Equipment debugging management system and method based on big data

Abstract

The invention relates to the technical field of equipment debugging management. In particular to an equipment debugging management system and method based on big data. The equipment debugging management system comprises a data acquisition module, a data analysis module, an equipment adjustment management module and an alarm reminding module. The data acquisition module is used for acquiring historical data information of parts and historical data of wafer cleanliness through background data information of wafer cleaning equipment; the data analysis module is used for judging an abnormal cleaning process by constructing an abnormal detection model and generating an abnormal cleaning process sequence; the equipment adjustment management module is used for adjusting parts of the wafer cleaning equipment according to the abnormal cleaning process sequence and detecting the cleanliness of the adjusted wafer; and the alarm reminding module is used for adjusting the operating parameters of the abnormal parts or replacing the parts when the abnormal cleaning process is judged to give an alarm. By detecting and adjusting the abnormal parts, the replacement and debugging frequency of the parts is reduced, and the production efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment commissioning management, and specifically to a big data-based equipment commissioning management system and method. Background Technique

[0002] With the booming development of the global semiconductor industry and the continuous progress of semiconductor manufacturing processes, wafer cleaning equipment is widely used in the semiconductor, power electronics, and optoelectronics industries, and the requirements for cleaning efficiency and energy consumption are getting higher and higher; during the operation of wafer cleaning equipment, components will experience normal wear due to long-term use and friction, etc. For example, components such as nozzles will experience wear due to physical friction during long-term high-speed spraying and flushing processes. This wear will gradually reduce the performance of the components, such as spraying pressure, flow rate, etc., and may also cause problems such as leakage and blockage during the operation of the equipment, which will affect the cleaning effect; in order to ensure the normal operation and production efficiency of the equipment, it is necessary to regularly inspect and maintain the equipment, and replace damaged or aged components in a timely manner according to the actual situation.

[0003] Under the existing technology, wafer cleaning equipment is usually equipped with a real-time monitoring system that can detect the operating status and various parameters of the equipment. When an abnormality occurs in a component, the system will automatically alarm. However, the sensors in the monitoring system may be damaged or aged after long-term use, resulting in a decrease in its detection accuracy or complete failure; and when the alarm device issues an alarm, it is difficult for the operator to directly identify the faulty component and replace it with a new or repaired component, which increases the maintenance cost, and frequent replacement of components may also affect the overall stability and lifespan of the equipment. Summary of the Invention

[0004] The purpose of the present invention is to provide a big data-based equipment commissioning management system and method to solve the problems raised in the above background technique.

[0005] To solve the above technical problems, the present invention provides the following technical solution: A big data-based equipment commissioning management method, the equipment commissioning management method specifically includes the following steps:

[0006] S100. Regularize and sort out the background data information obtained from the wafer cleaning equipment through system logs, and construct a historical cleaning database for the wafer cleaning equipment. The historical cleaning database includes: cleaning procedures, historical data information of components of the wafer cleaning equipment, and historical data of wafer cleaning. Among them, the historical cleaning database comes from the background data information obtained from the system logs of the wafer cleaning equipment, and the historical data information of the components of the wafer cleaning equipment and the historical data of wafer cleaning are obtained based on the historical cleaning database. The historical data information of the components includes historical operating parameters, historical maintenance records and reports. The historical data of wafer cleaning is the cleanliness of the wafers sampled and inspected after each cleaning process in the cleaning process group of the wafers, and the final inspection data of the cleanliness of the wafers after the cleaning process group ends (the cleanliness of the wafers after each cleaning process is usually sampled inspection data, the same wafers are sampled and inspected after each cleaning process, and the final inspection data of the cleanliness of the wafers after the cleaning process group ends is the full inspection data of the wafers).

[0007] It should be noted that the cleaning procedures (managed by numbers) are different cleaning process groups formulated according to customer requirements. The cleaning processes include: chemical components of the cleaning solution, temperature of the cleaning solution, usage amount of the cleaning solution, circulation amount of the cleaning solution, cleaning time, etc. The cleaning processes in all cleaning tanks form a cleaning process group, and the corresponding components are different for different cleaning process groups. The cleanliness of the wafers usually refers to the size and quantity of impurities (such as particles) on the wafers after a certain cleaning process. The wafer cleanliness is usually obtained by detecting the wafers passing through each cleaning process group with a detection device (such as a vision detection device). The wafer cleanliness is defined as ten grades from 1 to 10. The higher the wafer cleanliness, the fewer impurities on the wafers. A large number of wafers are cleaned simultaneously each time. The components of the wafer cleaning equipment refer to components such as nozzles, heaters, and sensors that can adjust operating parameters on site.

[0008] S200. Determine whether the cleaning process group in the cleaning procedure of the wafer cleaning equipment is abnormal based on the historical data of wafer cleaning, and give an alarm reminder when the cleaning process group is abnormal.

[0009] S300. Determine which component in the abnormal cleaning process group of the wafer cleaning equipment is abnormal based on the historical cleaning database of the wafer cleaning equipment.

[0010] S400 adjusts the operating parameters of the abnormal components determined in S300 or replaces the components. After the adjustment or replacement, the wafer cleanliness of the wafers produced by the equipment is detected, and the determination of whether there is an abnormality is carried out according to the method in the steps of S200. When the determination is non-abnormal, the adjusted operating parameters are written into the cleaning program to replace the operating parameters of the corresponding components in the original cleaning program. The adjustment of the operating parameters of the abnormal components or the replacement of the components need to be recorded to facilitate the subsequent maintenance personnel to understand the current operating state of the equipment and the previous adjustments. If the determination of whether there is an abnormality is carried out according to the method in the steps of S200 and the determination is abnormal, then further adjustment of the operating parameters of the components or replacement of the components is required until the determination is non-abnormal.

