Intelligent flow regulation and control type pipeline system of single-chip wet cleaning machine

Abstract

The invention relates to the technical field of semiconductor manufacturing and precision mechanical cleaning equipment, and discloses an intelligent flow control type monolithic wet cleaning machine pipeline system which comprises a high-precision PID flow control module, a pressure self-adaptive adjusting mechanism, an automatic control system and an intelligent fault diagnosis and prevention maintenance system. Wherein the high-precision PID flow control module adopts a high-precision turbine flowmeter, selects an electric regulating valve with a closed-loop control function, and is combined with a PID control algorithm, specifically a position type PID algorithm, so that the flow is quickly and accurately regulated, and the flow in each cleaning stage is stabilized in a range of a preset value + / -1%. According to the intelligent flow regulation and control type monolithic wet cleaning machine pipeline system, the problem that the cleaning effect is unstable due to the fact that flow distribution is uneven and pressure fluctuation is large in a traditional pipeline system is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor manufacturing and precision mechanical cleaning equipment, and in particular to an intelligent flow control type single-chip wet cleaning machine pipeline system. Background Art

[0002] In the field of semiconductor manufacturing and precision machining, single-wafer wet cleaning machines play an indispensable role. Their cleaning efficiency is directly related to the quality and performance of the final product. This type of high-precision equipment uses advanced wet cleaning technology, which can deeply remove particles, organic residues, metal impurities and natural oxide layers on the surface of chips or precision parts, ensuring that the surface of the material reaches extremely high cleanliness standards. By precisely controlling the temperature, concentration, flow rate and spray angle of the cleaning liquid, the single-wafer wet cleaning machine can realize customized cleaning solutions for different materials and process requirements, effectively avoid chemical corrosion and mechanical damage, and protect the surface integrity of the workpiece. In addition, the automated and intelligent cleaning process design not only improves production efficiency, but also significantly reduces human operation errors, ensuring the consistency and stability of the cleaning process. Therefore, optimizing the cleaning effect of the single-wafer wet cleaning machine is of far-reaching significance for improving the yield of semiconductor devices, extending the service life of precision mechanical parts, and promoting technological progress and industrial upgrading in related industries.

[0003] In the current multi-stage cleaning process of single-wafer wet cleaning machines, the demand for cleaning fluid flow in each stage is precise and variable, aiming to meet the specific requirements of different pollutant removal. However, traditional piping systems often lack flexibility and precision in design and are unable to dynamically adjust flow distribution according to the actual needs of each stage. This directly leads to unnecessary waste in some cleaning areas due to excess flow in actual operations, and may even cause surface damage caused by excessive cleaning; while other areas are not cleaned thoroughly due to insufficient flow, and residues remain, which seriously affects the cleaning effect. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of unstable cleaning effect caused by uneven flow distribution and large pressure fluctuations in traditional piping systems, and to propose an intelligent flow control type single-chip wet cleaning machine piping system.

[0005] The technical solution of the present invention to solve the above technical problems is as follows: Intelligent flow control type single-chip wet cleaning machine piping system, including high-precision PID flow control module, pressure adaptive adjustment mechanism, automatic control system and intelligent fault diagnosis and preventive maintenance system; The high-precision PID flow control module uses a high-precision turbine flowmeter and an electric regulating valve with closed-loop control function. It is combined with a PID control algorithm, specifically a position-based PID algorithm, to achieve fast and precise flow regulation, ensuring that the flow rate in each cleaning stage is stable within the preset value ±1% range. The pressure adaptive adjustment mechanism includes: installing Honeywell 26PCA series piezoelectric pressure sensors at key nodes to provide high-precision pressure monitoring of ±0.1%FS to ensure the accuracy of system pressure data, and using an intelligent pressure regulator with integrated adaptive algorithm to keep the system pressure within the set value ±0.05MPa range; The automatic control system integrates an embedded central control unit based on the ARM Cortex-A series processor to achieve rapid data processing and decision-making, and provides a 7-inch color touch screen user interface that supports graphical display of the cleaning process, real-time flow and pressure curves, and has early warning and alarm prompt functions, making the operation simple and intuitive; The intelligent fault diagnosis and preventive maintenance system uses historical data to train the SVM model for abnormal state identification, and combines it with the decision tree model for fault classification to improve the accuracy and efficiency of fault identification. Based on sensor data and system operating status, it uses built-in algorithm analysis to predict maintenance needs, automatically push maintenance prompts and system optimization suggestions to users, and achieve preventive maintenance.

[0006] Based on the above technical solution, the present invention can also be improved as follows.

[0007] Furthermore, the turbine flowmeter uses Krohne Optiflux series high-precision turbine flowmeters. The flowmeter should be installed at key positions in the piping system to ensure that it can accurately measure and reflect the actual flow conditions in each cleaning stage. During installation, factors such as the flow direction, temperature, and pressure of the fluid should be considered to ensure accurate measurement. The turbine flowmeter converts the flow into an electrical signal output through a built-in sensor. The signal is a 4-20mA analog signal or a digital signal. The signal is transmitted to the central control unit through a dedicated cable. The built-in signal processing module of the central control unit filters and amplifies the received signal to eliminate noise interference and improve data accuracy.

