Intelligent closed sampling safety guarantee system for diethyl zinc

Abstract

The invention relates to the field of chemical sampling safety guarantee, in particular to an intelligent closed sampling safety guarantee system for diethyl zinc. The device comprises a magnetofluid sealing assembly, an intelligent deformation sealing assembly, a gas replacement protection layer, a double-wall structure negative pressure isolation cavity and a condensation trapping assembly which are arranged in sequence, each assembly forms a dynamic complementary enhancement loop through a state sensing unit comprising a quantum cascade laser detection module and an electrochemical array detection module and a control unit integrated with a fuzzy control algorithm, and when the performance parameter of any assembly deviates from a threshold value, a grading compensation strategy is automatically triggered. And the specific composition and parameters of each component are limited. The technical effects that the safety of the diethyl zinc sampling process is effectively guaranteed, dynamic complementary enhancement is achieved through multi-assembly cooperation and intelligent control, assembly performance changes can be coped with in time, the leakage condition can be accurately detected, emergency response can be triggered, and meanwhile the residual life of the material can be accurately predicted are achieved.

Description

Technical Field

[0001] This application relates to the field of chemical sampling safety assurance, and in particular to a smart closed sampling safety assurance system for diethylzinc. Background Technology

[0002] In the field of chemical production and testing, diethylzinc (DEZ) is an important chemical raw material with a wide range of applications, playing a key role in industries such as organic synthesis and semiconductor manufacturing. With the continuous development of the chemical industry, the demand for quality testing and safety management of DEZ is increasing. Accurately and safely obtaining DEZ samples is crucial for ensuring product quality, production safety, and meeting environmental protection requirements. This not only relates to the production efficiency and economic benefits of enterprises but is also closely related to the sustainable development of the entire industry.

[0003] In traditional diethylzinc sampling, various methods are typically employed to prevent leakage and ensure sampling safety. One common method is the use of fixed physical barriers, such as constructing barriers using common sealing materials and structures to prevent leakage. Simultaneously, single detection technologies, such as simple gas detection devices, are also used to monitor for leaks. Furthermore, the operation often relies on operator experience and fixed procedures, lacking the ability to adapt to different operating conditions and sample characteristics.

[0004] However, these traditional methods have significant drawbacks. Traditional fixed barriers lack dynamic adjustment capabilities and cannot cope with sudden minor failures; once a barrier malfunctions, it can easily trigger a leakage risk. Single detection technologies are susceptible to environmental interference, leading to false alarms or missed alarms, and cannot accurately identify diethylzinc leaks. Furthermore, due to the lack of accurate prediction of sealing component lifespan, in corrosive environments, the lifespan estimate of sealing components is inaccurate, potentially leading to sudden failure during use. Additionally, traditional operating methods cannot adaptively adjust to batch variations and characteristics of samples, easily resulting in contamination or detection errors. Summary of the Invention

[0005] The purpose of this application is to overcome the above-mentioned technical problems and provide a smart, closed-loop sampling safety assurance system for diethylzinc. A diethylzinc intelligent sealed sampling safety assurance system includes a magnetohydrodynamic sealing component, an intelligent deformation sealing component, a gas replacement protective layer, a double-wall structure negative pressure isolation chamber, and a condensation trapping component arranged in sequence. Each component forms a dynamic complementary enhancement loop through a state sensing unit containing a quantum cascade laser detection module and an electrochemical array detection module, and a control unit integrating a fuzzy control algorithm. When the performance parameter of any component deviates from the threshold, a graded compensation strategy is automatically triggered.

[0006] By adopting the above technical solutions, a dynamic complementary five-layer sealed barrier is constructed, forming an adaptive protection system of "state awareness - dynamic enhancement". This solves the leakage risk caused by sudden minor failures and achieves the four-dimensional sampling goal of "zero exposure, zero pollution, zero accidents, and high reliability", reducing the risk of single barrier failure and the probability of system leakage. The quantum cascade laser-electrochemical array fusion leakage detection technology realizes dual-modal detection, accurately identifies DEZ bulk and reaction products, and solves the problem of false alarms caused by environmental interference. The fuzzy control algorithm can automatically activate the hierarchical complementary strategy when the performance of a certain barrier deteriorates, ensuring the safe and stable operation of the system.

[0007] Preferably, the magnetic fluid sealing assembly includes a neodymium iron boron annular permanent magnet, a titanium alloy magnetic fluid reservoir, and a magnetic field gradient controller of model MGC-200. The reservoir is filled with Fe3O4 nanoparticle magnetic fluid containing 5% fluorocarbon anti-corrosion additive. The differential pressure sensor is of model with an accuracy of 0.001MPa and the working differential pressure is maintained within the range of -0.1 to 0.5MPa through a closed-loop control system.

[0008] By adopting the above technical solution, the neodymium iron boron annular permanent magnet, titanium alloy magnetic fluid reservoir, and magnetic field gradient controller in the magnetic fluid sealing assembly work together. The reservoir is filled with Fe3O4 nanoparticle magnetic fluid containing fluorocarbon anti-corrosion additives, which can enhance the sealing performance. The high-precision differential pressure sensor and closed-loop control system maintain the working differential pressure within the range of -0.1~0.5MPa, ensuring stable system operation and reducing the risk of single barrier failure and system leakage probability.

[0009] Preferably, the intelligent deformation sealing assembly is made of a composite material of polyimide-based shape memory polymer and conductive carbon nanotubes, and has a built-in fiber optic deformation sensor connected to a nickel-chromium heating wire and a PID temperature control system. The temperature control system has control parameters of proportional coefficient 0.8 and integral time 10 seconds, which can make the diameter shrinkage range of the sealing ring 0-5mm and the response time ≤30 seconds.

