Intelligent cooperative control system and method for warped product cleaning
By collecting data and breaking down the process of warped products, a feature database was established, and the deviations and influence coefficients of sub-processes were analyzed. The defect rate was monitored in real time, which solved the problems of unstable quality and low efficiency in the cleaning of warped products and achieved efficient and controllable cleaning process management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGSU FERROTEC SEMICON TECH CO LTD
- Filing Date
- 2026-02-09
- Publication Date
- 2026-05-08
AI Technical Summary
Existing technologies for cleaning warped products lack precise and coordinated control throughout the entire process, resulting in unstable cleaning quality, low efficiency, and high costs. They also fail to achieve dynamic adaptation between product characteristics and cleaning parameters, making it difficult to quickly locate cleaning defect rate issues and develop targeted control solutions.
By collecting data on warped products, establishing a product characteristic database, dividing the cleaning process into multiple sub-steps, analyzing the deviation and impact coefficient of each sub-step, and combining historical defect rate data, parameters are monitored and updated in real time to achieve precise control of the cleaning process.
It enables precise and collaborative management of the cleaning process, significantly reduces the defect rate, improves production efficiency and equipment stability, reduces rework costs, and provides scientific data support and rapid problem location capabilities.
Smart Images
Figure CN121995885A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent cleaning control technology, specifically to an intelligent collaborative control system and method for cleaning warped products. Background Technology
[0002] In the field of warped product cleaning technology, a core technical problem has long existed: the lack of a precise, collaborative management and control mechanism across the entire process. This problem directly leads to a technical dilemma in the cleaning process, characterized by poor quality stability, low production efficiency, and high overall costs. Specifically, due to the lack of support from precise, collaborative management and control across the entire process, existing technologies cannot establish a dynamic adaptation relationship between product characteristics and cleaning parameters. Instead, they can only adopt a unified and fixed cleaning parameter scheme. This unified and fixed cleaning parameter scheme fails to achieve a refined breakdown and collaborative linkage of the entire process, and cannot clearly define the correlation and impact mechanism between each sub-cleaning step and the final product quality. When the cleaning defect rate becomes abnormal, all cleaning parameters and steps can only be blindly investigated. This makes it difficult to quickly locate the root cause of the problem or formulate targeted control measures, resulting in a persistently high percentage of defective products and a significant increase in rework and repair costs. Furthermore, the lack of a precise and collaborative management and control mechanism throughout the entire process also means that the existing cleaning management system lacks the ability to regularly iterate data and link equipment early warnings: on the one hand, it is impossible to integrate and utilize various data in the actual cleaning process to achieve dynamic optimization of the cleaning plan; on the other hand, it is difficult to predict potential equipment failures through process data monitoring, and often it is necessary to wait until the equipment stops and fails before it can be dealt with, which in turn causes the production process to be interrupted, further exacerbating the dual dilemma of quality and efficiency. Summary of the Invention
[0003] The purpose of this invention is to provide an intelligent collaborative control system and method for cleaning warped products, so as to solve the problems mentioned in the background art.
[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent collaborative control method for cleaning warped products, comprising the following steps: S1. Collect data on the warped product to obtain its characteristics; when cleaning the warped product, obtain the actual cleaning parameters. S2. Based on the warped product cleaning process, divide the warped product cleaning sub-process, calculate the warped product deviation degree of the sub-process, compare the actual state and standard state of each batch of products after passing through the corresponding sub-stage based on the warped product deviation degree of the sub-process, determine the deviation degree of a single batch of products in the sub-stage, consider the actual cleaning parameters, and analyze the influence of the cleaning parameters on the sub-stage. S3. Based on the cleaning defect rate in historical data, analyze the influence coefficient of the cleaning sub-process on the cleaning defect rate; S4. When cleaning a new batch of warped products, analyze the degree of influence of the actual cleaning parameters of the new batch of warped products on the product defect rate, and record the degree of influence in the defect rate anomaly emergency handling database. S5. Monitor product defect rate in real time, and issue alarms to managers to rectify actual cleaning parameters and equipment malfunctions based on the defect rate; S6. Update the database in real time according to the preset cycle.
[0005] When analyzing the influence coefficient of cleaning parameters on the cleaning sub-process based on the deviation of warped products in the sub-process, the cleaning process of warped products is divided into multiple sub-steps. The existing product feature database is called up, and the actual state of each batch of products after passing through the corresponding sub-step is compared with the standard state to determine the degree of deviation of a single batch of products in this sub-step. By combining the preset standard cleaning parameters with the actual cleaning parameters of each batch, the strength of the effect of each parameter on the cleaning effect of the sub-step is measured, and then the relative influence coefficient of the cleaning parameters on the sub-step is obtained.
[0006] Furthermore, in step S1, after authorization, data is collected from the warped products to obtain warped product data. For any type of warped product, the warped product data is preprocessed and normalized to obtain the warped product characteristics of that type of warped product, thereby establishing a warped product characteristic database. Pre-set cleaning parameters are used as standard cleaning parameters. When cleaning any batch of warped products, the actual cleaning parameters during the cleaning process are collected in real time. The authorized collection process ensures the legality and accuracy of the data source. The collected warped product data is preprocessed to extract features, constructing a comprehensive product characteristic database, providing reliable data support for the precise control of the subsequent cleaning process. At the same time, the pre-set standard cleaning parameters and the synchronous real-time collection of actual cleaning parameters form a benchmark for parameter comparison, avoiding parameter setting deviations caused by the lack of a unified reference. This lays a solid foundation for subsequent analysis of the correlation between cleaning parameters and product characteristics, improving the scientific nature and controllability of the cleaning process from the source.