[0008] Furthermore, the electric control valve uses Siemens SIPART PS2 series electric control valve. The valve needs to be selected according to the flow range, pressure level and medium characteristics of the pipeline system to ensure that the valve can meet the system requirements. The electric control valve has a built-in closed-loop control system, which receives the flow set value signal from the central control unit and compares it with the actual flow signal fed back by the turbine flowmeter. According to the comparison result, the closed-loop control system changes the flow by adjusting the valve opening until the actual flow reaches or approaches the set value, thereby achieving precise control of the flow.

[0009] Furthermore, the PID control algorithm is implemented in an embedded central control unit. The logic and calculation process of the algorithm are realized by writing dedicated control software. The control software receives the flow signal from the turbine flowmeter and the flow set value signal from the user interface in real time, calculates the control quantity through the PID algorithm, and outputs it to the closed-loop control system of the electric control valve to achieve precise control of the flow.

[0010] Furthermore, the intelligent pressure regulator has a built-in integrated adaptive algorithm, which can automatically adjust the control strategy according to the actual changes in the system pressure to achieve precise control of the system pressure. The algorithm can gradually adapt to the dynamic changes of the system through continuous learning and optimization, and improve the stability and reliability of the system. When the system pressure deviates from the set value, the intelligent pressure regulator will automatically adjust the control quantity such as the opening of the regulating valve according to the size and direction of the deviation to restore the system pressure to the set value. At the same time, the algorithm will also consider the dynamic characteristics and hysteresis of the system to ensure that the adjustment process is smooth, fast and accurate. The user can also set the set value of the system pressure through the user interface of the central control unit. Once the set value is determined, the intelligent pressure regulator will perform pressure control according to the value. Through the intelligent pressure regulator with an integrated adaptive algorithm, the system pressure can be kept within the range of ±0.05MPa of the set value. This range is set according to system requirements and accuracy requirements, ensuring the stability and consistency of the system pressure. The system also has fault monitoring and alarm functions. When a sensor or regulator fails, the system will automatically send an alarm signal and take corresponding protective measures such as cutting off the power supply and closing the valve to prevent the fault from expanding and causing damage to the system.

[0011] Furthermore, at the data processing and decision-making level, the system uses stable and reliable operating systems such as embedded Linux or Android, supports multi-task concurrent processing, and through the development of dedicated data processing algorithms, the system can filter, detect anomalies and predict trends in real time on the data collected by sensors to ensure the accuracy and reliability of the data. Subsequently, the decision support system automatically adjusts system parameters and sends control instructions based on the processed data and preset thresholds and rules to realize automated control processes.

[0012] Furthermore, the automatic adjustment system collects the operation data of the equipment in real time through various sensors deployed on the equipment, such as temperature, pressure, vibration, current and other sensors. The data collection frequency is set according to the equipment characteristics and maintenance requirements to ensure the comprehensiveness and accuracy of the data. The collected raw data is cleaned to remove noise, outliers and redundant information, and data compression and dimensionality reduction processing are performed to improve the efficiency of subsequent processing. Key features such as time domain feature mean, variance, peak, etc., frequency domain feature power spectrum, spectrum entropy, etc. and time-frequency domain feature wavelet transform coefficients, energy spectrum density, etc. are extracted. The normal state and abnormal state samples in the historical data are used to train the support vector machine (SVM) model. The classification performance and generalization ability of the SVM model are improved through feature selection and optimization algorithms. Appropriate kernel functions and parameters are set to ensure that the model can accurately identify the abnormal state of the equipment. After the SVM model identifies the abnormal state, the decision tree model is used to further classify the abnormal state. The decision tree model constructs classification rules according to the fault characteristics and subdivides the abnormal state into specific fault types. The classification accuracy and efficiency of the decision tree model are improved through pruning and optimization algorithms.

[0013] Furthermore, based on the recognition results of SVM and decision tree models, combined with the operation history and current status of the equipment, the time series analysis algorithm is used to predict the future failure trend of the equipment, analyze the prediction results, determine the possible time of the failure, and determine the impact of the failure. According to the failure prediction results, a preventive maintenance plan is formulated to determine the priority, time and required resources of maintenance, evaluate the cost and benefit of the maintenance plan, ensure the economy and feasibility of maintenance activities, design an intuitive and easy-to-use user interaction interface, display the real-time operation status, fault identification results and maintenance plan of the equipment, and provide users with the functions of querying, modifying and confirming maintenance plans. When the system detects potential failure risks or needs to execute maintenance plans, it automatically pushes maintenance reminders and system optimization suggestions to users, and ensures that users can obtain relevant information in a timely manner through SMS, email or system notifications. According to user feedback and maintenance effects, the system's algorithms and models are continuously optimized, the data collection and processing processes are improved, and the overall performance and stability of the system are improved. Through online learning and update mechanisms, the timeliness and accuracy of the model are maintained.