[0010] By adopting the above technical solution, the intelligent deformation sealing component uses a composite material of polyimide-based shape memory polymer and conductive carbon nanotubes, and has a built-in fiber optic deformation sensor to monitor the deformation of the sealing ring. When connected to a nickel-chromium heating wire and a PID temperature control system with specific control parameters, the sealing ring diameter shrinkage range is 0-5mm with a response time ≤30 seconds. Combined with a dynamic complementary reinforcement circuit, it can adaptively reinforce when a single barrier fails slightly, reducing the risk of single barrier failure and reducing the probability of system leakage.

[0011] Preferably, the gas replacement protective layer includes an electromagnetic flow regulating valve of model KV-10 and an OXY-910 oxygen content analyzer. The flow regulating valve can adaptively adjust the flow rate of ultrapure argon gas in the range of 1-5L / min. The micro positive pressure maintenance system controls the pressure in the cavity at +50~100Pa and the oxygen content is less than 1ppm through a pressure sensor.

[0012] By adopting the above technical solution, based on the sequential arrangement of a magnetic fluid sealing component, an intelligent deformation sealing component, a gas replacement protective layer, a double-wall structure negative pressure isolation chamber, and a condensation trapping component, and forming a dynamic complementary enhancement circuit, the gas replacement protective layer uses an electromagnetic flow regulating valve to adaptively adjust the flow rate of ultrapure argon gas within the range of 1-5 L / min. In conjunction with an oxygen content analyzer and a pressure sensor, the pressure inside the chamber is controlled at +50-100 Pa and the oxygen content is below 1 ppm. This can solve the problem of contamination or detection error caused by batch differences, prevent diethylzinc leakage, and reduce the probability of system leakage.

[0013] Preferably, the double-walled negative pressure isolation chamber is made of 316L stainless steel with an inner wall thickness of 5mm and an outer wall thickness of 3mm. The molecular pump and mechanical pump linkage system, in conjunction with the vacuum gauge for real-time monitoring, can adjust the vacuum level of the inner chamber within the range of 0.5-1Pa and automatically alarm when the pressure rise rate exceeds 0.1Pa / min.

[0014] By adopting the above technical solution, the double-walled negative pressure isolation chamber is made of 316L stainless steel of a specific thickness. With the linkage system of molecular pump and mechanical pump and real-time monitoring by vacuum gauge, the vacuum degree of the inner chamber can be adjusted, and an automatic alarm is triggered when the pressure rise rate exceeds the threshold, which enhances the sealing and safety of the system and reduces the risk of leakage.

[0015] Preferably, the condensation and collection assembly is equipped with a liquid nitrogen cooling module and an electric heating regeneration module. The cooling module achieves controllable cooling of -196 to -200°C through a TC-300 temperature controller. The regeneration module uses a 120°C desorption temperature combined with anhydrous ethanol solvent for cleaning, with a regeneration cycle of ≤15 minutes and a collection efficiency of ≥99.995%.

[0016] By adopting the above technical solution, the liquid nitrogen cooling module of the condensation and collection component can achieve controllable cooling from -196 to -200℃ using a temperature controller. Combined with the electric heating regeneration module, which uses a 120℃ desorption temperature and anhydrous ethanol solvent for cleaning, the regeneration cycle is ≤15 minutes and the collection efficiency is ≥99.995%, which can effectively treat diethylzinc and improve the safety and reliability of the system.

[0017] Preferably, the control unit integrates a multiphysics coupling model that includes four sub-models: fluid-heat transfer-stress-chemical reaction. The fluid model adopts the k-ω SST turbulence model of ANSYS Fluent. The remaining life prediction of the sealing component is corrected by real-time data interaction of temperature field, concentration field and mechanical field, with a prediction accuracy of ±5%.

[0018] By adopting the above technical solution, a multiphysics coupling model is integrated, which includes four sub-models: fluid, heat transfer, stress, and chemical reaction. The fluid model adopts the k-ω SST turbulence model of ANSYS Fluent. The remaining life prediction of the sealing component can be corrected by real-time data interaction of temperature field, concentration field, and mechanical field, so that the prediction accuracy reaches ±5%, realizing accurate quantitative prediction of the life of sealing components and solving the problem of inaccurate life estimation in corrosive environments.

[0019] Preferably, the material health index is calculated using a weighted average algorithm, which integrates three parameters: corrosion rate (Hastelloy ≤ 0.01 mm / year), deformation rate (≤ 5%), and stress value (≤ 300 MPa), with weights of 0.4, 0.3, and 0.3, respectively, and uses LSTM network training to predict remaining lifetime.

[0020] By adopting the above technical solution and using a weighted average algorithm to calculate the material health index, the material health status can be comprehensively and accurately assessed by comprehensively considering three parameters: corrosion rate, deformation rate, and stress value, and assigning them corresponding weights. Based on LSTM network training, the remaining life prediction can be realized, which can accurately predict the remaining life of sealing components with a prediction accuracy of ±5%, identify potential risks in advance, provide a scientific basis for the maintenance and replacement of sealing components, and ensure the stable operation of the diethylzinc intelligent sealed sampling safety assurance system.

[0021] Preferably, the quantum cascade laser detection module uses a QCL laser with a center wavelength of 10.6 μm in conjunction with an MCT detector, the electrochemical array detection module detects Zn²⁺ and ethanol concentrations through a three-electrode system, and the DS evidence theory fusion algorithm assigns a weight of 0.6 to QCL detection and a weight of 0.4 to electrochemical detection. An emergency response is triggered when the leakage confidence threshold is ≥0.8.

[0022] By adopting the above technical solution, the quantum cascade laser detection module uses a QCL laser with a center wavelength of 10.6μm in conjunction with an MCT detector, while the electrochemical array detection module detects Zn²⁺ and ethanol concentrations through a three-electrode system. The combination of these two technologies achieves dual-modal detection, enabling accurate identification of DEZ bulk and reaction products. The D-S evidence theory fusion algorithm assigns a weight of 0.6 to QCL detection and 0.4 to electrochemical detection, achieving high-precision leak identification and reducing false alarms caused by environmental interference. When the leak confidence threshold is ≥0.8, an emergency response is triggered, allowing for timely handling of leaks and ensuring system safety.

