data, unify the data format, and obtain preprocessed data; Step 3: Fluctuation feature extraction: Receive pre-processed data and extract the fluctuation period and amplitude characteristics of the exposure time and photoresist thickness changes in the lithography process. Use wavelet transform technology to decompose the data into different frequency components to identify high-frequency noise fluctuations and low-frequency trend fluctuations. Step 4: Establish a fluctuation model: Based on the extracted fluctuation characteristics and combined with the principles of semiconductor manufacturing process, a process fluctuation prediction model is established. Specifically, based on historical data and physical models, a mathematical model is constructed between the fluctuation of photoresist thickness and exposure energy and temperature in the photolithography process. This model simulates the fluctuation of process parameters under different conditions. Step 5: Formulate and implement a fluctuation suppression strategy: Based on the fluctuation prediction model, formulate a corresponding fluctuation suppression strategy. That is, when a large fluctuation in photoresist thickness is predicted, the exposure energy and temperature control parameters are automatically adjusted. The feedback control system intervenes in the manufacturing process in real time to suppress fluctuations in process parameters and ensure the stability of the semiconductor manufacturing process. The specific working logic for the formulation and implementation of volatility suppression strategies is as follows: Threshold setting: Based on historical data and the quality standards of the photolithography process, the normal fluctuation range of the photoresist thickness is determined, and the average normal photoresist thickness is set to , the photoresist thickness fluctuation range is ,in A reasonable fluctuation threshold set according to process accuracy requirements; Prediction and judgment: Use the established photoresist thickness fluctuation prediction model , input the current exposure energy and temperature data in real time, predict the photoresist thickness, when the predicted value Extraordinary fluctuation range When , it is determined that the photoresist thickness will fluctuate greatly; Strategy formulation: Use PID control algorithm to formulate control strategy, specifically, its output By the proportional term , integral item and the differential term , the formula is: , is the error at the current moment, that is, the average value of normal photoresist thickness and predicted value The difference , 、 and They are proportional coefficient, integral coefficient and differential coefficient respectively, among which the proportional term According to the size of the current error, the control amount is adjusted immediately; the integral term Used to eliminate the steady-state error of the system; differential term Adjust the control amount in advance according to the changing trend of the error; The above coefficients are specifically improved according to the characteristics of the photolithography process and the response of the equipment. The improvement is based on the deviation degree and change rate of the predicted photoresist thickness from the normal range, and the coefficients are dynamically adjusted; Consider multi-factor interference compensation: introduce interference compensation terms into the PID algorithm and establish a multi-factor interference model; collect data on the impact of different factors on the photoresist thickness, and analyze the relationship between each factor and the change in photoresist thickness. Specifically, assume that the interference factor set is , the corresponding compensation coefficient is , then the control quantity after compensation is ,in is the control quantity after dynamic adjustment of the coefficient; Strategy execution: Control quantity calculated according to PID algorithm , automatically adjust the exposure energy control device of the exposure equipment and the control parameters of the temperature regulating equipment; described 、 and Specific improvements are made based on the characteristics of the lithography process and the response of the equipment. The specific implementation steps are as follows: Step A, real-time data acquisition: During the photolithography process, the photoresist thickness data is collected in real time through the sensor. The photoresist thickness is measured as , the normal photoresist thickness range is , and obtain the exposure energy at the corresponding moment and temperature ; Step B, calculate the rate of change: calculate the photoresist thickness error Rate of change ,in is the time interval for data collection, and the error change rate, i.e., the change trend of the photoresist thickness error, is obtained by calculating the error values ​​at adjacent moments; Step C: Construct coefficient adjustment function: Determine the function form: Use multiple linear regression to construct the coefficient adjustment function, where the proportional coefficient The adjustment function is , let the function form be ,in 、 and is the coefficient to be determined; similarly, the integral coefficient The adjustment function is set to , differential coefficient The adjustment function is set to ; Coefficient training: Through the preset experimental data, the coefficients in the above function are determined, and several sets of experimental data under different photolithography process conditions are prepared. Each set of data contains the measured value of the photoresist thickness and the corresponding error. , error change rate and the best performing 、 and value, use the least square method to solve the coefficient; for Solve the adjustment function coefficients, assuming that the experimental data are group, target , respectively 、 and Taking the partial derivatives and setting them equal to zero, we obtain the system of equations: Solving this system of equations, we get 、 and The value of The adjustment function of and Adjust the coefficients in the function 、 、 and 、 、 ; Step D, dynamic coefficient update: calculate the error , error change rate , substitute into the determined coefficient adjustment function to calculate the adjusted coefficient to be used at the current moment 、 and , substitute the adjusted coefficient into the PID control algorithm formula , get the control quantity after dynamic adjustment coefficient , calculate the control quantity ; Step 6. Data storage: Store the preprocessed data in step 2 and the fluctuation suppression strategy data in step 5 in a distributed database, and classify and store them according to different process links, equipment types, and time series for quick query and call.