[0014] Compared with the prior art, the technical solution of this application has the following beneficial technical effects: The intelligent flow control type single-chip wet cleaning machine piping system of the present invention integrates high-precision PID flow control, pressure adaptive regulation, automatic control, intelligent fault diagnosis and preventive maintenance, and improves the shortcomings of traditional piping systems. The system uses a high-precision turbine flowmeter and a closed-loop controlled electric regulating valve, combined with a position PID algorithm, to achieve accurate and rapid regulation of the cleaning liquid flow, ensuring the uniformity and stability of the flow distribution in each cleaning stage, preventing waste and damage caused by excess flow, and eliminating the problem of incomplete cleaning caused by insufficient flow, greatly improving the consistency of the cleaning effect. At the same time, the combination of Honeywell 26PCA piezoelectric pressure sensor and intelligent pressure regulator accurately monitors and adjusts the internal pressure of the system, reduces the interference of pressure fluctuations on the cleaning quality, and enhances the stability and controllability of the system. The system is also equipped with an ARM-based The powerful control unit of the Cortex-A processor is equipped with a 7-inch color touch screen, which uses a graphical interface to intuitively display the cleaning process, real-time flow and pressure data, and provides early warning alarm functions. It is easy to operate and the information is clear at a glance, which improves production efficiency and safety. More importantly, the intelligent fault diagnosis and preventive maintenance system uses SVM and decision tree models to accurately identify and classify faults, and pushes predictive maintenance suggestions, realizing the transformation from passive maintenance to active prevention, greatly reducing equipment failure rate, extending service life, reducing downtime losses, and significantly improving overall operational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a structural block diagram of the pipeline system of the intelligent flow control type single-chip wet cleaning machine of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] Combination Figure 1 As shown, the intelligent flow control type single-chip wet cleaning machine pipeline system of the present invention includes a high-precision PID flow control module, a pressure adaptive adjustment mechanism, an automatic control system and an intelligent fault diagnosis and preventive maintenance system; The high-precision PID flow control module adopts a high-precision turbine flowmeter, an electric control valve with a closed-loop control function, and a PID control algorithm, specifically a position PID algorithm. The parameters are optimized by experiments to Kp=0.8, Ki=0.05, Kd=0.01, to achieve fast and precise adjustment of the flow rate, ensuring that the flow rate in each cleaning stage is stable within the preset value ±1% range; the position PID control algorithm is an algorithm based on the proportional (P), integral (I), and differential (D) control of the deviation. The algorithm calculates the proportional, integral, and differential terms of the deviation to obtain the control quantity, and then adjusts the opening of the electric control valve. The proportional term (P) is used to quickly respond to deviation changes; the integral term (I) is used to eliminate static deviations; the differential term (D) is used to predict the deviation change trend, make adjustments in advance, and improve the stability and response speed of the system. The parameters of the PID control algorithm (Kp, Ki, Kd) need to be optimized according to the system characteristics. In this technical solution, through a large number of experiments and data analysis, it is finally determined that Kp=0.8, Ki=0.05, Kd=0.01 is the optimal parameter combination. These parameter values ​​enable the system to achieve rapid and precise flow adjustment while ensuring stability, ensuring that the flow in each cleaning stage is stable within the preset value ±1%.

[0018] The pressure adaptive adjustment mechanism includes: installing Honeywell 26PCA series piezoelectric pressure sensors at key nodes to provide high-precision pressure monitoring of ±0.1%FS to ensure the accuracy of system pressure data, and using an intelligent pressure regulator with integrated adaptive algorithm to keep the system pressure within the set value ±0.05MPa range; The automatic control system integrates an embedded central control unit based on the ARM Cortex-A series processor to achieve rapid data processing and decision-making, and provides a 7-inch color touch screen user interface that supports graphical display of the cleaning process, real-time flow and pressure curves, and has early warning and alarm prompt functions, making the operation simple and intuitive; The intelligent fault diagnosis and preventive maintenance system uses historical data to train the SVM model for abnormal state identification, and combines it with the decision tree model for fault classification to improve the accuracy and efficiency of fault identification. Based on sensor data and system operating status, it uses built-in algorithm analysis to predict maintenance needs, automatically push maintenance prompts and system optimization suggestions to users, and achieve preventive maintenance.