[0011] Further, the specific method for obtaining the wafer cleaning historical data in S100 includes:

[0012] Obtain the background data information of the wafer cleaning equipment through the system log and regularize and sort it to construct the wafer cleaning equipment historical cleaning database. Obtain the wafer cleaning historical data based on the wafer cleaning equipment historical cleaning database. Among them, the historical data information of the components in the wafer cleaning equipment screened from the wafer cleaning equipment historical cleaning database refers to the historical data information of the components in the wafer cleaning equipment that can be adjusted in operating parameters and replaced. After sorting, it is recorded as the set Wr, Wr = {Wr1, Wr2, Wr3... Wrf... WrF}; where Wr represents the historical detection data set of the cleanliness of the wafer in the r-th cleaning process group; Wrf represents the historical spot-check data of the wafer passing through the f-th cleaning process in the r-th cleaning process group, r = 0, 1, 2, 3... R, and R represents the total number of cleaning process groups included in the cleaning program in the wafer cleaning equipment, and R is a constant; where when r = 0, W0 represents the historical detection data of the cleanliness corresponding to the wafer before entering the cleaning program in the wafer cleaning equipment; f = 1, 2, 3... F, and F represents the total number of cleaning processes included in the cleaning process group, and F is a constant; where r and f are only the numbers of the cleaning process groups and cleaning processes in the wafer cleaning equipment.

[0013] Further, the specific method for determining whether there is an abnormality in the cleaning process group to be judged in the wafer cleaning equipment in S200 includes: (The cleaning process group to be judged is usually the current cleaning process group)

[0014] S201. Build an anomaly detection model. Whenever a new wafer cleaning data is generated after a new cleaning process group is completed, it is recorded as data Wr1 (that is, when the cleaning process group and wafer cleanliness data are recorded, this record is usually also entered into the historical cleaning database at the same time). Then, compare it with the historical cleaning database with the same cleaning process group and the most recent occurrence time in the historical cleaning database, which is recorded as data Wr, to determine the similarity degree of the final detection data of the wafer cleanliness in data Wr1 and data Wr, so as to determine whether this cleaning process group is abnormal. Determine whether the cleaning process group is abnormal through the anomaly detection model. The determination formula is:

[0015] K[Wr, Wr1] = {∑ F f＝1 [|Wrf’ ∩ Wr1f’| / |Wrf’ ∪ Wr1f’|]} / F

[0016] The determination formula is used to calculate the anomaly degree of the wafer cleanliness of the cleaning process group in the wafer cleaning equipment; K[Wr, Wr1] represents the correlation degree of the two detection data sets of the cleaning process group r of Wr and Wr1; Wr1 represents the final detection data set of the cleanliness of the newly generated wafers in the rth cleaning process group of the wafer cleaning equipment; Wr represents the historical final detection data set of the cleanliness that is the same as the cleaning process group r of Wr1 in the wafer cleaning equipment and has the most recent occurrence time; Wrf’ represents the sampling detection data set of the wafer cleanliness of any cleaning process f’ in the rth cleaning process group of the wafer cleaning equipment, and Wr1f’ represents the sampling detection data set of the cleanliness of the newly generated wafers of any cleaning process f’ in the rth cleaning process group of the wafer cleaning equipment;

[0017] Among them, |Wrf’ ∩ Wr1f’| represents the number of wafers with the same wafer cleanliness in any cleaning process f’ in the cleaning process group r, and |Wrf’ ∪ Wr1f’| represents the total number of wafers in the sampling detection data set of any cleaning process f’ in the cleaning process group r;

[0018] S202. For the cleaning process group r that needs to determine whether it is abnormal, when K[Wr, Wr1] ≥ Q, it means that the difference in the historical final detection data set of the cleanliness with the same cleaning process group r and the most recent occurrence time is small. When K[Wr, Wr1] < Q, it means that the difference in the historical final detection data set of the cleanliness with the same cleaning process group r and the most recent occurrence time is large. Determine that the cleaning process group r is abnormal and give an alarm reminder. Q represents the difference degree threshold of the cleaning process group.

[0019] Further, the specific method for determining which component in the abnormal cleaning process group of the wafer cleaning equipment in S300 includes: By the formula: prf’ = (|Wrf’ ∩ Wr1f’| / |Wrf’ ∪ Wr1f’|), determine the abnormal score of the wafer cleanliness of any cleaning process f’ in the abnormal cleaning process group r. prf’ represents the abnormal score of the wafer cleanliness of any cleaning process f’ in the abnormal cleaning process group r; the higher the score, the lower the abnormal degree of the cleaning process in the abnormal cleaning process group r.

[0020] Construct a comprehensive scoring model based on the cleaning process component data information in the historical cleaning database: Prf’ = prf’ / (Vrf’ / Trf’), and calculate the comprehensive abnormal score of the wafer cleanliness of any cleaning process f’ in the abnormal cleaning process group r as Prf’; Trf’ and Vrf’ respectively represent the component replacement frequency and usage duration of any cleaning process f’ in the cleaning process group r; Sort the cleaning processes in the cleaning process group from high to low according to the abnormal comprehensive score to generate an abnormal cleaning process sequence. By analyzing the abnormal comprehensive score of the cleaning processes in the abnormal cleaning process group, it avoids the misdetection of the abnormal cleaning process group due to the error of the detection equipment, reduces the misdetection risk, and improves the abnormal recognition rate.