2. The industrial big data management method according to claim 1, characterized in that: The process parameter data in step 1 specifically include the following: photolithography link: installing a temperature sensor near the photoresist coating area of ​​the photolithography equipment to monitor the ambient temperature of the photoresist during the photolithography process; adding a pressure sensor to the photoresist coating equipment to collect pressure data applied to the photoresist during coating; through a high-precision photolithography precision sensor, the accuracy deviation between the photolithography pattern and the actual photoresist molding is recorded in real time; related link: installing a flow sensor on the etching equipment to collect etching gas flow data, as the etching gas flow affects the etching rate.

3. The industrial big data management method according to claim 1, characterized in that: The specific steps of extracting the fluctuation characteristics are as follows: Step S31, data preparation: collect the exposure time and photoresist thickness data of the photolithography step in step 2 for a period of time, so as to obtain continuous Exposure time series at each time point and the photoresist thickness sequence ; Step S32: Time series analysis algorithm: Periodic analysis: Use Fourier transform to extract periodic features. The Fourier transform formula is: ,in is the original time series, i.e., the exposure time and photoresist thickness , It is the transformed frequency domain representation, which is obtained by exposing the time series Perform Fourier transform to obtain its amplitude at different frequencies. In the frequency domain, the period corresponding to the frequency component with larger amplitude is the fluctuation period of exposure time. The same principle applies to the photoresist thickness sequence , find the fluctuation period; Amplitude feature extraction: Extract the amplitude characteristics of the fluctuation and calculate the range of the time series, that is, for the exposure time series , calculate the range Indicates the maximum fluctuation range of exposure time within the observation period; for photoresist thickness series , very poor Indicates the fluctuation amplitude of photoresist thickness; Step 33: Decompose the frequency components by wavelet transform: Use discrete wavelet transform to process the exposure time and photoresist thickness data. Discrete wavelet transform selects appropriate wavelet basis function ,in is the scale parameter, is the translation parameter, for the exposure time series Perform discrete wavelet transform , we get the wavelet coefficients at different scales and positions. At high-frequency scales, the wavelet coefficients reflect the fast-changing part in the exposure time series, that is, the high-frequency noise fluctuation. Similarly, for the photoresist thickness series Perform discrete wavelet transform , separate its high-frequency noise fluctuations and low-frequency trend fluctuations, that is, fully grasp the fluctuation characteristics of the lithography process parameter data, and provide an accurate data basis for the subsequent establishment of the fluctuation model.

4. The industrial big data management method according to claim 1, characterized in that: The specific steps for establishing the fluctuation model are as follows: Step 41: Data collation and preparation: Collect historical data from the photolithography process, including the photoresist thickness measurements during the production of different batches of products, the corresponding exposure energy settings, and the ambient temperature or temperature data of key parts of the equipment. Set data and record the photoresist thickness as , the exposure energy is recorded as , the temperature is recorded as ; Step 42: Use multiple linear regression algorithm to build a model; Based on the principle of photolithography, the thickness of photoresist is affected by the combined effect of exposure energy and temperature. The model is built using a multivariate linear regression algorithm, specifically: ,in is the dependent variable, i.e., the photoresist thickness, and , is the intercept, and is the regression coefficient, is a random error term, where 、 and Estimates based on historical data; Step 43, parameter estimation: use the least square method to estimate the regression coefficient so that the photoresist thickness and the model prediction value The sum of squared errors between Minimum; through About 、 and Find the partial derivatives and set them to zero to obtain a system of equations: Solving the above system of equations, we get 、 and Estimated value of 、 and The final photoresist thickness fluctuation prediction model is ,in is the predicted value, is the exposure energy, is temperature; Step 44: Model Testing and Optimization: Using the Coefficient of Determination The photoresist thickness fluctuation prediction model is evaluated, where ,in yes The average value of The closer it is to 1, the better the model fits the data.

5. An industrial big data governance system, characterized by: According to the industrial big data governance method according to any one of claims 1 to 4, the system comprises: The data acquisition module collects process parameters and equipment operating status data in the manufacturing process in real time through sensors, and cleans the collected data of errors, fills in missing values, and standardizes the data format to obtain pre-processed data; The semiconductor process fluctuation suppression module uses a time series algorithm to process process parameter data, using wavelet transforms to distinguish high-frequency noise from low-frequency trend fluctuations. It then constructs a prediction model based on the fluctuation characteristics and process principles, namely a mathematical model of photoresist thickness, exposure energy, and temperature. Based on the model predictions, it automatically adjusts the exposure energy parameters to stabilize the process. The data storage module is used to store the pre-processed data in the acquisition module and the fluctuation suppression strategy data in the semiconductor process fluctuation suppression module in a distributed database, and classify and store them according to different process links, equipment types and time series.

Images