[0019] In a preferred embodiment, the present invention can be further configured as follows: Krohne Optiflux series high-precision turbine flowmeters are selected, and the flowmeters should be installed at key positions in the pipeline system to ensure that the actual flow conditions of each cleaning stage can be accurately measured and reflected. Factors such as the flow direction, temperature, and pressure of the fluid should be considered during installation to ensure accurate measurement. The turbine flowmeter converts the flow into an electrical signal output through a built-in sensor. The signal is a 4-20mA analog signal or a digital signal. The signal is transmitted to the central control unit through a dedicated cable. The signal processing module built into the central control unit filters and amplifies the received signal to eliminate noise interference and improve data accuracy. Krohne The Optiflux series of high-precision turbine flowmeters are installed at key positions in the pipeline system. This design ensures that the flowmeter can accurately measure and reflect the actual flow conditions of each cleaning stage in real time. By accurately monitoring the flow data, the system can more accurately control the flow distribution of each cleaning stage, thereby avoiding the difference in cleaning effect caused by uneven flow distribution in traditional systems. Secondly, the impact of factors such as fluid flow direction, temperature, and pressure on flow measurement is fully considered during installation to ensure the accuracy of the measurement results. This measure effectively reduces the interference of external environmental factors on flow measurement, improves the reliability of data, and provides a solid foundation for subsequent flow control. Furthermore, the turbine flowmeter converts the flow into electrical signal output through a built-in sensor and supports 4-20mA analog signal or digital signal transmission. This signal transmission method is not only stable and reliable, but also convenient for seamless docking with the central control unit. After the signal is transmitted to the central control unit through a dedicated cable, the built-in signal processing module will filter and amplify the received signal to eliminate noise interference and further improve the accuracy of the data. This process ensures that every link from flow measurement to signal processing meets high-precision requirements, providing strong support for the intelligent control of the system.

[0020] In a preferred embodiment, the present invention can be further configured as follows: Siemens SIPART PS2 series electric control valve is selected, and the valve needs to be selected according to the flow range, pressure level and medium characteristics of the pipeline system to ensure that the valve can meet the system requirements. The electric control valve has a built-in closed-loop control system, which receives the flow set value signal from the central control unit and compares it with the actual flow signal fed back by the turbine flowmeter. According to the comparison result, the closed-loop control system changes the flow by adjusting the opening of the valve until the actual flow reaches or approaches the set value, thereby realizing precise control of the flow. Siemens SIPART The selection of PS2 series electric control valves fully considers the flow range, pressure level and medium characteristics of the pipeline system, ensuring that the valve can fully meet the system requirements. This customized selection method based on the actual needs of the system effectively avoids the problem of inaccurate flow control caused by valve mismatch, and provides a strong guarantee for the stable operation of the system. Secondly, the electric control valve has a built-in closed-loop control system. This innovative design makes the flow control process more intelligent and precise. The system can receive the flow set value signal from the central control unit in real time and compare it with the actual flow signal fed back by the turbine flowmeter. By continuously monitoring and comparing these two signals, the closed-loop control system can accurately calculate the flow deviation and automatically adjust the valve opening accordingly to change the flow until the actual flow reaches or approaches the set value. This dynamic and continuous adjustment process ensures the accuracy and stability of flow control and effectively solves the problem of inaccurate flow control in traditional systems.

[0021] In a preferred embodiment, the present invention can be further configured as follows: the PID control algorithm is implemented in an embedded central control unit, and the logic and calculation process of the algorithm are realized by writing dedicated control software. The control software receives the flow signal from the turbine flowmeter and the flow setting value signal from the user interface in real time, calculates the control amount through the PID algorithm, and outputs it to the closed-loop control system of the electric control valve to realize accurate control of the flow rate. First of all, the PID control algorithm is a mature and effective control strategy, and its introduction provides powerful algorithm support for flow control. By writing dedicated control software, the logic and calculation process of the PID algorithm are embedded in the central control unit, realizing the real-time operation and accurate calculation of the algorithm. This integrated design not only simplifies the system structure, but also The execution efficiency and response speed of the control algorithm are improved, so that the system can adapt to flow changes more quickly and achieve more accurate flow control. Secondly, the control software receives the flow signal from the turbine flowmeter and the flow setting value signal from the user interface in real time. This design ensures the real-time and accuracy of the data. The high-precision flow data provided by the turbine flowmeter provides a reliable input basis for the PID algorithm, and the setting value of the user interface reflects the actual needs of the operator. By comparing these two signals in real time, the PID algorithm can quickly calculate the required control amount and output it to the closed-loop control system of the electric control valve. This closed-loop feedback mechanism enables the system to continuously adjust and optimize the flow control strategy until the actual flow reaches or approaches the set value, thereby achieving precise control of the flow.