[0023] Preferably, the fuzzy control algorithm adopts a triangular membership function, and the input parameters include the magnetohydrodynamic pressure difference attenuation rate (judgment threshold 5%-15%), the deformation rate decrease value (judgment threshold 3%-8%), and the oxygen content exceeding the standard value (judgment threshold 1.0-2.0ppm). The output compensation strategy includes three types of control actions: increasing the magnetic field strength by 0.1T, reducing the vacuum degree to 0.3Pa, and reducing the condensation temperature to -200℃, with a response time ≤5 seconds.

[0024] By adopting the above technical solution, the fuzzy control algorithm uses a triangular membership function with the magnetohydrodynamic pressure differential attenuation rate, deformation rate decrease, and oxygen content exceeding the standard as input parameters. It can accurately judge the situation based on the performance changes of each component. The output includes compensation strategies for increasing magnetic field strength, decreasing vacuum, and decreasing condensation temperature. It can automatically trigger a graded compensation strategy when the component performance parameters deviate from the threshold, realize adaptive reinforcement when a single barrier fails slightly, reduce the risk of single barrier failure and the probability of system leakage, and respond to component performance degradation issues in no more than 5 seconds.

[0025] In summary, this application includes at least one of the following beneficial technical effects: 1. The dynamic complementary five-layer airtight barrier technology constructs an adaptive protection system of "state perception - dynamic enhancement". It adds a closed-loop mechanism of "barrier state perception - dynamic complementary enhancement" and analyzes the data of each barrier sensor through fuzzy control algorithm. When the performance of a certain barrier deteriorates, the complementary strategy is automatically activated to reduce the risk of single barrier failure and the probability of system leakage. 2. Intelligent sealing life prediction technology integrating materials and digital twins integrates real-time material health monitoring and digital twin simulation to achieve accurate quantitative prediction of the life of sealing components and solve the problem of inaccurate life estimation in corrosive environments; 3. The working condition adaptive sampling strategy optimization technology is based on the "one vehicle, one policy" operation mode with real-time perception of sample characteristics, which solves the problem of contamination or detection error caused by batch differences; 4. Quantum cascade laser (QCL) - electrochemical array fusion leakage detection technology enables dual-mode detection, accurately identifying DEZ bulk and reaction products, and solving the problem of false alarms caused by environmental interference. Attached Figure Description

[0026] Figure 1 System overall architecture and dynamic complementary enhancement logic diagram.

[0027] Figure 2 : Logic diagram of dual-modal leakage detection and graded compensation.

[0028] Figure 3 Logic diagram for material health monitoring and remaining life prediction.

[0029] Figure 4 : Full lifecycle traceability logic diagram. Detailed Implementation

[0030] The technical solutions in the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. The described embodiments are only possible technical implementations of the present invention, but are not limited thereto. Other embodiments obtained by those skilled in the art in conjunction with the embodiments of the present invention without creative effort are also within the protection scope of the present invention.

[0031] This application mainly adopts a scheme of setting up a five-fold sealed component and a dynamic complementary enhancement circuit, which achieves the effect of improving the sampling safety of diethylzinc and reducing the risk of leakage. The following is a further detailed description of this application. Example

[0032] The intelligent closed sampling safety system for diethylzinc provided in this application includes a magnetohydrodynamic sealing component, an intelligent deformation sealing component, a gas replacement protective layer, a double-walled negative pressure isolation chamber, and a condensation trapping component arranged sequentially. Each component forms a dynamic complementary enhancement loop through a state sensing unit containing a quantum cascade laser detection module and an electrochemical array detection module, and a control unit integrating a fuzzy control algorithm. When the performance parameter of any component deviates from the threshold, a graded compensation strategy is automatically triggered. This allows for timely compensation when the performance of a component degrades, improving the safety and reliability of the entire system and reducing the risk of diethylzinc leakage.

[0033] Specifically, the magnetohydrodynamic (MHD) sealing assembly comprises a neodymium iron boron (NdFeB) ring-shaped permanent magnet, a titanium alloy MHD reservoir, and a magnetic field gradient controller (model MGC-200). The NdFeB ring-shaped permanent magnet has high remanence, reaching up to 1.2T, providing a stable magnetic field. Its ring structure ensures a more uniform magnetic field distribution. Alternatively, other permanent magnet materials with high remanence, such as samarium cobalt (SMC), can be used. The titanium alloy MHD reservoir, with a volume of 50 mL, stores the MHD fluid and is made of titanium alloy, offering excellent corrosion resistance. Reservoirs made of other corrosion-resistant metals, such as stainless steel, can also be used. The reservoir is filled with Fe3O4 nanoparticle MHD fluid containing 5% fluorocarbon-based anti-corrosion additives. This MHD fluid consists of Fe3O4 nanoparticles (10 nm, 99.9% purity), a perfluoropolyether carrier fluid (viscosity 20 mm² / s, 25°C), and fluorocarbon-based anti-corrosion additives. Fe3O4 nanoparticles provide excellent sealing under magnetic field conditions, while the perfluoropolyether carrier fluid offers a suitable environment. Fluorocarbon-based anti-corrosion additives enhance the corrosion resistance of the magnetic fluid. An HMC5883L magnetic field strength sensor monitors the magnetic field gradient in real time with a sampling frequency of 100Hz, accurately tracking magnetic field changes. A differential pressure sensor with an accuracy of 0.001MPa maintains the operating differential pressure within the range of -0.1 to 0.5MPa through a closed-loop control system, ensuring real-time monitoring of the sealing assembly's differential pressure and guaranteeing a tight seal. This magnetic fluid sealing assembly achieves a sealing accuracy of 10⁻¹² Pa·m³ / s, an operating temperature range of -40 to 80℃, a self-healing time of <1 second, a magnetic field strength monitoring accuracy of ±0.01T, and a continuous operating life of >8000 hours. These components, combined with a permanent magnet providing the magnetic field, enable the magnetic fluid to form a zero-gap liquid seal. This, along with the differential pressure sensor, magnetic field strength sensor, and magnetic field gradient controller, achieves a stable sealing function.