[0007] Furthermore, in step S2, based on the warped product cleaning process, the warped product cleaning process is divided into M warped product cleaning sub-processes. For the m-th cleaning sub-process, m=1,2,…M, the standard characteristic of the warped product in the m-th cleaning sub-process is {A}. 1_m A 2_m ,…,A n_m ,…,A N_m}, where N represents the number of standard features of the warped product, A n_mLet {B} represent the standard feature of the nth warped product. The standard feature of the mth cleaning sub-process represents the standard feature specified after the warped product has passed through the mth cleaning sub-process. The warped product feature database is called, and the number of batches of warped products after cleaning is P. The warped product feature of the pth batch after the mth cleaning sub-process is {B}. 1_m_p B 2_m_p ,…,B n_m_p ,…,B N_m_p}, where B n_m_p This represents the characteristics of the nth warped product after the mth cleaning sub-process of the pth batch of warped products, where p = 1, 2, ..., P. This leads to the sub-process warped product deviation C of the mth cleaning sub-process. p_m : ; Substituting p = 1, 2, ..., P into the equations, we obtain the sub-process warpage deviation of the m-th cleaning sub-process for batches P of warped products. The sub-process warpage deviation of the m-th cleaning sub-process for batches P of warped products is {C}. 1_m C 2_m ,…,C p_m ,…,C P_m The standard cleaning parameters for the cleaning process are {R1, R2, ..., R}. y ,…,R Y}, where R y This represents the y-th standard cleaning parameter, which calls the actual cleaning parameters of the p-th batch of warped products. The actual cleaning parameters of the p-th batch of warped products are {R}. 1_p ,R 2_p ,…,R y_p ,…,R Y_p}, where R y_p Let y be the actual cleaning parameter of the p-th batch of warped products, and then calculate the degree of influence D of the y-th cleaning parameter on the m-th cleaning sub-process. y_m : ; Where C m Represents {C 1_m C 2_m ,…,C p_m ,…,C P_m Substituting the average of all values in} into y=1,2,…Y, we obtain the degree of influence of the Y cleaning parameters on the m-th cleaning subprocess {D}. 1_m D 2_m ,…,D y_m ,…,D Y_m}, and then obtain the influence coefficient d of the y-th cleaning parameter on the m-th cleaning sub-process. y_mThe influence coefficient of the y-th cleaning parameter on the m-th cleaning sub-process is the ratio of the influence of the y-th cleaning parameter on the m-th cleaning sub-process to the sum of the influences of the Y cleaning parameters on the m-th cleaning sub-process. Substituting each value into y=1,2,…Y, we obtain the influence coefficient of the Y cleaning parameters on the m-th cleaning sub-process. The influence coefficient of the Y cleaning parameters on the m-th cleaning sub-process is {d 1_m ,d 2_m ,…,d y_m ,…,d Y_m By meticulously breaking down the overall cleaning process, targeted control over each sub-cleaning step is achieved, avoiding the crude drawbacks of traditional holistic management models. By comparing the standard and actual characteristics of products in each sub-process, deviations in product characteristics are accurately measured. Simultaneously, by combining correlation analysis between standard and actual cleaning parameters, the specific impact weight of various cleaning parameters on each sub-process is clarified, forming a quantifiable basis for parameter influence. This provides scientific support for subsequent optimization and adjustment of cleaning parameters, effectively avoiding blind parameter control and improving the controllability and adaptability of the cleaning process.
[0008] Furthermore, in step S3, based on historical data, the cleaning defect rate of batch P warped products after M cleaning sub-processes is obtained, and the cleaning defect rate of batch P warped products after M cleaning sub-processes is {E1, E2, ..., E p ,…,E P}, where E p This represents the cleaning defect rate of the p-th batch of warped products after M cleaning sub-processes, and then the influence F of the m-th cleaning sub-process on the cleaning defect rate is calculated. m : ; Where e represents {E1, E2, ..., E...} p ,…,E P Substituting the average of all values in} into m=1,2,…M, we obtain the degree of influence of the M cleaning sub-processes on the cleaning defect rate {F1,F2,…,F}. m ,…,F M}, and thus obtain the influence coefficients {f1,f2,…,f} of the M cleaning sub-processes on the cleaning defect rate. m ,…,f M}, where f m f is the influence coefficient of the m-th cleaning sub-process on the cleaning defect rate. m For F m With {F1,F2,…,F m ,…,F MThe ratio of the sum of all values in the dataset, analyzed based on historical cleaning data, clarifies the impact weight of each cleaning sub-process on the overall cleaning defect rate, solving the problem of difficulty in identifying key influencing factors in the traditional management model. By quantifying the correlation between each sub-process and the defect rate, core focus areas are identified for subsequent defect rate control, avoiding indiscriminate investigation of all cleaning processes and significantly improving the efficiency and accuracy of tracing the source of defects. Simultaneously, the analysis results provide clear direction for optimizing the cleaning process, helping to improve by starting with key aspects affecting the defect rate and fundamentally reducing the risk of cleaning defects.