[0022] In a preferred embodiment, the present invention can be further configured as follows: the intelligent pressure regulator has a built-in integrated adaptive algorithm, which can automatically adjust the control strategy according to the actual changes in the system pressure to achieve accurate control of the system pressure. The algorithm can gradually adapt to the dynamic changes of the system through continuous learning and optimization, and improve the stability and reliability of the system. When the system pressure deviates from the set value, the intelligent pressure regulator will automatically adjust the control quantity such as the opening of the regulating valve according to the size and direction of the deviation to restore the system pressure to the set value. At the same time, the algorithm will also take into account the dynamic characteristics and hysteresis of the system to ensure that the adjustment process is smooth, fast and accurate. The user can also set the system pressure through the user interface of the central control unit. The set value of the system pressure. Once the set value is determined, the intelligent pressure regulator will control the pressure according to this value. Through the intelligent pressure regulator with integrated adaptive algorithm, the system pressure can be kept within the range of ±0.05MPa of the set value. This range is set according to the system requirements and accuracy requirements to ensure the stability and consistency of the system pressure. The system also has fault monitoring and alarm functions. When a sensor or regulator fails, the system will automatically send out an alarm signal and take corresponding protective measures such as cutting off the power supply and closing the valve to prevent the fault from expanding and causing damage to the system. The built-in integrated adaptive algorithm of the intelligent pressure regulator can monitor the actual changes of the system pressure in real time and These changes automatically adjust the control strategy. This adaptive ability enables the system to flexibly respond to various pressure fluctuations. Whether caused by changes in external loads or fluctuations in internal component performance, precise control can be achieved through algorithm optimization and adjustment. Through continuous learning and optimization, the algorithm can gradually adapt to the dynamic changes of the system, improve the stability and reliability of the system, and thus ensure the smooth progress of the cleaning process. Secondly, when the system pressure deviates from the set value, the intelligent pressure regulator can respond quickly and automatically adjust the control amount according to the size and direction of the deviation, such as the opening of the regulating valve, to restore the system pressure to the set value. This adjustment process is not only fast, but also accurate, ensuring the stability and consistency of the system pressure. At the same time, the algorithm will fully consider the dynamic characteristics and hysteresis of the system. Through reasonable control logic and parameter settings, the adjustment process is smooth and fast, avoiding system oscillation or overshoot caused by improper adjustment. In addition, users can easily set the set value of system pressure through the user interface of the central control unit. This design enables operators to flexibly adjust the pressure set value according to actual needs and process requirements without complex physical adjustments or programming operations. Once the set value is determined, the intelligent pressure regulator will accurately control it according to the value to ensure that the system pressure always fluctuates within the set range. According to system requirements and accuracy requirements, the system pressure can be maintained at the set value ±0.05MPa. This high-precision control range ensures the stability and consistency of the cleaning effect. Finally, the system also has a complete fault monitoring and alarm function. When a key component such as a sensor or regulator fails, the system will automatically send out an alarm signal and take corresponding protective measures such as cutting off the power supply and closing the valve to prevent the failure from expanding and causing damage to the system. This design not only improves the safety and reliability of the system, but also reduces maintenance costs and downtime, providing users with a more convenient and efficient use experience.

[0023] In a preferred embodiment, the present invention can be further configured as follows: at the data processing and decision-making level, the system adopts a stable and reliable operating system such as embedded Linux or Android, supports multi-task concurrent processing, and by developing a dedicated data processing algorithm, the system can filter, detect anomalies and predict trends on the data collected by the sensor in real time to ensure the accuracy and reliability of the data. Subsequently, the decision support system automatically adjusts the system parameters and sends control instructions according to the processed data and preset thresholds and rules to realize the automated control process. First, the system adopts a stable and reliable operating system such as embedded Linux or Android, which not only has powerful multi-task concurrent processing capabilities, but also has high stability and security, providing a solid operating foundation for the system. The multi-task concurrent processing mechanism enables the system to process multiple data streams and tasks at the same time, improves data processing efficiency and response speed, and ensures the stable operation of the system in complex environments. Secondly, by developing a dedicated data processing algorithm, the system The system can filter, detect anomalies and predict trends of data collected by sensors in real time. The filtering algorithm can effectively remove noise and interference in the data, and improve the accuracy and reliability of the data. The anomaly detection algorithm can promptly detect and handle anomalies in the data to avoid the impact of erroneous data on system decisions. The trend prediction algorithm can predict future trends based on historical data and provide forward-looking decision support for the system. The comprehensive application of these data processing algorithms has significantly improved the system's data processing and analysis capabilities. Finally, the decision support system automatically adjusts system parameters and sends control instructions based on the processed data and preset thresholds and rules, realizing an automated control process. This innovative design enables the system to make autonomous decisions based on real-time data and preset rules, quickly respond to system changes, and automatically adjust parameters to optimize system performance. Compared with traditional manual intervention or simple threshold control methods, this intelligent decision support mechanism is more efficient, accurate and reliable, and significantly improves the system's automation level and cleaning effect.