[0034] Specifically, the intelligent deformation sealing component is made of a composite material of polyimide-based shape memory polymer and conductive carbon nanotubes. The polyimide-based shape memory polymer possesses shape memory properties and a glass transition temperature of 60°C. The conductive carbon nanotube network (resistivity 10⁻³Ω·cm) increases the material's conductivity and strength. An integrated FBG-1550 fiber optic deformation sensor is connected to a nichrome heating wire and a PID temperature control system. The fiber optic deformation sensor monitors the deformation of the sealing ring in real time, while the nichrome heating wire heats the ring to induce deformation. The temperature control system has control parameters of a proportional coefficient of 0.8, an integral time of 10 seconds, and a derivative time of 2 seconds, allowing the sealing ring diameter to shrink within a range of 0-5 mm with a response time ≤30 seconds. Alternatively, other polymer materials with shape memory functions, such as polyurethane-based shape memory polymers, can be used in other solutions. This intelligent deformation sealing component has a maximum sealing pressure of 2MPa, a cycle life of >1 million cycles, a deformation monitoring accuracy of ±0.01mm, and a DEZ corrosion resistance rate of ≤0.001mm / year. This component enhances the sealing effect by controlling the deformation of the sealing ring through temperature control.

[0035] Specifically, the gas replacement protective layer includes a KV-10 electromagnetic flow control valve and an OXY-910 oxygen analyzer. The electromagnetic flow control valve adaptively adjusts the ultrapure argon flow rate within the range of 1-5 L / min, allowing for adjustments based on actual conditions. The oxygen analyzer monitors the oxygen content within the chamber in real time. A micro-positive pressure maintenance system uses a pressure sensor to control the chamber pressure at +50~100 Pa and maintain an oxygen content below 1 ppm. Ultrapure argon (99.999% purity, dew point ≤ -70℃) serves as the protective medium, effectively preventing the reaction between diethylzinc and oxygen. The replacement mode consists of initial three pulse replacements (5 L / min each, lasting 10 seconds) followed by continuous micro-replenishment (basic flow rate 1 L / min), with the flow rate adaptively adjustable from 1-5 L / min. In other cases, other inert gases, such as helium, can also be used as the protective medium. Flow regulation and pressure control ensure effective protection.

[0036] Specifically, the double-walled negative pressure isolation chamber is constructed of 316L stainless steel with an inner wall thickness of 5mm and an outer wall thickness of 3mm. 316L stainless steel possesses excellent corrosion resistance and strength. Equipped with a PG-200 vacuum gauge for real-time monitoring and a DP-80 molecular pump + 2XZ-4 mechanical pump linkage system, the inner chamber vacuum level can be adjusted within the range of 0.5-1Pa, while the outer chamber vacuum level is <10Pa. An automatic alarm is triggered if the pressure rise rate exceeds 0.1Pa / min, and the evacuation time is ≤5 minutes (from atmospheric pressure to 1Pa). In some alternative solutions, other metal materials with good corrosion resistance, such as titanium alloy, can be used to construct the isolation chamber. This structure, by creating negative pressure through evacuation, further enhances the isolation effect.

[0037] Specifically, the condensation and collection assembly is equipped with a liquid nitrogen cooling module and an electrically heated regeneration module. The cooling module uses a TC-300 temperature controller to achieve controllable cooling from -196 to -200℃, condensing diethylzinc vapor. The regeneration module uses a 120℃ desorption temperature combined with 99.99% anhydrous ethanol solvent cleaning (flow rate 10 mL / min), and is equipped with a DEZ-100 collection efficiency monitoring instrument, achieving a regeneration cycle of ≤15 minutes and a collection efficiency ≥99.995%. Other cooling and regeneration methods can also be used in other schemes, such as dry ice cooling and hot nitrogen purging regeneration. Through cooling and regeneration, effective collection of diethylzinc is achieved.

[0038] Specifically, the control unit integrates a multiphysics coupling model that includes four sub-models: fluid-heat transfer-stress-chemical reaction. The fluid model uses the k-ω SST turbulence model from ANSYS Fluent, incorporating real-time data from a VM-100 viscosity sensor (sampling frequency 10Hz) for correction, with an iteration step of 0.001s. The heat transfer model uses COMSOL Multiphysics to establish a transient heat transfer model (thermal conductivity: Hastelloy 15W / (m·K), seal 0.2W / (m·K)), combined with real-time data from a PT1000 temperature sensor (accuracy ±0.1℃), and is linked with the fluid model through a temperature field coupling interface. The stress model uses ABAQUS to establish a material corrosion stress model (corrosion expansion coefficient 1.05), importing stress data from an FBG sensor (measurement range 0-500MPa, accuracy ±1MPa), and is linked with the heat transfer model through a thermal stress coupling interface. The chemical reaction model is based on the Arrhenius equation to establish a kinetic model of the DEZ reaction with water and oxygen (reaction activation energy 80kJ / mol), calibrated with real-time data from a leak detection system, and linked with the fluid model through a concentration field coupling interface. The remaining life prediction of the sealing component is corrected through real-time data interaction of temperature, concentration, and mechanical fields, achieving a prediction accuracy of ±5%. The Material Health Monitoring (PHM) system uses FBG sensors (model FBG-1550, wavelength resolution ≤0.1pm, measurement range 0-2000με) installed via a combination of surface bonding and internal implantation, with one sensor placed per square centimeter at key stress points on the sealing component. Armored, corrosion-resistant optical fibers are used for transmission. Data processing utilizes the Python Scikit-learn library to construct the Material Health Index (MHI, 0-1), integrating parameters such as corrosion rate (Hastelloy ≤0.01mm / year), deformation rate (≤5%), and stress value (≤300MPa). A weighted average algorithm (weights of 0.4, 0.3, and 0.3) is employed, and the remaining life prediction is achieved based on LSTM network training. Through comprehensive calculation and analysis of these parameters, the remaining life of the sealing component can be predicted more accurately.