[0021] Further, the specific method for adjusting the operating parameters of the abnormal component or replacing the component in S400 is as follows: Adjust the operating parameters of the component or replace the component according to the sorting result of the abnormal cleaning process sequence, and then judge according to the steps of S200. When it is determined to be non-abnormal, write the adjusted operating parameters into the cleaning program to replace the operating parameters of the corresponding component in the original cleaning program. Record the parameters adjusted for the component or the replacement of the component, so that the subsequent maintenance personnel can understand the current operating state of the equipment and the previous adjustments, and provide a reliable basis for adjusting the operating parameters when the component is abnormal next time; The components include nozzles, heaters, and sensors, and the parameters include: cleaning liquid concentration, cleaning liquid flow rate, and temperature. When it is determined to be abnormal, further adjust the operating parameters of the component or replace the component until it is determined to be non-abnormal.

[0022] Further, the device debugging management system includes a data acquisition module, a data analysis module, a device adjustment management module, and an alarm reminder module; the output end of the data acquisition module is connected to the input end of the data analysis module, the output end of the data analysis module is connected to the input end of the device debugging management module, and the output end of the device adjustment management module is connected to the input end of the alarm reminder module; the data acquisition module obtains the historical data information of components and the historical data of wafer cleanliness through the background data information of the wafer cleaning device, and the abnormal situation of the cleaning process group of the wafer cleaning device can be judged through the historical data of wafer cleanliness; the data analysis module constructs an anomaly detection model through the historical data of wafer cleaning obtained by the detection device, analyzes the current process group to complete the determination of the abnormal cleaning process and generates an abnormal cleaning process sequence. Subsequently, the operating parameters of the components of the wafer cleaning device can be adjusted through the abnormal cleaning process sequence to improve the adjustment efficiency of the components of the wafer cleaning device; the device adjustment management module adjusts the components of the wafer cleaning device according to the abnormal cleaning process sequence and detects the cleanliness of the wafer after adjustment, and determines whether the device adjustment is completed through the cleanliness of the wafer detected subsequently; the alarm reminder module is to alarm and remind when an abnormal cleaning process is judged, adjust the operating parameters of the judged abnormal components or replace the components. After the adjustment or replacement, the wafer cleanliness of the wafers produced by the device is detected, and the determination of whether it is abnormal is carried out according to the method of step S200. When the determination is non-abnormal, the adjusted operating parameters are written into the cleaning program to replace the operating parameters of the corresponding components in the original cleaning program. The adjustment of the operating parameters of the abnormal components or the replacement of the components need to be recorded to facilitate the subsequent maintenance personnel to understand the current operating state of the device and the previous adjustments. If the determination of whether it is abnormal is carried out according to the method of step S200 and the determination is non-abnormal, then further adjustment of the operating parameters of the components or replacement of the components is required until the determination is non-abnormal.

[0023] Further, the data acquisition module includes a historical data information acquisition unit and a cleaning historical data acquisition unit; the historical data information acquisition unit acquires historical data information of the components of the wafer cleaning equipment from the historical cleaning database, where the historical data information of the components includes historical operation parameters and historical maintenance records and reports; the cleaning historical data acquisition unit acquires the wafer cleaning historical data from the historical cleaning database, and the wafer cleaning historical data is the cleanliness of the wafers sampled after each cleaning process under the cleaning process group of the wafers, and the final inspection data of the cleanliness of the wafers after the cleaning process group ends (the cleanliness of the wafers after each cleaning process is usually sampled inspection data, and the same wafers are sampled after each cleaning process. The final inspection data of the cleanliness of the wafers after the cleaning process group ends is the full inspection data of the wafers).

[0024] Further, the data analysis module includes an anomaly detection model construction unit, an abnormal cleaning process judgment unit, and an abnormal cleaning process sequence generation unit; the anomaly detection model construction unit constructs an anomaly detection model for the cleaning process group through the wafer cleaning historical data, and through the anomaly detection model of the cleaning process group, it can analyze and judge the abnormal cleaning process group in the wafer cleaning equipment; the abnormal cleaning process judgment unit preliminarily determines the abnormality of the cleaning process group in the wafer cleaning equipment, and then determines the abnormal cleaning process. By performing an abnormal comprehensive score analysis on the cleaning processes in the abnormal cleaning process group, it avoids the misdetection of the abnormal cleaning process group due to the error of the detection equipment, reduces the misdetection risk, and improves the anomaly recognition rate; the abnormal cleaning process sequence generation unit sorts the abnormal cleaning processes according to the abnormal comprehensive score to generate an abnormal cleaning process sequence, and then adjusts the operation parameters of the components or replaces the components according to the sorting result of the abnormal cleaning process sequence.

[0025] Further, the equipment adjustment management module includes a component operation parameter adjustment unit and a wafer cleanliness detection unit; the component operation parameter adjustment unit adjusts the operation parameters of the components or replaces the components according to the sorting result of the abnormal cleaning process sequence; the wafer cleanliness detection unit acquires the detection of the wafer cleanliness after the operation parameters of the components of the wafer cleaning equipment are adjusted.

[0026] Further, the alarm reminder module includes an alarm reminder unit and a data storage and recording unit; the alarm reminder unit gives an alarm reminder when there is an abnormal cleaning process group in the wafer cleaning equipment; the data storage and recording unit, when the operating parameters of the components are adjusted and the cleaning process is determined to be normal, writes the adjusted operating parameters into the cleaning program, replacing the operating parameters of the corresponding components in the original cleaning program, records the adjusted parameters of the components or the replacement of the components, facilitating subsequent maintenance personnel to understand the current operating state of the equipment and the previous adjustments, and providing a reliable basis for adjusting the operating parameters when the components are abnormal next time.