[0024] In a preferred embodiment, the present invention can be further configured as follows: the system collects the operation data of the equipment in real time through various sensors deployed on the equipment, such as temperature, pressure, vibration, current and other sensors. The data collection frequency is set according to the equipment characteristics and maintenance requirements to ensure the comprehensiveness and accuracy of the data, clean the collected raw data, remove noise, outliers and redundant information, perform data compression and dimensionality reduction processing, improve the efficiency of subsequent processing, extract key features, such as time domain feature mean, variance, peak value, etc., frequency domain feature power spectrum, spectrum entropy, etc. and time-frequency domain feature wavelet transform coefficients, energy spectrum density, etc., use normal state and abnormal state samples in historical data to train the support vector machine (SVM) model, improve the classification performance and generalization ability of the SVM model through feature selection and optimization algorithms, set appropriate kernel functions and parameters to ensure that the model can accurately The system can accurately identify the abnormal state of the equipment. After the SVM model identifies the abnormal state, the decision tree model is used to further classify the abnormal state. The decision tree model builds classification rules according to the fault characteristics and subdivides the abnormal state into specific fault types. The classification accuracy and efficiency of the decision tree model are improved through pruning and optimization algorithms. The system realizes the comprehensive and real-time collection of equipment operation data through various sensors deployed on the equipment, such as temperature, pressure, vibration, current, etc. The data collection frequency is flexibly set according to the equipment characteristics and maintenance requirements to ensure the comprehensiveness and accuracy of the data. This improvement not only solves the problem of incomplete data collection, but also provides a solid foundation for subsequent data processing and analysis. Secondly, the system strictly cleans and preprocesses the collected raw data, and improves the data quality by removing noise, outliers and redundant information.Through data compression and dimensionality reduction, the complexity of the data is reduced and the efficiency of subsequent processing is improved. This step is crucial to ensure the accuracy and reliability of data analysis. Then, the system extracts a variety of key features of the data, including time domain features, frequency domain features, and time-frequency domain features. These features not only cover multiple aspects of equipment operation, but also provide rich information for subsequent abnormal state identification and fault classification. Through feature extraction, the system can have a more comprehensive understanding of the operating status of the equipment, providing strong support for subsequent analysis and decision-making. Then, the system uses the normal state and abnormal state samples in the historical data to train the support vector machine (SVM). M) model, through feature selection and optimization algorithms, improves the classification performance and generalization ability of the SVM model. This model can accurately identify the abnormal state of the equipment, making it possible to take maintenance measures in a timely manner. Finally, after the SVM model identifies the abnormal state, the system uses the decision tree model to further classify the abnormal state. The decision tree model constructs classification rules based on the fault characteristics and subdivides the abnormal state into specific fault types. Through pruning and optimization algorithms, the classification accuracy and efficiency of the decision tree model are improved. This step not only solves the problem of unclear fault classification, but also provides maintenance personnel with specific fault information, reducing the difficulty and cost of maintenance. ;

[0025] In a preferred embodiment, the present invention can be further configured as follows: based on the recognition results of the SVM and decision tree models, combined with the operation history and current status of the equipment, using the time series analysis algorithm to predict the future failure trend of the equipment, analyzing the prediction results, determining the possible time of the failure, and the degree of impact, according to the failure prediction results, formulating a preventive maintenance plan, determining the maintenance priority, time and required resources, evaluating the cost and benefit of the maintenance plan, ensuring the economy and feasibility of the maintenance activities, designing an intuitive and easy-to-use user interaction interface, displaying the real-time operation status of the equipment, the fault identification results and the maintenance plan, providing users with the functions of querying, modifying and confirming the maintenance plan, and automatically reporting to the system when the system detects a potential failure risk or the need to execute the maintenance plan. Maintenance reminders and system optimization suggestions are pushed to users through SMS, email or system notifications to ensure that users can obtain relevant information in a timely manner. According to user feedback and maintenance effects, the system's algorithms and models are continuously optimized, data collection and processing processes are improved, and the overall performance and stability of the system are improved. Through online learning and update mechanisms, the timeliness and accuracy of the model are maintained. First, based on the recognition results of SVM and decision tree models, the system uses time series analysis algorithms to further predict future failure trends of equipment. This innovation not only improves the accuracy and reliability of fault prediction, but also provides a scientific basis for the formulation of preventive maintenance plans. By analyzing the prediction results, the system can determine the possible time and impact of the failure, so as to provide targeted Formulate maintenance plans to reduce unplanned downtime and improve equipment utilization. Secondly, the preventive maintenance plan formulated based on the fault prediction results fully considers the maintenance priority, time and required resources. This plan not only ensures the economy and feasibility of maintenance activities, but also optimizes resource allocation and improves maintenance efficiency. At the same time, the system also has the function of evaluating the cost and benefit of maintenance plans to help users make more informed decisions. Furthermore, the system is designed with an intuitive and easy-to-use user interaction interface, which greatly improves the user experience. The interface not only displays the real-time operating status of the equipment, fault identification results and maintenance plans, but also provides users with the function of querying, modifying and confirming maintenance plans. This interactive design allows users to participate more conveniently. During the maintenance management process, the transparency and efficiency of maintenance work are improved. In addition, the system also has the function of automatically pushing maintenance reminders and system optimization suggestions. When the system detects potential failure risks or needs to execute maintenance plans, it can promptly remind users through text messages, emails or system notifications. This instant communication mechanism ensures that users can obtain relevant information and take corresponding measures in the first place, reducing the possibility of failures. Finally, the system maintains the timeliness and accuracy of the model through online learning and update mechanisms, continuously optimizes algorithms and models based on user feedback and maintenance effects, improves data collection and processing processes, and improves the overall performance and stability of the system. This continuous optimization capability enables the system to adapt to changing environments and needs.Always maintain its competitiveness and advancement.