[0039] Specifically, the quantum cascade laser detection module uses a QCL laser with a center wavelength of 10.6 μm in conjunction with an MCT detector to accurately detect diethylzinc bulk. The electrochemical array detection module uses a three-electrode system (working electrode: Pt / C nanocomposite material, area 0.5 cm²; reference electrode: Ag / AgCl (saturated KCl); counter electrode: platinum wire) to detect Zn²⁺ and ethanol concentrations, and can detect the reaction products of diethylzinc. The D-S evidence theory fusion algorithm assigns a weight of 0.6 to QCL detection and 0.4 to electrochemical detection, triggering an emergency response when the leakage confidence threshold is ≥0.8. This dual-modal detection and fusion algorithm improves the accuracy of leak detection.

[0040] Specifically, the fuzzy control algorithm uses a triangular membership function. The input parameters include the magnetohydrodynamic pressure differential attenuation rate (first-level judgment threshold: decrease of 5% - 10%; second-level judgment threshold: >10% - 15%), the deformation rate decrease value (first-level judgment threshold: decrease of 3% - 5%; second-level judgment threshold: >5% - 8%), and the oxygen content exceeding the standard value (first-level judgment threshold: 1.0 - 1.5ppm; second-level judgment threshold: >1.5 - 2.0ppm). The output compensation strategy includes three types of control actions: increasing the magnetic field strength by 0.1T, reducing the vacuum degree to 0.3Pa, and reducing the condensation temperature to -200℃. The response time is ≤5 seconds. When the magnetohydrodynamic sealing differential pressure decreases by one level (5% - 10%), the argon flow rate is increased to 2L / min; for the second level (>10% - 15%), it is increased to 3L / min + the magnetic field strength is increased by 0.1T. The oxygen content must be stable <1ppm, the sealing differential pressure must recover to ≥80% of the rated value, the first-level response time <0.5 seconds, and the second-level response time <1 second. When the sealing ring deformation rate decreases by one level (3% - 5%), the vacuum degree of the negative pressure chamber drops to 0.5Pa; for the second level (>5% - 8%), the vacuum degree is 0.3Pa + the sealing ring heating temperature is increased by 5℃. The pressure inside the chamber must be stable <0.5Pa, the deformation rate must recover to ≥90% of the rated value, the first-level response time <1 second, and the second-level response time <2 seconds. When the oxygen content exceeds the standard at Level 1 (1.0 - 1.5 ppm), the condensation temperature drops to -200℃; at Level 2 (>1.5 - 2.0 ppm), the gas is replaced twice with argon pulses at -200℃, requiring the exhaust gas DEZ concentration to be <0.05 ppm, the oxygen content to drop back to <1 ppm, the Level 1 response time to be <3 seconds, and the Level 2 response time to be <5 seconds. Through a fuzzy control algorithm, timely compensation measures can be taken based on the performance changes of each component.

[0041] The implementation principle of this embodiment is as follows: The system, through the setting of five layers of sealed components, forms a multi-layered protection system, effectively preventing the leakage of diethylzinc. The dynamic complementary enhancement loop formed by the state sensing unit and the control unit can monitor the performance of each component in real time. When the performance of a component deteriorates, a graded compensation strategy is triggered in a timely manner, improving the reliability and safety of the system. The multiphysics coupling model and material health index calculation method can accurately predict the remaining life of the sealing components, facilitating early maintenance and replacement. Dual-modal detection and fusion algorithms improve the accuracy of leak detection and reduce the probability of false alarms and missed alarms. The fuzzy control algorithm can quickly respond to changes in component performance and take timely and effective compensation measures, ensuring the stable operation of the system. Compared with traditional diethylzinc sampling systems, this embodiment has significantly improved in terms of safety, reliability, and intelligence, solving the problems of high leakage risk, inaccurate life estimation, and large detection errors existing in traditional systems. Example

[0042] The difference between this embodiment and the above embodiments is as follows: In the magnetohydrodynamic sealing assembly, the permanent magnet is replaced by a different permanent magnet material with high remanence, such as samarium cobalt permanent magnets; the liquid storage tank is replaced by a different corrosion-resistant metal material, such as stainless steel, instead of titanium alloy; the intelligent deformation sealing assembly is replaced by a different polymer material with shape memory function, such as polyurethane-based shape memory polymers; the gas replacement protective layer is replaced by a different inert gas, such as helium, instead of ultrapure argon; the double-walled negative pressure isolation chamber is replaced by a different metal material with good corrosion resistance, such as titanium alloy; and the condensation trapping assembly is replaced by a different cooling method and regeneration method, such as dry ice cooling and hot nitrogen purging regeneration, instead of liquid nitrogen cooling and electric heating regeneration.

[0043] The implementation principle of this embodiment is as follows: although some components use replaceable materials and methods, the overall system architecture and working principle remain unchanged. By using different materials and methods, costs may be reduced or specific performance may be improved in certain situations. For example, using other permanent magnet materials may be more cost-effective, and using other cooling methods may result in lower energy consumption. At the same time, the dynamic complementary enhancement loop and control strategy can still ensure that the system takes timely compensatory measures when component performance degrades, maintaining the system's safety and reliability. Compared with traditional systems, it can still solve problems such as high leakage risk and inaccurate lifespan estimation, and can be optimized to a certain extent according to actual needs. Example