[0009] Furthermore, in step S4, when performing a cleaning task on the new batch of warped products, the actual cleaning parameters of the new batch of warped products are obtained. The actual cleaning parameters of the new batch of warped products are {S1, S2, ..., S...} y ,…,S Y}, where S y This represents the y-th actual cleaning parameter of the new batch of warped products, and the deviation T of the y-th actual cleaning parameter is calculated. y T y =|(S y -R y ) / R y |, and thus obtain the participation coefficient U of the y-th actual cleaning parameter to the m-th cleaning sub-process. y_m U y_m Let d be the deviation of the y-th actual cleaning parameter and the influence coefficient d of the y-th cleaning parameter on the m-th cleaning sub-process. y_m The product of these factors yields the defect rate participation coefficient V of the y-th actual cleaning parameter on the defect rate. y V y Let U be the sum of the products of the participation coefficient of the y-th actual cleaning parameter on the M cleaning sub-processes and the influence coefficients of the corresponding M cleaning sub-processes on the cleaning defect rate, where the participation coefficient U y_m With influence coefficient f m Correspondingly, substituting each parameter into y=1,2,…Y, we obtain the defect rate participation coefficients of the Y actual cleaning parameters on the defect rate. These coefficients are {V1,V2,…,V...} y ,…,V YThe system records the contribution coefficients of Y actual cleaning parameters to the defect rate in real time into the defect rate anomaly emergency handling database. Parameter analysis is conducted on the cleaning of new batches of warped products. By comparing the deviations between actual and standard cleaning parameters, and combining this with previously obtained parameter influence coefficients, the system clarifies the degree of influence of various parameter deviations on the cleaning defect rate, thus forming corresponding defect rate contribution coefficients. Simultaneously, these coefficients are entered into the defect rate anomaly emergency handling database in real time, providing accurate data support for the rapid handling of subsequent defect rate anomalies. This avoids blind investigation when defects occur, significantly improves the efficiency of anomaly tracing, and lays a solid data foundation for timely adjustment of cleaning parameters and control of defect rate spread.
[0010] Furthermore, in step S5, during the cleaning task of the new batch of warped products, the cleaning defect rate of the warped products is monitored in real time. If the cleaning defect rate is lower than the preset cleaning defect rate threshold, the actual defect rate is judged to be normal, and the cleaning of the new batch of warped products continues; otherwise, the defect rate is judged to be abnormal, the cleaning task is suspended, and the actual cleaning parameters are recommended to the management personnel for investigation based on the defect rate participation coefficient of Y actual cleaning parameters from high to low. After investigation, the cleaning task is re-entered, and the number of re-entries is calculated for each cleaning task. If the number of re-entries is lower than the preset number of re-entries threshold, the cleaning task is judged to be normal; otherwise, the cleaning equipment is judged to be faulty, and a cleaning equipment fault alarm is issued to the management personnel. By monitoring the defect rate of the new batch of warped products in real time, dynamic control of cleaning quality is achieved. It can issue an early warning and suspend the cleaning operation as soon as the defect rate is abnormal, effectively avoiding the generation of batch defective products and reducing subsequent rework costs. At the same time, targeted parameter investigation is carried out based on the defect rate participation coefficient obtained in the early stage, without the need to verify all parameters one by one, which greatly improves the efficiency of problem location. In addition, by monitoring the number of times the cleaning task is restarted, potential equipment failures can be identified in a timely manner, avoiding continuous quality abnormalities caused by equipment problems and ensuring the stable operation of the cleaning process.
[0011] Furthermore, in step S6, the database is updated after a warped product cleaning task management cycle.
[0012] The intelligent collaborative control system for cleaning warped products includes: a warped product data acquisition module, a cleaning parameter impact analysis module, a sub-process defect rate analysis module, a new batch parameter participation analysis module, a cleaning defect rate monitoring and early warning module, and a cleaning data periodic update module. The warped product data acquisition module is used to acquire data from the warped product and obtain its characteristics; when cleaning the warped product, it acquires the actual cleaning parameters. The cleaning parameter impact analysis module is used to divide the warped product cleaning process into sub-processes, calculate the warped product deviation degree of each sub-process, compare the actual state and standard state of each batch of products after passing through the corresponding sub-stage based on the warped product deviation degree of each sub-process, determine the degree of deviation of a single batch of products in the sub-stage, and analyze the degree of impact of cleaning parameters on the sub-stage, taking into account the actual cleaning parameters. The sub-process defect rate analysis module is used to analyze the influence coefficient of the cleaning sub-process on the cleaning defect rate based on the cleaning defect rate in historical data. The new batch parameter participation analysis module is used to analyze the participation degree of the actual cleaning parameters of the new batch of warped products on the product defect rate when the new batch of warped products is cleaned, and record the participation degree in the defect rate abnormal emergency handling database. The cleaning defect rate monitoring and early warning module is used to monitor the product defect rate in real time and issue alarms to managers to rectify the actual cleaning parameters and equipment malfunctions based on the defect rate. The cleaning data periodic update module is used to update the database in real time according to a preset period.
[0013] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: Firstly, it enables precise and collaborative control of the cleaning process. This method establishes a comprehensive product feature database by conducting full-process data collection and feature extraction on warped products. Simultaneously, by combining real-time monitoring of relevant parameters during the cleaning process, it performs correlation analysis between the product's own characteristics and cleaning parameters, clarifying the impact of various parameters in different cleaning stages on the post-cleaning characteristics of the product. This targeted control method based on product characteristics allows the cleaning operation to better adapt to the actual needs of different types of warped products, effectively improving the adaptability and accuracy of the cleaning process and avoiding poor cleaning results caused by improper parameter settings.