[0027] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The present invention analyzes and judges the abnormal cleaning process group in the wafer cleaning equipment through the detection data of the wafer cleanliness, determines the abnormal components, and determines whether the components need to be replaced by adjusting the operating parameters of the abnormal components; reduces the maintenance cost, optimizes the working state of the equipment, reduces the replacement rate of components, and in case of emergency, quickly adjusting the parameters can quickly restore the normal operation of the equipment, reduce the production interruption time, and avoid the reduction of work efficiency. Brief Description of the Drawings

[0028] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0029] Figure 1 is a schematic structural diagram of a device debugging management system based on big data according to the present invention;

[0030] Figure 2 is a schematic flow diagram of a device debugging management method based on big data according to the present invention. Detailed Embodiments

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0032] Please refer to Figure 1 and Figure 2 As shown, the present invention provides a technical solution: a device debugging management method based on big data, and the device debugging management method specifically includes the following steps:

[0033] S100. Regularize and sort out the background data information obtained from the wafer cleaning equipment through system logs to construct a historical cleaning database for the wafer cleaning equipment. The historical cleaning database includes: cleaning procedures, historical data information of components of the wafer cleaning equipment, and historical data of wafer cleaning. Among them, the historical cleaning database comes from the background data information obtained from the system logs of the wafer cleaning equipment. The historical data information of the components and the historical data of wafer cleaning are obtained based on the historical cleaning database. The historical data information of the components includes historical operating parameters, historical maintenance records and reports. The historical data of wafer cleaning are the cleanliness of the wafers sampled and inspected after each cleaning process in the cleaning process group of the wafers, and the final inspection data of the cleanliness of the wafers after the cleaning process group ends (the cleanliness of the wafers after each cleaning process is usually sampled inspection data, and the same wafers are sampled and inspected after each cleaning process. The final inspection data of the cleanliness of the wafers after the cleaning process group ends are the full inspection data of the wafers).

[0034] It should be noted that the cleaning procedures (managed by numbers) are different cleaning process groups formulated according to customer requirements. The cleaning processes include: chemical components of the cleaning solution, temperature of the cleaning solution, usage amount of the cleaning solution, circulation amount of the cleaning solution, cleaning time, etc. The cleaning processes in all cleaning tanks form a cleaning process group. Among them, the components corresponding to different cleaning process groups are inconsistent. The cleanliness of the wafers usually refers to the size and quantity of impurities (such as particulate matter) on the wafers after a certain cleaning process. The wafer cleanliness is usually obtained by detecting the wafers passing through each cleaning process group with a detection device (such as a vision detection device). The wafer cleanliness is defined as ten grades from 1 to 10. The higher the wafer cleanliness, the fewer impurities on the wafers. A large number of wafers are cleaned simultaneously each time. The components of the wafer cleaning equipment refer to components such as nozzles, heaters, and sensors that can adjust operating parameters on site.

[0035] S200. Determine whether the cleaning process group in the cleaning procedure of the wafer cleaning equipment is abnormal based on the historical data of wafer cleaning, and give an alarm reminder when the cleaning process group is abnormal.

[0036] S300. Determine which component in the abnormal cleaning process group of the wafer cleaning equipment is abnormal based on the historical cleaning database of the wafer cleaning equipment.

[0037] S400 adjusts the operating parameters of the abnormal components identified in S300 or replaces the components. After the adjustment or replacement, the wafer cleanliness of the wafers produced by the equipment is detected, and whether it is abnormal is determined according to the method in the steps of S200. When it is determined to be non-abnormal, the adjusted operating parameters are written into the cleaning program to replace the operating parameters of the corresponding components in the original cleaning program. Records shall be kept for the adjustment of the operating parameters of the abnormal components or the replacement of the components, so that the subsequent maintenance personnel can understand the current operating state of the equipment and the previous adjustments. If it is determined whether it is abnormal according to the method in the steps of S200 and it is determined to be abnormal, then further adjust the operating parameters of the components or replace the components until it is determined to be non-abnormal.

[0038] Further, the specific method for obtaining the wafer cleaning historical data in S100 includes:

[0039] Obtain the background data information of the wafer cleaning equipment through the system log and regularize and sort it to construct the historical cleaning database of the wafer cleaning equipment. Obtain the wafer cleaning historical data based on the historical cleaning database of the wafer cleaning equipment. Among them, the historical data information of the components in the wafer cleaning equipment screened from the historical cleaning database of the wafer cleaning equipment refers to the historical data information of the components in the wafer cleaning equipment that can adjust the operating parameters and be replaced. After sorting, it is recorded as the set Wr, Wr = {Wr1, Wr2, Wr3... Wrf... WrF}; where Wr represents the historical detection data set of the cleanliness of the wafer in the rth cleaning process group; Wrf represents the historical spot-check data of the wafer passing through the fth cleaning process in the rth cleaning process group, r = 0, 1, 2, 3... R, and R represents the total number of cleaning process groups included in the cleaning program in the wafer cleaning equipment, and R is a constant; where when r = 0, W0 represents the historical detection data of the cleanliness of the wafer before entering the cleaning program in the wafer cleaning equipment; f = 1, 2, 3... F, and F represents the total number of cleaning processes included in the cleaning process group, and F is a constant; where r and f are only the numbers of the cleaning process groups and cleaning processes in the wafer cleaning equipment.