[0044] This embodiment, based on the above embodiments, adds a non-contact magnetically driven sampler and an adaptive sampling system. The non-contact magnetically driven sampler adopts a magnetic coupling non-contact transmission method to avoid the risk of leakage at dynamic sealing points. Key components are made of DEZ corrosion-resistant materials and are equipped with a precision metering and contamination control system. Its drive system consists of a neodymium iron boron external magnet (magnetic field strength 0.5 - 1.2T, adjustment accuracy 0.01T) driving an internal samarium cobalt magnetic rotor, equipped with a RS-200 speed sensor for real-time monitoring (accuracy ±1rpm), and a transmission efficiency >95%. The sampling unit is composed of a BT-100 precision peristaltic pump (flow range 0.1 - 10mL / min, accuracy ±0.2%) and an MV-50 quantitative metering valve (repeatability ±0.1%), and the sampling volume can be continuously adjusted within the range of 1μL - 10mL. The DEZ contact components are made of Hastelloy C-276 (solution treated, hardness HB200), the sealing components are made of perfluoroether rubber (Shore hardness 70A, resistant to DEZ immersion for 1000 hours without swelling), and the tubing is made of 316L stainless steel (inner wall polished Ra≤0.2μm). This non-contact magnetically driven sampler has a minimum sampling volume of 1μL, an accuracy of ±0.5%, a repeatability RSD <1%, cross-contamination controllable to <0.005%, and no performance degradation after ≥100 consecutive samplings.

[0045] The adaptive sampling system integrates a sample characteristic sensing module and an adaptive control unit to dynamically optimize sampling parameters, paths, and cleaning procedures. Sample characteristic sensing utilizes an NIR-5000 near-infrared spectroscopy online detection module with a wavelength range of 1200-2500nm (8cm⁻¹ resolution), equipped with a fiber optic probe (corrosion-resistant Hastelloy material). It analyzes key indicators in real time, including DEZ purity (≥99.99%, detection accuracy ±0.001%), moisture (≤10ppm, detection limit 0.1ppm), and chloride (≤5ppm, detection limit 0.05ppm). The detection time is <3 seconds, and the data update frequency is 1Hz. Regarding the parameter adaptation strategy, the impurity adaptation is as follows: when the moisture content is 10-15 ppm, a first-level enhanced cleaning is initiated (5 solvent rinses + 3 minutes of vacuum drying); when the moisture content is >15 ppm, a second-level enhanced cleaning is initiated (7 rinses + 5 minutes of drying); the purity adaptation is as follows: when the purity is 99.99%-99.999%, a standard flow rate of 1 mL / min is used; when the purity is ≥99.999%, the flow rate is reduced to 0.5 mL / min, the preheating time is shortened to 5 minutes, and inert gas protection is used to protect the sampling pipeline; the batch adaptation is as follows: based on the random forest algorithm (100 decision trees), historical data from 500+ batches are analyzed to construct a "sample characteristics-operating parameters" mapping model, with a sampling parameter matching accuracy of ≥98% for the same type of batch. The path optimization design is based on ANSYS Fluent CFD fluid simulation (turbulence model k - ε, mesh count 1 million+), optimizing the sampling pipeline path (total length ≤ 1.5m, bends ≤ 3), and the total dead volume < 0.1mL; a three-step cleaning method is adopted (anhydrous ethanol solvent rinsing - nitrogen purging - vacuum drying), which is automatically executed after each sampling, and the cleaning time can be adaptively adjusted according to the impurity content (3 - 8 minutes).

[0046] The implementation principle of this embodiment is as follows: the non-contact magnetically driven sampler avoids the risk of leakage at dynamic sealing points and improves sampling safety through magnetic coupling non-contact transmission. The condition-adaptive sampling system can adjust sampling parameters, paths, and cleaning procedures in real time according to the characteristics of the sample, solving the problem of contamination or detection errors caused by batch differences and improving the accuracy and reliability of sampling. Example

[0047] This embodiment further improves the system's safety features and human-computer interaction functions based on the above embodiments. Regarding safety, the system has physical and chemical safety protection measures. In terms of explosion-proof and mechanical safety design, the electrical system as a whole reaches ExdIICT4 Ga level, the junction box is explosion-proof (model BXJ-51), the cable is cross-linked polyethylene insulated PVC sheathed copper core cable (model YJV-0.6 / 1kV, cross-section 2.5mm²), the wiring is run through galvanized steel conduit (nominal diameter 20mm), and the explosion-proof gap is ≤0.1mm; it is equipped with an ion fan of model SF-200 (wind speed 1-3m / s, ion balance ±5V), and the equipment shell and pipelines are coated with an antistatic coating (surface resistance 10 ohms). 6 - 10 8 The system equipotential bonding resistance is <0.1Ω, and the grounding resistance is <4Ω (using copper grounding electrodes, depth 2.5m); the safety valve is a spring type (model A42Y-16C, opening pressure 0.6MPa, reseating pressure 0.55MPa), the rupture disc is made of graphite (model LPB-50, burst pressure 0.7MPa, burst accuracy ±5%), and equipped with a PT-200 pressure transmitter (measuring range 0-1MPa, accuracy ±0.2%FS) for real-time monitoring; the equipment guardrail is made of 304 stainless steel (height 1.2m, spacing 10cm), equipped with a GR-100 infrared grating (protection height 1.5m, response time <10ms, detection distance 5m), and the maintenance door uses electromagnetic interlocking (power failure unlocking, unlocking force <50N). In terms of the chemical emergency response system, when a leak occurs, it automatically sprays inert gas (argon, flow rate 10L / min) + negative pressure suction (negative pressure -0.08MPa), with a suction efficiency >99%; it uses high-pressure atomization injection of ethanol-water solution (volume ratio 1:1) (pressure 2MPa) for chemical neutralization, with a neutralization efficiency >99.9%; by-products are treated by activated carbon adsorption (adsorption capacity ≥50mg / g) + catalytic oxidation (oxidation temperature 300℃) to ensure that the exhaust gas emissions meet the standards.