[0014] On the one hand, it significantly reduces the defect rate of warped products after cleaning. The method breaks down the entire cleaning process into multiple sub-stages, analyzing the deviations between the standard characteristics of each sub-stage and the actual characteristics of the cleaned products. It also quantifies the impact of various cleaning parameters on each sub-stage and the weight of each sub-stage's contribution to the overall cleaning defect rate. When an abnormal cleaning defect rate occurs, targeted investigations can be conducted based on the priority of parameter impact on the defect rate, quickly identifying key parameters causing defects, and promptly adjusting and optimizing the cleaning strategy to reduce the generation of defective products from the source and improve the overall cleaning pass rate.
[0015] On the other hand, a closed-loop optimization system for the cleaning process is constructed. After completing a cleaning task management cycle, the relevant database is updated in a timely manner, incorporating new product characteristic data, cleaning parameter data, and defect rate data, so that the entire control method can be continuously iterated and improved with actual application. At the same time, by monitoring the number of cleaning task re-entries, potential faults in the cleaning equipment can be identified in a timely manner, and early warning prompts can be issued, reducing production stoppages caused by equipment failures, ensuring the long-term stable operation of the cleaning process, and improving overall production efficiency and management level. Attached Figure Description
[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a structural diagram of the intelligent collaborative control system for cleaning warped products according to the present invention; Figure 2 This is a flowchart of the intelligent collaborative control method for cleaning warped products according to the present invention. Detailed Implementation
[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] Please see Figure 1 and Figure 2 This invention provides a technical solution: an intelligent collaborative control method for cleaning warped products, comprising the following steps: S1. Collect data on the warped product to obtain its characteristics; when cleaning the warped product, obtain the actual cleaning parameters. S2. Based on the warped product cleaning process, divide the warped product cleaning sub-process, calculate the warped product deviation degree of the sub-process, compare the actual state and standard state of each batch of products after passing through the corresponding sub-stage based on the warped product deviation degree of the sub-process, determine the deviation degree of a single batch of products in the sub-stage, consider the actual cleaning parameters, and analyze the influence of the cleaning parameters on the sub-stage. S3. Based on the cleaning defect rate in historical data, analyze the influence coefficient of the cleaning sub-process on the cleaning defect rate; S4. When cleaning a new batch of warped products, analyze the degree of influence of the actual cleaning parameters of the new batch of warped products on the product defect rate, and record the degree of influence in the defect rate anomaly emergency handling database. S5. Monitor product defect rate in real time, and issue alarms to managers to rectify actual cleaning parameters and equipment malfunctions based on the defect rate; S6. Update the database in real time according to the preset cycle.
[0019] When analyzing the influence coefficient of cleaning parameters on the cleaning sub-process based on the deviation of warped products in the sub-process, the cleaning process of warped products is divided into multiple sub-steps. The existing product feature database is called up, and the actual state of each batch of products after passing through the corresponding sub-step is compared with the standard state to determine the degree of deviation of a single batch of products in this sub-step. By combining the preset standard cleaning parameters with the actual cleaning parameters of each batch, the strength of the effect of each parameter on the cleaning effect of the sub-step is measured, and then the relative influence coefficient of the cleaning parameters on the sub-step is obtained.
[0020] In step S1, after authorization, data is collected from warped products to obtain warped product data. For any type of warped product, the warped product data is preprocessed and normalized to obtain the warped product characteristics of that type of warped product, thereby establishing a warped product characteristic database. Preset cleaning parameters are used as standard cleaning parameters. When cleaning any batch of warped products, the actual cleaning parameters during the cleaning process are collected in real time. The authorized collection process ensures the legality and accuracy of the data source. The collected warped product data is preprocessed to extract features, constructing a comprehensive product characteristic database, providing reliable data support for the precise control of the subsequent cleaning process. Simultaneously, the preset standard cleaning parameters and the real-time collection of actual cleaning parameters form a benchmark for parameter comparison, avoiding parameter setting deviations caused by the lack of a unified reference. This lays a solid foundation for subsequent analysis of the correlation between cleaning parameters and product characteristics, improving the scientific nature and controllability of the cleaning process from the source.