[0040] Further, the specific method for determining whether there is an abnormality in the cleaning process group to be judged in the wafer cleaning equipment in S200 includes: (The cleaning process group to be judged is usually the current cleaning process group)

[0041] S201. Build an anomaly detection model. Whenever a new wafer cleaning data is generated after a new cleaning process group is completed, it is recorded as data Wr1 (that is, when the cleaning process group and wafer cleanliness data are recorded, this record is usually also entered into the historical cleaning database at the same time). Then, compare it with the historical cleaning database with the same cleaning process group and the most recent occurrence time in the historical cleaning database, which is recorded as data Wr, to determine the similarity degree of the final detection data of the wafer cleanliness in data Wr1 and data Wr to determine whether the cleaning process group is abnormal. Determine whether the cleaning process group is abnormal through the anomaly detection model. The determination formula is:

[0042] K[Wr, Wr1] = {∑ F f＝1 [|Wrf’ ∩ Wr1f’| / |Wrf’ ∪ Wr1f’|]} / F

[0043] The determination formula is used to calculate the anomaly degree of the wafer cleanliness of the cleaning process group in the wafer cleaning equipment; K[Wr, Wr1] represents the correlation degree of the two detection data sets of the cleaning process group r of Wr and Wr1; Wr1 represents the final detection data set of the cleanliness of the newly generated wafers in the rth cleaning process group of the wafer cleaning equipment; Wr represents the historical final detection data set of cleanliness that is the same as the cleaning process group r of Wr1 in the wafer cleaning equipment and has the most recent occurrence time; Wrf’ represents the sampling detection data set of the wafer cleanliness of any cleaning process f’ in the rth cleaning process group of the wafer cleaning equipment, and Wr1f’ represents the sampling detection data set of the cleanliness of the newly generated wafers of any cleaning process f’ in the rth cleaning process group of the wafer cleaning equipment;

[0044] Among them, |Wrf’ ∩ Wr1f’| represents the number of wafers with the same wafer cleanliness in any cleaning process f’ in the cleaning process group r, and |Wrf’ ∪ Wr1f’| represents the total number of wafers in the sampling detection data set of any cleaning process f’ in the cleaning process group r;

[0045] S202. For the cleaning process group r that needs to determine whether it is abnormal, when K[Wr, Wr1] ≥ Q, it means that the difference in the historical final detection data set of cleanliness with the same cleaning process group r and the most recent occurrence time is small. When K[Wr, Wr1] < Q, it means that the difference in the historical final detection data set of cleanliness with the same cleaning process group r and the most recent occurrence time is large. Determine that the cleaning process group r is abnormal and give an alarm reminder. Q represents the difference degree threshold of the cleaning process group.

[0046] Further, the specific method for determining which component in the abnormal cleaning process group of the wafer cleaning equipment in S300 includes: By the formula: prf’ = (|Wrf’ ∩ Wr1f’| / |Wrf’ ∪ Wr1f’|), determine the abnormal score of the wafer cleanliness of any cleaning process f’ in the abnormal cleaning process group r. prf’ represents the abnormal score of the wafer cleanliness of any cleaning process f’ in the abnormal cleaning process group r; the higher the score, the lower the abnormal degree of the cleaning process in the abnormal cleaning process group r.

[0047] Construct a comprehensive scoring model based on the cleaning process component data information in the historical cleaning database: Prf’ = prf’ / (Vrf’ / Trf’), and calculate the comprehensive abnormal score of the wafer cleanliness of any cleaning process f’ in the abnormal cleaning process group r as Prf’; Trf’ and Vrf’ respectively represent the component replacement frequency and usage duration of any cleaning process f’ in the cleaning process group r; Sort the cleaning processes in the cleaning process group from high to low according to the abnormal comprehensive score to generate an abnormal cleaning process sequence. By analyzing the abnormal comprehensive score of the cleaning processes in the abnormal cleaning process group, it avoids the misdetection of the abnormal cleaning process group due to the error of the detection equipment, reduces the misdetection risk, and improves the abnormal recognition rate.

[0048] Further, the specific method for adjusting the operating parameters of the abnormal component or replacing the component in S400 is as follows: Adjust the operating parameters of the component or replace the component according to the sorting result of the abnormal cleaning process sequence, and then judge according to the steps of S200. When it is determined to be non-abnormal, write the adjusted operating parameters into the cleaning program to replace the operating parameters of the corresponding component in the original cleaning program. Record the parameters adjusted for the component or the replacement of the component, so that the subsequent maintenance personnel can understand the current operating state of the equipment and the previous adjustments, and provide a reliable basis for adjusting the operating parameters when the component is abnormal next time; The components include spray nozzles, heaters, and sensors, and the parameters include: cleaning liquid concentration, cleaning liquid flow rate, and temperature. When it is determined to be abnormal, further adjust the operating parameters of the component or replace the component until it is determined to be non-abnormal.