[0048] In terms of human-computer interaction, the system possesses intent-state collaborative human-computer safe interaction technology. The state perception module uses a HW-800 smart bracelet (heart rate measurement range 30-200 bpm, accuracy ±1 bpm; blood oxygen saturation 90%-100%, accuracy ±1%) and EG-300 smart glasses (blood pressure measurement range 60 / 40-200 / 120 mmHg, accuracy ±2 mmHg) for physiological monitoring, transmitting data via Bluetooth 5.0. A high-definition industrial camera (MV-1300, resolution 1280×720, frame rate 30fps) captures operational actions, and a CNN algorithm (12 layers, recognition accuracy ≥99%) is used to analyze valve rotation accuracy (±1°), operation speed (±0.1 m / s), and hand tremor detection accuracy ≤0.1 mm / s. An ASR- The 500 speech recognition module (recognizes Chinese and English languages, with an accuracy of >98% and a response time of <300ms) extracts speech features through Mel-frequency cepstral coefficients (MFCC) to determine speech rate (normal 60-120 words / minute) and intonation changes (fluctuations ±50Hz determine state changes). The adaptive interaction strategy optimizes based on the operator's state. When the operator is fatigued (heart rate < 60 bpm, action response delay > 1 second), operation steps are merged, voice reminders are repeated, a status confirmation step is added, the system background calls a simplified process script, and the TTS voice module (model TTS-300) broadcasts reminders, reducing the operation error rate by 50% and shortening the operation time by 15%. When the operator is stressed (heart rate > 100 bpm, rapid tone (> 150 words / min)), AR animation magnifies the guidance, annotates operation details, plays soothing music, the AR engine (Unity 3D) renders the animation, and the audio module plays 40-60 BPM music, reducing operation errors by 40% and shortening the time for the heart rate to return to the normal range by 40%. When the operator is proficient (smooth movements (completion time < 80% of standard time), stable voice), basic prompts are hidden, one-click shortcut operations are enabled, a custom parameter interface is provided, the UI dynamically loads configuration files, and shortcut key customization is supported, reducing the operation time by 25% and improving interaction satisfaction to 95%. Meanwhile, based on the RNN network (64 neurons in the hidden layer, 10,000+ operation records in the training sample), the historical operation sequence is analyzed to predict the next intention (accuracy ≥ 95%), and the relevant parameter interface (such as the argon flow rate setting interface) is loaded 0.5 seconds in advance. Offline prediction mode is supported (caching the most recent 100 operation data).

[0049] The implementation principle of this embodiment is as follows: Physical and chemical safety protection measures ensure the safety of the system from multiple aspects. Explosion-proof design prevents explosion accidents, electrostatic control avoids dangers caused by static electricity, pressure safety devices ensure that the system pressure is within a safe range, and mechanical protection protects the safety of operators. The chemical emergency response system can handle leaks in a timely manner, reducing the hazards caused by leaks. The intent-state collaborative human-machine safety interaction technology adaptively interacts based on the operator's physiological and behavioral states, reducing the risk of human error and improving the accuracy and efficiency of operation. Example

[0050] This embodiment, based on the above embodiments, improves the system's modular architecture, traceability management, and parameter collaborative control. The system adopts a modular system architecture design, including a sampling execution module (dynamic five-layer barrier + non-contact magnetic drive sampler + adaptive cleaning unit, realizing the core function of sealed sampling), an intelligent control module (digital twin platform + AI decision-making system + sensor data acquisition unit, responsible for state prediction and parameter control), a safety assurance module (dual-modal leak detection + explosion-proof protection + chemical emergency treatment, realizing risk prevention and emergency response), a human-computer interaction module (AR augmented reality interface + voice interaction + wearable status monitoring, realizing intelligent collaborative operation), and a traceability management module (blockchain node + data on-chain unit + traceability query interface, realizing full lifecycle traceability). The modules are connected via standardized interfaces. The hardware interfaces use standard DN25 flange connections and tongue and groove structures for sealing surfaces, with interchangeability ≥95%. The software interfaces use the OPC UA industrial communication protocol, supporting integration with MES and ERP systems. The data interfaces use the JSON standardized format, and blockchain data is transmitted using hash encryption. The control interfaces use the Modbus-RTU protocol, with control signal response time <100ms.

[0051] The system adopts a full lifecycle traceability technology that integrates blockchain and digital twins. In terms of blockchain architecture, the consortium blockchain adopts Hyperledger Fabric 2.4, with node deployment using Docker containerization technology (each node configured with 4 cores and 8GB of memory), and the PBFT consensus mechanism (4 nodes, consensus latency <500ms). Data transmission uses AES-256 encryption algorithm, block hashing uses SHA-256 algorithm, and digital signature uses ECDSA algorithm (key length 256 bits). The key is automatically rotated every 7 days. Production / transportation / storage / sampling / testing data is stored in "batch-timestamp" partitions, with a single batch data size ≤50MB, block generation time <10 seconds, data retention period ≥3 years, and support for offline data retransmission (1000 data entries cached after network outage). The node deployment scheme is as follows: manufacturer nodes are deployed on production workshop servers (industrial-grade servers, model IBM System x3650), transportation company nodes are deployed in the logistics monitoring center (cloud servers, Alibaba Cloud ECS), and user unit nodes are deployed in the sampling site control cabinet (embedded servers, model Advantech ARK-). The testing agency nodes are deployed on a laboratory server (Dell PowerEdge R740), and the nodes communicate with each other via a VPN leased line (bandwidth ≥100Mbps). Full lifecycle data traceability covers the production stage (manufacturers upload production batches, synthesis temperatures, initial purity, etc., generating unique batch hash codes), the transportation stage (GPS + temperature sensors upload transportation data in real time (temperature 25±5℃, vibration ≤2g), generating a data block every 30 minutes), the storage stage (the warehouse system uploads storage temperature (-20℃), storage time, and synchronizes environmental parameters with the digital twin model), the sampling stage (automatically uploads sampling parameters, test results, equipment health data, and associates operator information), and the testing stage (the testing agency uploads analysis results, cross-validates with sampling data, forming a complete quality chain).