[0021] In step S2, based on the warped product cleaning process, the warped product cleaning process is divided into M warped product cleaning sub-processes. For the m-th cleaning sub-process, m=1,2,…M, the standard characteristic of the warped product in the m-th cleaning sub-process is {A}. 1_m A 2_m ,…,A n_m ,…,A N_m}, where N represents the number of standard features of the warped product, A n_m Let {B} represent the standard feature of the nth warped product. The standard feature of the mth cleaning sub-process represents the standard feature specified after the warped product has passed through the mth cleaning sub-process. The warped product feature database is called, and the number of batches of warped products after cleaning is P. The warped product feature of the pth batch after the mth cleaning sub-process is {B}. 1_m_p B 2_m_p ,…,B n_m_p ,…,BN_m_p}, where B n_m_p This represents the characteristics of the nth warped product after the mth cleaning sub-process of the pth batch of warped products, where p = 1, 2, ..., P. This leads to the sub-process warped product deviation C of the mth cleaning sub-process. p_m : ; Substituting p = 1, 2, ..., P into the equations, we obtain the sub-process warpage deviation of the m-th cleaning sub-process for batches P of warped products. The sub-process warpage deviation of the m-th cleaning sub-process for batches P of warped products is {C}. 1_m C 2_m ,…,C p_m ,…,C P_m The standard cleaning parameters for the cleaning process are {R1, R2, ..., R}. y ,…,R Y}, where R y This represents the y-th standard cleaning parameter, which calls the actual cleaning parameters of the p-th batch of warped products. The actual cleaning parameters of the p-th batch of warped products are {R}. 1_p ,R 2_p ,…,R y_p ,…,R Y_p}, where R y_p Let y be the actual cleaning parameter of the p-th batch of warped products, and then calculate the degree of influence D of the y-th cleaning parameter on the m-th cleaning sub-process. y_m : ; Where C m Represents {C 1_m C 2_m ,…,C p_m ,…,C P_m Substituting the average of all values in} into y=1,2,…Y, we obtain the degree of influence of the Y cleaning parameters on the m-th cleaning subprocess {D}. 1_m D 2_m ,…,D y_m ,…,D Y_m}, and then obtain the influence coefficient d of the y-th cleaning parameter on the m-th cleaning sub-process. y_m The influence coefficient of the y-th cleaning parameter on the m-th cleaning sub-process is the ratio of the influence of the y-th cleaning parameter on the m-th cleaning sub-process to the sum of the influences of the Y cleaning parameters on the m-th cleaning sub-process. Substituting each value into y=1,2,…Y, we obtain the influence coefficient of the Y cleaning parameters on the m-th cleaning sub-process. The influence coefficient of the Y cleaning parameters on the m-th cleaning sub-process is {d 1_m ,d 2_m ,…,d y_m,…,d Y_m By meticulously breaking down the overall cleaning process, targeted control over each sub-cleaning step is achieved, avoiding the crude drawbacks of traditional holistic management models. By comparing the standard and actual characteristics of products in each sub-process, deviations in product characteristics are accurately measured. Simultaneously, by combining correlation analysis between standard and actual cleaning parameters, the specific impact weight of various cleaning parameters on each sub-process is clarified, forming a quantifiable basis for parameter influence. This provides scientific support for subsequent optimization and adjustment of cleaning parameters, effectively avoiding blind parameter control and improving the controllability and adaptability of the cleaning process.
[0022] In step S3, based on historical data, the cleaning defect rate of batch P warped products after M cleaning sub-processes is obtained. The cleaning defect rate of batch P warped products after M cleaning sub-processes is {E1, E2, ..., E...} p ,…,E P}, where E p This represents the cleaning defect rate of the p-th batch of warped products after M cleaning sub-processes, and then the influence F of the m-th cleaning sub-process on the cleaning defect rate is calculated. m : ; Where e represents {E1, E2, ..., E...} p ,…,E P Substituting the average of all values in} into m=1,2,…M, we obtain the degree of influence of the M cleaning sub-processes on the cleaning defect rate {F1,F2,…,F}. m ,…,F M}, and thus obtain the influence coefficients {f1,f2,…,f} of the M cleaning sub-processes on the cleaning defect rate. m ,…,f M}, where f m f is the influence coefficient of the m-th cleaning sub-process on the cleaning defect rate. m For F m With {F1,F2,…,F m ,…,F M The ratio of the sum of all values in the dataset, analyzed based on historical cleaning data, clarifies the impact weight of each cleaning sub-process on the overall cleaning defect rate, solving the problem of difficulty in identifying key influencing factors in the traditional management model. By quantifying the correlation between each sub-process and the defect rate, core focus areas are identified for subsequent defect rate control, avoiding indiscriminate investigation of all cleaning processes and significantly improving the efficiency and accuracy of tracing the source of defects. Simultaneously, the analysis results provide clear direction for optimizing the cleaning process, helping to improve by starting with key aspects affecting the defect rate and fundamentally reducing the risk of cleaning defects.
[0023] In step S4, when performing a cleaning task on the new batch of warped products, the actual cleaning parameters of the new batch of warped products are obtained. The actual cleaning parameters of the new batch of warped products are {S1, S2, ..., S...} y ,…,S Y}, where S y This represents the y-th actual cleaning parameter of the new batch of warped products, and the deviation T of the y-th actual cleaning parameter is calculated. y T y =|(S y -R y ) / R y |, and thus obtain the participation coefficient U of the y-th actual cleaning parameter to the m-th cleaning sub-process. y_m U y_m Let d be the deviation of the y-th actual cleaning parameter and the influence coefficient d of the y-th cleaning parameter on the m-th cleaning sub-process. y_m The product of these factors yields the defect rate participation coefficient V of the y-th actual cleaning parameter on the defect rate. y V y Let U be the sum of the products of the participation coefficient of the y-th actual cleaning parameter on the M cleaning sub-processes and the influence coefficients of the corresponding M cleaning sub-processes on the cleaning defect rate, where the participation coefficient U y_m With influence coefficient f m Correspondingly, substituting each parameter into y=1,2,…Y, we obtain the defect rate participation coefficients of the Y actual cleaning parameters on the defect rate. These coefficients are {V1,V2,…,V...} y ,…,V Y The system records the contribution coefficients of Y actual cleaning parameters to the defect rate in real time into the defect rate anomaly emergency handling database. Parameter analysis is conducted on the cleaning of new batches of warped products. By comparing the deviations between actual and standard cleaning parameters, and combining this with previously obtained parameter influence coefficients, the system clarifies the degree of influence of various parameter deviations on the cleaning defect rate, thus forming corresponding defect rate contribution coefficients. Simultaneously, these coefficients are entered into the defect rate anomaly emergency handling database in real time, providing accurate data support for the rapid handling of subsequent defect rate anomalies. This avoids blind investigation when defects occur, significantly improves the efficiency of anomaly tracing, and lays a solid data foundation for timely adjustment of cleaning parameters and control of defect rate spread.