[0026] The intelligent flow control type single-chip wet cleaning machine pipeline system integrates high-precision PID flow control module, pressure adaptive adjustment mechanism, automatic control system and intelligent fault diagnosis and preventive maintenance system, which work together to achieve efficient, accurate and reliable cleaning operations. The system first uses a high-precision PID flow control module to monitor the flow in real time using a high-precision turbine flowmeter, and combines an electric control valve with a closed-loop control function and a positional PID control algorithm to quickly and accurately adjust the flow to ensure that the flow at each stage in the entire cleaning process can be stabilized within the range of ±1% of the preset value, thereby ensuring the accuracy of the cleaning effect and process parameters. At the same time, the pressure adaptive adjustment mechanism deploys Honeywell 26PCA series piezoelectric pressure sensors at key nodes of the system. These sensors can provide high-precision pressure monitoring of ±0.1%FS to ensure the accuracy of the system pressure data. Through the intelligent pressure regulator with integrated adaptive algorithm, the system can automatically adjust to keep the pressure within the range of ±0.05MPa of the set value, effectively cope with pressure fluctuations, and ensure the stability of the cleaning process. The automatic control system is the core of the entire system and integrates ARM-based The embedded central control unit of the Cortex-A series processor has powerful data processing and decision-making capabilities. Through the 7-inch color touch screen user interface, the system not only supports graphical display of cleaning processes, real-time flow and pressure curves, but also provides early warning and alarm prompt functions, making operation simple and intuitive, and convenient for users to monitor and intervene in real time. The intelligent fault diagnosis and preventive maintenance system further enhances the intelligence level of the system. The system uses historical data to train the SVM model for abnormal state identification, and combines the decision tree model for fault classification, which significantly improves the accuracy and efficiency of fault identification. According to real-time sensor data and system operating status, the system predicts maintenance needs through built-in algorithm analysis. The system can automatically push maintenance prompts and system optimization suggestions to users, realizing the transition from passive maintenance to active prevention, reducing the occurrence of failures, extending the service life of equipment, and ensuring the continuity and efficiency of cleaning operations.

[0027] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0028] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. Intelligent flow control type single-chip wet cleaning machine pipeline system, characterized in that: It includes high-precision PID flow control module, pressure adaptive adjustment mechanism, automatic control system and intelligent fault diagnosis and preventive maintenance system; The high-precision PID flow control module adopts a high-precision turbine flowmeter, an electric regulating valve with closed-loop control function, and a PID control algorithm, specifically a position-based PID algorithm, to achieve fast and precise flow regulation, ensuring that the flow in each cleaning stage is stable within the preset value ±1% range; The pressure adaptive adjustment mechanism includes: installing Honeywell 26PCA series piezoelectric pressure sensors at key nodes to provide high-precision pressure monitoring of ±0.1%FS to ensure the accuracy of system pressure data, and using an intelligent pressure regulator with integrated adaptive algorithm to keep the system pressure within the set value ±0.05MPa range; The automatic control system integrates an embedded central control unit based on the ARM Cortex-A series processor to achieve rapid data processing and decision-making, and provides a 7-inch color touch screen user interface that supports graphical display of the cleaning process, real-time flow and pressure curves, and has early warning and alarm prompt functions, making the operation simple and intuitive; The intelligent fault diagnosis and preventive maintenance system uses historical data to train the SVM model for abnormal state identification, and combines it with the decision tree model for fault classification to improve the accuracy and efficiency of fault identification. Based on sensor data and system operating status, it uses built-in algorithm analysis to predict maintenance needs, automatically push maintenance prompts and system optimization suggestions to users, and achieve preventive maintenance.

2. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 1 is characterized in that: The turbine flowmeter uses Krohne Optiflux series high-precision turbine flowmeters. The flowmeter should be installed at key positions in the pipeline system to ensure that it can accurately measure and reflect the actual flow conditions in each cleaning stage. During installation, factors such as the flow direction, temperature, and pressure of the fluid need to be considered to ensure accurate measurement. The turbine flowmeter converts the flow into an electrical signal output through a built-in sensor. The signal is a 4-20mA analog signal or a digital signal. The signal is transmitted to the central control unit through a dedicated cable. The built-in signal processing module of the central control unit filters and amplifies the received signal to eliminate noise interference and improve data accuracy.

3. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 1 is characterized in that: The electric control valve uses Siemens SIPART PS2 series electric control valve. The valve needs to be selected according to the flow range, pressure level and medium characteristics of the pipeline system to ensure that the valve can meet the system requirements. The electric control valve has a built-in closed-loop control system, which receives the flow set value signal from the central control unit and compares it with the actual flow signal fed back by the turbine flowmeter. According to the comparison result, the closed-loop control system changes the flow by adjusting the valve opening until the actual flow reaches or approaches the set value, thereby achieving precise control of the flow.

4. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 1 is characterized in that: The PID control algorithm is implemented in an embedded central control unit. The logic and calculation process of the algorithm are realized by writing dedicated control software. The control software receives the flow signal from the turbine flowmeter and the flow set value signal from the user interface in real time, calculates the control quantity through the PID algorithm, and outputs it to the closed-loop control system of the electric control valve to achieve precise control of the flow.

5. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 1 is characterized in that: The intelligent pressure regulator has a built-in integrated adaptive algorithm, which can automatically adjust the control strategy according to the actual changes in the system pressure to achieve accurate control of the system pressure. The algorithm can gradually adapt to the dynamic changes of the system through continuous learning and optimization, and improve the stability and reliability of the system. When the system pressure deviates from the set value, the intelligent pressure regulator will automatically adjust the control quantity such as the opening of the regulating valve according to the size and direction of the deviation to restore the system pressure to the set value. At the same time, the algorithm will also consider the dynamic characteristics and hysteresis of the system to ensure that the adjustment process is smooth, fast and accurate. The user can also set the set value of the system pressure through the user interface of the central control unit. Once the set value is determined, the intelligent pressure regulator will perform pressure control according to the value. Through the intelligent pressure regulator with an integrated adaptive algorithm, the system pressure can be kept within the range of ±0.05MPa of the set value. This range is set according to system requirements and accuracy requirements, ensuring the stability and consistency of the system pressure. The system also has fault monitoring and alarm functions. When a sensor or regulator fails, the system will automatically send an alarm signal and take corresponding protective measures such as cutting off the power supply and closing the valve to prevent the failure from expanding and causing damage to the system.

6. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 1 is characterized in that: At the data processing and decision-making level, the system uses stable and reliable operating systems such as embedded Linux or Android, supports multi-task concurrent processing, and through the development of dedicated data processing algorithms, the system can filter, detect anomalies and predict trends in real time on data collected by sensors to ensure data accuracy and reliability. Subsequently, the decision support system automatically adjusts system parameters and sends control instructions based on the processed data and preset thresholds and rules to realize automated control processes.

7. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 1 is characterized in that: The automatic adjustment system collects the operation data of the equipment in real time through various sensors deployed on the equipment, such as temperature, pressure, vibration, current and other sensors. The data collection frequency is set according to the characteristics of the equipment and maintenance requirements to ensure the comprehensiveness and accuracy of the data. The collected raw data is cleaned to remove noise, abnormal values ​​and redundant information, and data compression and dimensionality reduction processing are performed to improve the efficiency of subsequent processing. Key features such as time domain feature mean, variance, peak value, frequency domain feature power spectrum, spectrum entropy, etc. and time-frequency domain feature wavelet transform coefficients, energy spectrum density, etc. are extracted. The normal state and abnormal state samples in the historical data are used to train the support vector machine (SVM) model. The classification performance and generalization ability of the SVM model are improved through feature selection and optimization algorithms. Appropriate kernel functions and parameters are set to ensure that the model can accurately identify the abnormal state of the equipment. After the SVM model identifies the abnormal state, the decision tree model is used to further classify the abnormal state. The decision tree model constructs classification rules according to the fault characteristics and subdivides the abnormal state into specific fault types. The classification accuracy and efficiency of the decision tree model are improved through pruning and optimization algorithms.

8. The intelligent flow control type single-wafer wet cleaning machine pipeline system according to claim 7 is characterized in that: Based on the recognition results of SVM and decision tree models, combined with the operation history and current status of the equipment, the time series analysis algorithm is used to predict the future failure trend of the equipment, analyze the prediction results, determine the possible time of the failure, and determine the impact of the failure. According to the failure prediction results, a preventive maintenance plan is formulated to determine the priority, time and required resources of maintenance, evaluate the cost and benefit of the maintenance plan, ensure the economy and feasibility of maintenance activities, design an intuitive and easy-to-use user interaction interface, display the real-time operation status, fault identification results and maintenance plan of the equipment, and provide users with the functions of querying, modifying and confirming maintenance plans. When the system detects potential failure risks or needs to execute maintenance plans, it automatically pushes maintenance reminders and system optimization suggestions to users, and ensures that users can obtain relevant information in a timely manner through SMS, email or system notifications. According to user feedback and maintenance effects, the system's algorithms and models are continuously optimized, the data collection and processing processes are improved, and the overall performance and stability of the system are improved. Through online learning and update mechanisms, the timeliness and accuracy of the model are maintained.

Images