[0052] The system employs a multivariable coupled control algorithm to achieve dynamic and coordinated optimization of various system parameters. For example, in barrier-vacuum linkage, when the magnetic fluid sealing differential pressure decreases, the vacuum level of the negative pressure isolation chamber is automatically increased, with a response time of <1 second; in temperature-pressure coordination, when the DEZ temperature increases by 10℃, the sampling flow rate is automatically reduced by 20% and the condensation effect is enhanced to maintain stable system pressure; in sampling-cleaning synchronization, the cleaning solvent flow rate and time are adjusted in real time based on the impurity content detected by NIR, achieving coordinated optimization of sampling and cleaning; and in detection-emergency linkage, when the leak detection confidence level is ≥0.6, the inert gas spray system is pre-activated to shorten the emergency response time.

[0053] The implementation principle of this embodiment is as follows: Modular system architecture design ensures good independence and scalability of each part of the system, facilitating system maintenance and upgrades. Standardized interfaces guarantee compatibility and interoperability between modules. Blockchain-digital twin integrated full lifecycle traceability technology enables reliable traceability of samples from production to testing, solving the problems of quality responsibility identification and authenticity control. Multivariable coupled control algorithms can adjust various parameters in real time according to different system operating conditions, achieving dynamic collaborative optimization of system parameters and improving the overall performance and security of the system.

[0054] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A smart, sealed sampling safety system for diethylzinc, characterized in that, It includes a magnetic fluid sealing assembly, an intelligent deformation sealing assembly, a gas replacement protective layer, a double-wall structure negative pressure isolation chamber, and a condensation trapping assembly arranged in sequence. Each assembly forms a dynamic complementary enhancement loop through a state sensing unit containing a quantum cascade laser detection module and an electrochemical array detection module, and a control unit integrating a fuzzy control algorithm. When the performance parameter of any component deviates from the threshold, a graded compensation strategy is automatically triggered.

2. The system according to claim 1, characterized in that, The magnetic fluid sealing assembly includes a neodymium iron boron annular permanent magnet, a titanium alloy magnetic fluid reservoir, and a magnetic field gradient controller of model MGC-200. The reservoir is filled with Fe3O4 nanoparticle magnetic fluid containing 5% fluorocarbon anti-corrosion additives. The differential pressure sensor is a model with an accuracy of 0.001MPa and the working differential pressure is maintained within the range of -0.1 to 0.5MPa through a closed-loop control system.

3. The system according to claim 1, characterized in that, The intelligent deformation sealing assembly is made of polyimide-based shape memory polymer and conductive carbon nanotube composite material. It has a built-in fiber optic deformation sensor and is connected to a nickel-chromium heating wire and a PID temperature control system. The temperature control system has control parameters with a proportional coefficient of 0.8 and an integral time of 10 seconds, which can make the diameter of the sealing ring shrink within a range of 0-5mm and a response time of ≤30 seconds.

4. The system according to claim 1, characterized in that, The gas replacement protective layer includes a KV-10 electromagnetic flow regulating valve and an OXY-910 oxygen content analyzer. The flow regulating valve can adaptively adjust the flow rate of ultrapure argon gas in the range of 1-5 L / min. The micro positive pressure maintenance system controls the pressure in the cavity at +50~100 Pa and the oxygen content at less than 1 ppm through a pressure sensor.

5. The system according to claim 1, characterized in that, The double-walled negative pressure isolation chamber is made of 316L stainless steel with an inner wall thickness of 5mm and an outer wall thickness of 3mm. The molecular pump and mechanical pump linkage system, together with the vacuum gauge, monitor in real time, which can adjust the vacuum degree of the inner chamber in the range of 0.5-1Pa and automatically alarm when the pressure rise rate exceeds 0.1Pa / min.

6. The system according to claim 1, characterized in that, The condensation and collection assembly is equipped with a liquid nitrogen cooling module and an electric heating regeneration module. The cooling module achieves controllable cooling from -196 to -200°C through a TC-300 temperature controller. The regeneration module uses a 120°C desorption temperature combined with anhydrous ethanol solvent for cleaning. The regeneration cycle is ≤15 minutes and the collection efficiency is ≥99.995%.

7. The system according to claim 1, characterized in that, The control unit integrates a multiphysics coupling model that includes four sub-models: fluid, heat transfer, stress, and chemical reaction. The fluid model adopts the k-ω SST turbulence model of ANSYS Fluent. The remaining life prediction of the sealing component is corrected by real-time data interaction of temperature field, concentration field, and mechanical field, with a prediction accuracy of ±5%.

8. The system according to claim 7, characterized in that, The material health index is calculated using a weighted average algorithm, which integrates three parameters: corrosion rate (Hastelloy ≤ 0.01 mm / year), deformation rate (≤ 5%), and stress value (≤ 300 MPa), with weights of 0.4, 0.3, and 0.3, respectively. The remaining life prediction is achieved based on LSTM network training.

9. The system according to claim 1, characterized in that, The quantum cascade laser detection module uses a QCL laser with a center wavelength of 10.6 μm in conjunction with an MCT detector. The electrochemical array detection module detects Zn²⁺ and ethanol concentrations through a three-electrode system. The DS evidence theory fusion algorithm assigns a weight of 0.6 to QCL detection and a weight of 0.4 to electrochemical detection. An emergency response is triggered when the leakage confidence threshold is ≥0.

8.

10. The system according to claim 1, characterized in that, The fuzzy control algorithm uses a triangular membership function. The input parameters include the magnetohydrodynamic pressure differential attenuation rate (judgment threshold 5%-15%), the deformation rate decrease value (judgment threshold 3%-8%), and the oxygen content exceeding the standard value (judgment threshold 1.0-2.0ppm). The output compensation strategy includes three types of control actions: increasing the magnetic field strength by 0.1T, reducing the vacuum degree to 0.3Pa, and reducing the condensation temperature to -200℃. The response time is ≤5 seconds.

Images