[0024] In step S5, during the cleaning task of the new batch of warped products, the cleaning defect rate of the warped products is monitored in real time. If the cleaning defect rate is lower than the preset cleaning defect rate threshold, the actual defect rate is judged to be normal, and the cleaning of the new batch of warped products continues. Otherwise, the defect rate is judged to be abnormal, the cleaning task is suspended, and the actual cleaning parameters are recommended to the management personnel for investigation based on the defect rate participation coefficient of Y actual cleaning parameters from high to low. After investigation, the cleaning task is re-entered. The number of re-entries for each cleaning task is calculated. If the number of re-entries is lower than the preset re-entry number threshold, the cleaning task is judged to be normal. Otherwise, the cleaning equipment is judged to be faulty, and a cleaning equipment fault alarm is issued to the management personnel. By monitoring the defect rate of the new batch of warped products in real time, dynamic control of cleaning quality is achieved. It can issue an early warning and suspend the cleaning operation as soon as the defect rate is abnormal, effectively avoiding the generation of batch defective products and reducing subsequent rework costs. At the same time, targeted parameter investigation is carried out based on the defect rate participation coefficient obtained in the early stage, without the need to verify all parameters one by one, which greatly improves the efficiency of problem location. In addition, by monitoring the number of times the cleaning task is restarted, potential equipment failures can be identified in a timely manner, avoiding continuous quality abnormalities caused by equipment problems and ensuring the stable operation of the cleaning process.
[0025] In step S6, the database is updated after a warped product cleaning task management cycle.
[0026] An intelligent collaborative control system for cleaning warped products, the system comprising: a warped product data acquisition module, a cleaning parameter impact analysis module, a sub-process defect rate analysis module, a new batch parameter participation analysis module, a cleaning defect rate monitoring and early warning module, and a cleaning data periodic update module; The warped product data acquisition module is used to collect data on warped products and obtain their characteristics; when cleaning warped products, it also obtains the actual cleaning parameters. The cleaning parameter impact analysis module is used to divide the warped product cleaning process into sub-processes, calculate the warped product deviation degree of each sub-process, compare the actual state and standard state of each batch of products after passing through the corresponding sub-process based on the warped product deviation degree of each sub-process, determine the deviation degree of a single batch of products in the sub-process, and analyze the impact degree of cleaning parameters on the sub-process by considering the actual cleaning parameters. The sub-process defect rate analysis module is used to analyze the impact coefficient of the cleaning sub-process on the cleaning defect rate based on historical cleaning defect rate data. The new batch parameter participation analysis module is used to analyze the participation of the actual cleaning parameters of the new batch of warped products on the product defect rate when cleaning the new batch of warped products, and to record the participation in the defect rate anomaly emergency handling database. The cleaning defect rate monitoring and early warning module is used to monitor the product defect rate in real time and issue alarms to managers to rectify the actual cleaning parameters and equipment malfunctions based on the defect rate. The data cleaning and periodic update module is used to update the database in real time according to a preset period.
[0027] Example 1: In the data preparation stage, data collection for warped products is first conducted through an authorization mechanism to ensure the legality and accuracy of the data source. For different types of warped products, the collected data undergoes preprocessing and normalization to extract core characteristic information for each product type, thus building a comprehensive warped product characteristic database. Simultaneously, cleaning-related parameters are pre-set as standard references. When cleaning any batch of warped products, the actual cleaning parameters during the cleaning process are collected in real time, forming a comparison system between standard and actual parameters. This provides a reliable reference for subsequent analysis. Cleaning parameters include, for example, the relative value of cleaning pressure. Using the standard pressure value allowed by the corresponding cleaning sub-process as a benchmark, the ratio of the real-time collected pressure during the cleaning process to this benchmark pressure is calculated to obtain a dimensionless relative pressure value. This parameter reflects the difference between the actual and standard operating conditions of the cleaning pressure; and the relative proportion of ultrasonic power. If the cleaning process includes an ultrasonic cleaning step, the standard ultrasonic power of that step is used as a benchmark. The actual output ultrasonic power is then compared to the benchmark power to calculate the relative proportion of ultrasonic power. This parameter measures the rationality of the ultrasonic energy input and is unaffected by power unit dimensions.
[0028] In the parameter impact analysis phase, the overall cleaning process is broken down into several sub-operational steps, and corresponding product standard characteristics are set for each sub-step. The actual characteristics of products after past cleaning are retrieved from the feature database, and the deviations are compared. Then, combining the standard and actual cleaning parameters, the specific impact of various parameters on each sub-step is analyzed, and the impact weight of each parameter on each step is calculated. Subsequently, based on historical cleaning records, the defect rate of products after completing all cleaning steps is statistically analyzed to determine the impact of each sub-cleaning step on the overall defect rate, clarifying the impact weight of each step, thereby identifying the key control points in the cleaning process.
[0029] During the new batch cleaning control phase, actual cleaning parameters are collected during the cleaning of new batches of products. These parameters are compared with standard parameters to determine deviations. Combined with previously obtained parameter and process influence weights, the impact of various parameter deviations on the defect rate is calculated, generating corresponding adverse impact coefficients. These coefficients are then entered into the defect rate anomaly emergency handling database in real time. The defect rate is monitored in real time during cleaning. If it does not exceed a preset threshold, operation continues; if it does, cleaning is immediately paused. Relevant parameters are investigated in descending order of adverse impact coefficient. After investigation and adjustment, cleaning is restarted. The number of cleaning restarts is also recorded. If the number of restarts exceeds a preset limit, an equipment malfunction is identified, and an alarm is promptly issued to management.