[0049] Further, the device debugging management system includes a data acquisition module, a data analysis module, a device adjustment management module, and an alarm reminder module; the output end of the data acquisition module is connected to the input end of the data analysis module, the output end of the data analysis module is connected to the input end of the device debugging management module, and the output end of the device adjustment management module is connected to the input end of the alarm reminder module; the data acquisition module obtains the historical data information of components and the historical data of wafer cleanliness through the background data information of the wafer cleaning equipment, and the abnormal situation of the cleaning process group of the wafer cleaning equipment can be judged through the historical data of wafer cleanliness; the data analysis module constructs an abnormal detection model through the historical data of wafer cleaning obtained by the detection equipment, analyzes the current process group to complete the determination of the abnormal cleaning process and generates an abnormal cleaning process sequence. Subsequently, the operating parameters of the components of the wafer cleaning equipment can be adjusted through the abnormal cleaning process sequence to improve the adjustment efficiency of the components of the wafer cleaning equipment; the device adjustment management module adjusts the components of the wafer cleaning equipment according to the abnormal cleaning process sequence and detects the cleanliness of the wafer after adjustment, and determines whether the equipment adjustment is completed through the cleanliness of the wafer detected subsequently; the alarm reminder module alarms and reminds when an abnormal cleaning process is judged, adjusts the operating parameters of the judged abnormal components or replaces the components. After the adjustment or replacement, the wafer cleanliness of the wafers produced by the equipment is detected, and the determination of whether it is abnormal is carried out according to the method of step S200. When the determination is non-abnormal, the adjusted operating parameters are written into the cleaning program to replace the operating parameters of the corresponding components in the original cleaning program. The adjustment of the operating parameters of the abnormal components or the replacement of the components need to be recorded to facilitate the subsequent maintenance personnel to understand the current operating state of the equipment and the previous adjustments. If the determination of whether it is abnormal is carried out according to the method of step S200 and the determination is non-abnormal, then further, the operating parameters of the components need to be adjusted or the components need to be replaced until the determination is non-abnormal.

[0050] Further, the data acquisition module includes a historical data information acquisition unit and a cleaning historical data acquisition unit; the historical data information acquisition unit acquires the historical data information of the components of the wafer cleaning equipment from the historical cleaning database, where the historical data information of the components includes historical operation parameters and historical maintenance records and reports; the cleaning historical data acquisition unit acquires the wafer cleaning historical data from the historical cleaning database, and the wafer cleaning historical data is the cleanliness of the wafers sampled after each cleaning process under the cleaning process group of the wafers and the final inspection data of the cleanliness of the wafers after the cleaning process group ends (the cleanliness of the wafers after each cleaning process is usually sampled inspection data, the same wafers are sampled after each cleaning process, and the final inspection data of the cleanliness of the wafers after the cleaning process group ends is the full inspection data of the wafers).

[0051] Further, the data analysis module includes an anomaly detection model construction unit, an abnormal cleaning process judgment unit, and an abnormal cleaning process sequence generation unit; the anomaly detection model construction unit constructs an anomaly detection model for the cleaning process group through the wafer cleaning historical data, and through the anomaly detection model of the cleaning process group, it can analyze and judge the abnormal cleaning process group in the wafer cleaning equipment; the abnormal cleaning process judgment unit preliminarily determines the anomaly of the cleaning process group in the wafer cleaning equipment, and then determines the abnormal cleaning process. By performing an abnormal comprehensive score analysis on the cleaning processes in the abnormal cleaning process group, it avoids the misdetection of the abnormal cleaning process group due to the error of the detection equipment, reduces the misdetection risk, and improves the anomaly recognition rate; the abnormal cleaning process sequence generation unit sorts the abnormal cleaning processes according to the abnormal comprehensive score to generate an abnormal cleaning process sequence, and then adjusts the operation parameters of the components or replaces the components according to the sorting result of the abnormal cleaning process sequence.

[0052] Further, the equipment adjustment management module includes a component operation parameter adjustment unit and a wafer cleanliness detection unit; the component operation parameter adjustment unit adjusts the operation parameters of the components or replaces the components according to the sorting result of the abnormal cleaning process sequence; the wafer cleanliness detection unit acquires the detection of the wafer cleanliness after the operation parameters of the components of the wafer cleaning equipment are adjusted.

[0053] Further, the alarm reminder module includes an alarm reminder unit and a data storage and recording unit; the alarm reminder unit gives an alarm reminder when there is an abnormal cleaning process group in the wafer cleaning equipment; the data storage and recording unit, when the operating parameters of the components are adjusted and the cleaning process is determined to be normal, writes the adjusted operating parameters into the cleaning program, replacing the operating parameters of the corresponding components in the original cleaning program, records the adjusted parameters of the components or the replacement of the components, so as to facilitate subsequent maintenance personnel to understand the current operating state of the equipment and the previous adjustments, and provide a reliable basis for adjusting the operating parameters when the components are abnormal next time.

[0054] In this embodiment: the background data information of the wafer cleaning equipment is obtained through the system log and sorted out to construct the historical cleaning database of the wafer cleaning equipment. According to the historical cleaning database of the wafer cleaning equipment, the historical cleaning data of the wafers is obtained, and after sorting, it is recorded as the set Wr, Wr = {Wr1, Wr2, Wr3... Wrf... WrF}; where, Wr represents the historical detection data set of the cleanliness of the wafers in the r-th cleaning process group; Wrf represents the historical spot-check data of the wafers in the r-th cleaning process group after the f-th cleaning process, r = 0, 1, 2, 3... R, R represents the total number of cleaning process groups included in the cleaning program in the wafer cleaning equipment, R = 6; f = 1, 2, 3... F, F = 4.

[0055] S201. Construct an anomaly detection model, and determine whether the cleaning process group is abnormal through the anomaly detection model. The determination formula is:

[0056]

[0057] The determination formula is used to calculate the anomaly degree of the cleanliness of the wafers in the cleaning process group of the wafer cleaning equipment;

[0058] Embodiment 1: When there is a cleaning process group r = 3, such that K[Wr, Wr1] = 0.85 ≥ Q = 0.8, it indicates that the difference in the final detection data set of the cleanliness of the same and most recent occurrence events in the cleaning process group 3 is small, and it is determined that the cleaning process group is normal;

[0059] Embodiment 2: When there is a cleaning process group r = 3, such that K[Wr, Wr1] = 0.6 < Q = 0.8, it indicates that the difference in the final detection data set of the cleanliness of the same and most recent occurrence events in the cleaning process group r is large, and it is determined that the cleaning process group r is abnormal, and an alarm reminder is given. Q represents the difference degree threshold of the cleaning process group.