[0030] During the cycle summary phase, after completing a full cleaning task management cycle, all product characteristics, cleaning parameters, defect rate data, and abnormal handling records within that cycle are added to the corresponding database to achieve dynamic database updates and provide a more comprehensive reference for subsequent batch cleaning control.
[0031] The entire implementation process uses data-driven methods to achieve precise control of the cleaning process, effectively reducing the defect rate of warped products and improving operational stability.
[0032] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary sensing device embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.
Claims
1. An intelligent collaborative control method for cleaning warped products, characterized in that: The method includes the following steps: S1. Collect data on the warped product to obtain its characteristics; when cleaning the warped product, obtain the actual cleaning parameters. S2. Based on the warped product cleaning process, divide the warped product cleaning sub-process, calculate the warped product deviation degree of the sub-process, compare the actual state and standard state of each batch of products after passing through the corresponding sub-stage based on the warped product deviation degree of the sub-process, determine the deviation degree of a single batch of products in the sub-stage, consider the actual cleaning parameters, and analyze the influence of the cleaning parameters on the sub-stage. S3. Based on the cleaning defect rate in historical data, analyze the influence coefficient of the cleaning sub-process on the cleaning defect rate; S4. When cleaning a new batch of warped products, analyze the degree of influence of the actual cleaning parameters of the new batch of warped products on the product defect rate, and record the degree of influence in the defect rate anomaly emergency handling database. S5. Monitor product defect rate in real time, and issue alarms to managers to rectify actual cleaning parameters and equipment malfunctions based on the defect rate; S6. Update the database in real time according to the preset cycle.
2. The intelligent collaborative control method for cleaning warped products according to claim 1, characterized in that: When analyzing the influence coefficient of cleaning parameters on the cleaning sub-process based on the deviation of warped products in the sub-process, the cleaning process of warped products is divided into several sub-steps. The existing product feature database is called up, and the actual state of each batch of products after passing through the corresponding sub-step is compared with the standard state to determine the degree of deviation of a single batch of products in the sub-step. Combined with the preset standard cleaning parameters and the actual cleaning parameters of each batch, the strength of the effect of each parameter on the cleaning effect of the sub-step is measured, and then the relative influence coefficient of the cleaning parameters on the sub-step is obtained.
3. The intelligent collaborative control method for cleaning warped products according to claim 2, characterized in that: In step S1, after authorization, data is collected from the warped products to obtain warped product data. For any type of warped product, the warped product data is preprocessed and normalized to obtain the warped product characteristics of any type of warped product. Then, a warped product characteristic database is established. Preset cleaning parameters are used as standard cleaning parameters. When cleaning any batch of warped products, the actual cleaning parameters during the cleaning of the warped products are collected in real time.
4. The intelligent collaborative control method for cleaning warped products according to claim 3, characterized in that: In step S2, based on the warped product cleaning process, the warped product cleaning process is divided into M warped product cleaning sub-processes. For the m-th cleaning sub-process, m=1,2,…M, the standard characteristic of the warped product in the m-th cleaning sub-process is {A}. 1_m A 2_m ,…,A n_m ,…,A N_m }, where N represents the number of standard features of the warped product, A n_m Let {B} represent the standard feature of the nth warped product. The standard feature of the mth cleaning sub-process represents the standard feature specified after the warped product has passed through the mth cleaning sub-process. The warped product feature database is called, and the number of batches of warped products after cleaning is P. The warped product feature of the pth batch after the mth cleaning sub-process is {B}. 1_m_p B 2_m_p ,…,B n_m_p ,…,B N_m_p }, where B n_m_p This represents the characteristics of the nth warped product after the mth cleaning sub-process of the pth batch of warped products, where p = 1, 2, ..., P. This leads to the sub-process warped product deviation C of the mth cleaning sub-process. p_m : ； Substituting p = 1, 2, ..., P into the equations, we obtain the sub-process warpage deviation of the m-th cleaning sub-process for batches P of warped products. The sub-process warpage deviation of the m-th cleaning sub-process for batches P of warped products is {C}. 1_m C 2_m ,…,C p_m ,…,C P_m The standard cleaning parameters for the cleaning process are {R1, R2, ..., R}. y ,…,R Y }, where R y This represents the y-th standard cleaning parameter, which calls the actual cleaning parameters of the p-th batch of warped products. The actual cleaning parameters of the p-th batch of warped products are {R}. 1_p ,R 2_p ,…,R y_p ,…,R Y_p }, where R y_p Let y be the actual cleaning parameter of the p-th batch of warped products, and then calculate the degree of influence D of the y-th cleaning parameter on the m-th cleaning sub-process. y_m : ； Where C m Represents {C 1_m C 2_m ,…,C p_m ,…,C P_m Substituting the average of all values in} into y=1,2,…Y, we obtain the degree of influence of the Y cleaning parameters on the m-th cleaning subprocess {D}. 1_m D 2_m ,…,D y_m ,…,D Y_m }, and then obtain the influence coefficient d of the y-th cleaning parameter on the m-th cleaning sub-process. y_m The influence coefficient of the y-th cleaning parameter on the m-th cleaning sub-process is the ratio of the influence of the y-th cleaning parameter on the m-th cleaning sub-process to the sum of the influences of the Y cleaning parameters on the m-th cleaning sub-process. Substituting each value into y=1,2,…Y, we obtain the influence coefficient of the Y cleaning parameters on the m-th cleaning sub-process. The influence coefficient of the Y cleaning parameters on the m-th cleaning sub-process is {d 1_m ,d 2_m ,…,d y_m ,…,d Y_m } 5. The intelligent collaborative control method for cleaning warped products according to claim 4, characterized in that: In step S3, based on historical data, the cleaning defect rate of batch P warped products after M cleaning sub-processes is obtained. The cleaning defect rate of batch P warped products after M cleaning sub-processes is {E1, E2, ..., E...} p ,…,E P }, where E p This represents the cleaning defect rate of the p-th batch of warped products after M cleaning sub-processes, and then the influence F of the m-th cleaning sub-process on the cleaning defect rate is calculated. m : ； Where e represents {E1, E2, ..., E...} p ,…,E P Substituting the average of all values in} into m=1,2,…M, we obtain the degree of influence of the M cleaning sub-processes on the cleaning defect rate {F1,F2,…,F}. m ,…,F M }, and thus obtain the influence coefficients {f1,f2,…,f} of the M cleaning sub-processes on the cleaning defect rate. m ,…,f M }, where f m f is the influence coefficient of the m-th cleaning sub-process on the cleaning defect rate. m For F m With {F1,F2,…,F m ,…,F M The ratio of the sum of all values in}.
6. The intelligent collaborative control method for cleaning warped products according to claim 5, characterized in that: In step S4, when performing a cleaning task on the new batch of warped products, the actual cleaning parameters of the new batch of warped products are obtained. The actual cleaning parameters of the new batch of warped products are {S1, S2, ..., S...} y ,…,S Y }, where S y This represents the y-th actual cleaning parameter of the new batch of warped products, and the deviation T of the y-th actual cleaning parameter is calculated. y T y =|(S y -R y ) / R y |, and thus obtain the participation coefficient U of the y-th actual cleaning parameter to the m-th cleaning sub-process. y_m U y_m Let d be the deviation of the y-th actual cleaning parameter and the influence coefficient d of the y-th cleaning parameter on the m-th cleaning sub-process. y_m The product of these factors yields the defect rate participation coefficient V of the y-th actual cleaning parameter on the defect rate. y V y Let U be the sum of the products of the participation coefficient of the y-th actual cleaning parameter on the M cleaning sub-processes and the influence coefficients of the corresponding M cleaning sub-processes on the cleaning defect rate, where the participation coefficient U y_m With influence coefficient f m Correspondingly, substituting each parameter into y=1,2,…Y, we obtain the defect rate participation coefficients of the Y actual cleaning parameters on the defect rate. These coefficients are {V1,V2,…,V...} y ,…,V Y The defect rate participation coefficients of the Y actual cleaning parameters on the defect rate are recorded in real time into the defect rate anomaly emergency handling database.
7. The intelligent collaborative control method for cleaning warped products according to claim 6, characterized in that: In step S5, during the cleaning task of the new batch of warped products, the cleaning defect rate of the warped products is monitored in real time. If the cleaning defect rate is lower than the preset cleaning defect rate threshold, the actual defect rate is judged to be normal, and the cleaning of the new batch of warped products continues; otherwise, the defect rate is judged to be abnormal, the cleaning task is suspended, and the actual cleaning parameters are recommended to the management personnel for investigation based on the defect rate participation coefficient of Y actual cleaning parameters from high to low. After investigation, the cleaning task is re-entered. The number of re-entries is calculated for each cleaning task. If the number of re-entries is lower than the preset number of re-entries threshold, the cleaning task is judged to be normal. Otherwise, if the cleaning equipment is deemed to be malfunctioning, a cleaning equipment malfunction alarm will be issued to the management personnel.
8. The intelligent collaborative control method for cleaning warped products according to claim 6, characterized in that: In step S6, the database is updated after a warped product cleaning task management cycle.
9. An intelligent collaborative control system for cleaning warped products, said system being applied to the intelligent collaborative control method for cleaning warped products according to any one of claims 1-8, characterized in that: The system includes: a warped product data acquisition module, a cleaning parameter impact analysis module, a sub-process defect rate analysis module, a new batch parameter participation analysis module, a cleaning defect rate monitoring and early warning module, and a cleaning data periodic update module. The warped product data acquisition module is used to acquire data from the warped product and obtain its characteristics; when cleaning the warped product, it acquires the actual cleaning parameters. The cleaning parameter impact analysis module is used to divide the warped product cleaning process into sub-processes, calculate the warped product deviation degree of each sub-process, compare the actual state and standard state of each batch of products after passing through the corresponding sub-stage based on the warped product deviation degree of each sub-process, determine the degree of deviation of a single batch of products in the sub-stage, and analyze the degree of impact of cleaning parameters on the sub-stage, taking into account the actual cleaning parameters. The sub-process defect rate analysis module is used to analyze the influence coefficient of the cleaning sub-process on the cleaning defect rate based on the cleaning defect rate in historical data. The new batch parameter participation analysis module is used to analyze the participation degree of the actual cleaning parameters of the new batch of warped products on the product defect rate when the new batch of warped products is cleaned, and record the participation degree in the defect rate abnormal emergency handling database. The cleaning defect rate monitoring and early warning module is used to monitor the product defect rate in real time and issue alarms to managers to rectify the actual cleaning parameters and equipment malfunctions based on the defect rate. The cleaning data periodic update module is used to update the database in real time according to a preset period.