[0060] Determine the abnormal score of the wafer cleanliness of any cleaning process f' in the abnormal cleaning process group 3 through the formula: prf’ = (|Wrf’ ∩ Wr1f’| / |Wrf’ ∪ Wr1f’|). The calculated abnormal scores of the wafer cleanliness of the cleaning processes in the abnormal cleaning process group 3 are {0.8, 0.74, 0.65, 0.9};

[0061] According to the comprehensive scoring model: Prf’ = prf’ / (Vrf’ / Trf’), calculate that the comprehensive abnormal score of the wafer cleanliness of any cleaning process f' in the abnormal cleaning process group r is Prf’ = {98.4, 66.578, 80}; Trf’ and Vrf’ respectively represent the component replacement frequency and usage duration of any cleaning process f' in the cleaning process group r. Sort the cleaning processes in the cleaning process group from high to low according to the abnormal comprehensive score to generate an abnormal cleaning process sequence. Adjust the operating parameters of the components or replace the components according to the sorting result of the abnormal cleaning process sequence, and then make a judgment according to the steps of S200. When it is determined to be non-abnormal, write the adjusted operating parameters into the cleaning program to replace the operating parameters of the corresponding components in the original cleaning program.

[0062] Inspired by the ideal embodiments of the present application described above, through the above description, relevant staff can make various changes and modifications completely within the scope without departing from the technical idea of this application. The technical scope of this application is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

[0063] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0064] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementing in the process Figure 1 one process or multiple processes and / or blocksFigure 1 means for the functions specified in one or more boxes.

[0065] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one Figure 1 or more processes and / or boxes Figure 1 or more boxes.

[0066] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one Figure 1 or more processes and / or boxes Figure 1 or more boxes.

Claims

1. A device debugging management method based on big data, characterized by: The device debugging management method specifically comprises the following steps: S100, constructing a historical cleaning database of wafer cleaning equipment by regularly sorting out background data information obtained from the system log of the wafer cleaning equipment, and obtaining historical data information of parts of the wafer cleaning equipment and historical cleaning data of wafers according to the historical cleaning database; S200, judging whether a cleaning process group in a cleaning procedure of a wafer cleaning equipment is abnormal based on the wafer cleaning history data, and giving an alarm reminder when an abnormality exists in the cleaning process group; S300, determining which component in the abnormal cleaning process group in the wafer cleaning equipment has an abnormality according to a historical cleaning database of the wafer cleaning equipment; S400, adjust the operating parameters of the abnormal parts determined in S300 or replace the parts. After the adjustment or replacement, perform wafer cleanliness detection on the wafers produced by the equipment, and determine whether there is an abnormality according to the method of step S200. When it is determined to be non-abnormal, write the adjusted operating parameters into the cleaning program to replace the operating parameters of the corresponding parts of the original cleaning program.

2. According to the big data-based equipment debugging management method of claim 1, it is characterized by: The specific method of obtaining the wafer cleaning history data in S100 includes: The background data information of the wafer cleaning equipment is obtained through the system log and is regularly sorted to build a historical cleaning database of the wafer cleaning equipment. The wafer cleaning history data is obtained based on the historical cleaning database of the wafer cleaning equipment, and is recorded as a set Wr after sorting, Wr = {Wr1, Wr2, Wr3...Wrf...WrF}; wherein Wr represents the historical detection data set of the cleanliness of the wafer in the rth cleaning process group; Wrf represents the historical sampling data of the wafer after the fth cleaning process in the rth cleaning process group, r = 0, 1, 2, 3...R, R represents the total amount of cleaning process groups included in the cleaning program of the wafer cleaning equipment, and R is a constant; wherein when r = 0, W0 represents the historical detection data of the cleanliness of the wafer before entering the cleaning program in the wafer cleaning equipment; f = 1, 2, 3...F, F represents the total number of cleaning processes included in the cleaning process group, and F is a constant; wherein r and f are only the numbers of the cleaning process groups and cleaning processes in the wafer cleaning equipment.

3. The device debugging management method based on big data according to claim 2 is characterized by: The specific method for determining whether there is an abnormality in the cleaning process group to be determined in the wafer cleaning equipment in S200 includes: S201, construct an abnormality detection model, and use the abnormality detection model to determine whether the cleaning process group is abnormal. The determination formula is: K[Wr,Wr1]={∑ F f＝1 [|Wrf’∩Wr1f’| / |Wrf’∪Wr1f’|]} / F The determination formula is used to calculate the abnormal degree of wafer cleanliness of the cleaning process group in the wafer cleaning equipment; K[Wr, Wr1] represents the correlation degree of the two detection data sets of Wr and Wr1 cleaning process group r; Wr1 represents the cleanliness final detection data set of the newly generated wafer in the rth cleaning process group of the wafer cleaning equipment; Wr represents the cleanliness historical final detection data set of the same cleaning process group r as Wr1 in the wafer cleaning equipment and the most recent occurrence time; Wrf' represents the wafer cleanliness sampling detection data set of any cleaning process f' in the rth cleaning process group of the wafer cleaning equipment, and Wr1f' represents the wafer cleanliness sampling detection data set of any cleaning process f' in the rth cleaning process group of the wafer cleaning equipment; Where |Wrf'∩Wr1f'| represents the number of wafers with the same wafer cleanliness in any cleaning process f' in the cleaning process group r, and |Wrf'∪Wr1f'| represents the total number of wafers in the sampled test data set in any cleaning process f' in the cleaning process group r; S202. It is necessary to judge whether the cleaning process group r is abnormal. When K[Wr, Wr1]≥Q, it means that the cleaning process groups r are the same and the difference in the final detection data set of the cleanliness history of the most recent event is small. When K[Wr, Wr1]<Q, it means that the cleaning process groups r are the same and the difference in the final detection data set of the cleanliness history of the most recent event is large. The cleaning process group r is judged to be abnormal and an alarm is issued. Q is represented as the difference threshold of the cleaning process group.

4. The device debugging management method based on big data according to claim 3 is characterized by: The specific method for determining which component in the abnormal cleaning process group in the wafer cleaning equipment is abnormal in S300 includes: determining the abnormal score of wafer cleanliness of any cleaning process f' in the abnormal cleaning process group r by the formula: prf'=(|Wrf'∩Wr1f'| / |Wrf'∪Wr1f'|), where prf' is represented by the abnormal score of wafer cleanliness of any cleaning process f' in the abnormal cleaning process group r; A comprehensive scoring model is constructed based on the cleaning process parts data information in the historical cleaning database: Prf'=prf' / (Vrf' / Trf'), and the wafer cleanliness abnormality comprehensive score of any cleaning process f' in the abnormal cleaning process group r is calculated as Prf'; Trf' and Vrf' respectively represent the parts replacement frequency and usage time of any cleaning process f' in the cleaning process group r; the cleaning processes in the cleaning process group are sorted from high to low according to the abnormal comprehensive score to generate an abnormal cleaning process sequence.

5. The device debugging management method based on big data according to claim 4 is characterized in that: The specific method of adjusting the operating parameters of the abnormal parts or replacing the parts in S400 is as follows: adjusting the operating parameters of the parts or replacing the parts according to the sorting result of the abnormal cleaning process sequence, and then judging according to the steps of S200, when it is judged to be non-abnormal, writing the adjusted operating parameters into the cleaning program, replacing the operating parameters of the corresponding parts of the original cleaning program. When it is judged to be abnormal, further adjusting the operating parameters of the parts or replacing the parts until it is judged to be non-abnormal.

6. A device debugging management system based on big data, characterized by: The equipment debugging and management system comprises a data acquisition module, a data analysis module, an equipment adjustment management module and an alarm reminder module; the output end of the data acquisition module is connected to the input end of the data analysis module, the output end of the data analysis module is connected to the input end of the equipment debugging and management module, and the output end of the equipment adjustment management module is connected to the input end of the alarm reminder module; the data acquisition module obtains historical data information of parts and historical data of wafer cleanliness through background data information of wafer cleaning equipment; the data analysis module completes the judgment of abnormal cleaning process by constructing an abnormality detection model and generates an abnormal cleaning process sequence; The equipment adjustment management module adjusts the components of the wafer cleaning equipment according to the abnormal cleaning process sequence, and detects the cleanliness of the wafers after adjustment; the alarm reminder module issues an alarm reminder when it is determined that an abnormal cleaning process is present, adjusts the operating parameters of the abnormal components or replaces the components, and after the adjustment or replacement, determines whether there is an abnormality according to the method of step S200. When it is determined to be non-abnormal, the adjusted operating parameters are written into the cleaning program to replace the operating parameters of the corresponding components of the original cleaning program.

7. The device debugging management system based on big data according to claim 6 is characterized by: The data acquisition module includes a historical data information acquisition unit and a cleaning history data acquisition unit; the historical data information acquisition unit collects historical data information of wafer cleaning equipment parts from a historical cleaning database; the cleaning history data acquisition unit collects wafer cleaning history data through a historical cleaning database.

8. The device debugging management system based on big data according to claim 7 is characterized by: The data analysis module includes an abnormal detection model building unit, an abnormal cleaning process judgment unit and an abnormal cleaning process sequence generation unit; the abnormal detection model building unit builds an abnormal detection model of the cleaning process group through the wafer cleaning history data, and the abnormal detection model of the cleaning process group can analyze and judge the abnormal cleaning process group in the wafer cleaning equipment; the abnormal cleaning process judgment unit makes a preliminary judgment on the abnormality of the cleaning process group in the wafer cleaning equipment, and subsequently determines the abnormal cleaning process; the abnormal cleaning process sequence generation unit generates an abnormal cleaning process sequence by sorting the abnormal cleaning processes according to the comprehensive abnormality score.

9. The device debugging management system based on big data according to claim 8 is characterized by: The equipment adjustment management module includes a component operating parameter adjustment unit and a wafer cleanliness detection unit; the component operating parameter adjustment unit adjusts the component operating parameters or replaces the component according to the sorting results of the abnormal cleaning process sequence; the wafer cleanliness detection unit detects the cleanliness of the wafer after the operating parameters of the components of the wafer cleaning equipment are adjusted.

10. The device debugging management system based on big data according to claim 9 is characterized in that: The alarm reminder module includes an alarm reminder unit and a data storage and recording unit; the alarm reminder unit is used to give an alarm reminder when an abnormal cleaning process group exists in the wafer cleaning equipment; The data storage and recording unit writes the adjusted operating parameters into the cleaning program to replace the operating parameters of the corresponding parts of the original cleaning program when the operating parameters of the parts are adjusted and the cleaning process is determined to be normal.

Images