Efficient integrated nitrogen component analysis method and system

Abstract

The invention discloses an efficient integrated nitrogen component analysis method, and belongs to the technical field of chemical analysis, the efficient integrated nitrogen component analysis method comprises the following steps: introducing a nitrogen sample to be detected into an automatic flow control distribution device through an outlet of a nitrogen sampling steel cylinder; the automatic flow control distribution device comprises a programmable industrial control chip, a flow control module and a plurality of electromagnetic valves; the programmable industrial control chip controls the flow control module and the opening and closing of the plurality of electromagnetic valves according to a preset mode, and distributes a nitrogen sample to be tested to the trace oxygen analyzer, the gas chromatograph and / or the chilled mirror dew point instrument according to test requirements; a trace oxygen analyzer, a gas chromatograph and a chilled mirror dew-point meter are used for analyzing and detecting oxygen, other gas components except the oxygen and moisture components in the nitrogen sample respectively; and obtaining a component analysis result of the to-be-detected nitrogen sample. Multiple parameters are analyzed at the same time through one-time sample injection, and the sample and test cost is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of chemical analysis, and particularly to an efficient integrated nitrogen component analysis method and system. Background Art

[0002] In many industrial fields such as chemical engineering, electronics, and food, there are strict requirements for the purity and impurity content of nitrogen. For example, in the semiconductor manufacturing process, ultra-high purity nitrogen is required to protect the chip from impurity contamination, and any tiny impurity may affect the chip performance. Through nitrogen component analysis, the quality of nitrogen can be precisely controlled to ensure the smooth progress of industrial production and product quality. In food processing, such as puffed food, milk powder, beverages, etc. Nitrogen is mainly used as a processing aid to isolate oxygen, avoid the oxidation of ingredients in food, inhibit the growth of certain microorganisms, and improve the strength of food packaging, making it not easily damaged by extrusion during production and transportation. Currently, according to national standard requirements, multiple items need to be tested for nitrogen analysis, different equipment is used, and multiple injections are required, which requires a large amount of samples, and it is easy to contaminate the samples collected with air bags and difficult to transport. Summary of the Invention

[0003] An object of the present invention is to solve at least the above problems and / or defects and provide at least the advantages described hereinafter.

[0004] Another object of the present invention is to provide an efficient integrated nitrogen component analysis method, which aims at the problems in the prior art such as multiple injections required, a large amount of samples needed, and high testing costs, and realizes the simultaneous analysis of multiple parameters with one injection, saving samples and testing costs.

[0005] According to these objects and other advantages of the present invention, an efficient integrated nitrogen component analysis method is provided, which includes the following steps: 1) Connect the nitrogen sample to be tested to an automatic flow control and distribution device through the outlet of a nitrogen sampling steel cylinder. The automatic flow control and distribution device includes a programmable industrial control chip, a flow control module, and multiple solenoid valves; 2) The programmable industrial control chip controls the opening and closing of the flow control module and multiple solenoid valves according to a preset mode, and distributes the nitrogen sample to be tested to a trace oxygen analyzer, a gas chromatograph, and / or a cold mirror dew point meter according to the test requirements; 3) The trace oxygen analyzer, the gas chromatograph, and the cold mirror dew point meter respectively analyze and detect the oxygen, other gas components except oxygen, and moisture components in the nitrogen sample; 4) Collect and integrate the detection data of the gas chromatograph, the trace oxygen analyzer, and / or the cold mirror dew point meter to obtain the component analysis result of the nitrogen sample to be tested.

[0006] Preferably, the flow control module includes an electronic flow controller I and an electronic flow controller II, and a plurality of solenoid valves, namely solenoid valve I, solenoid valve II, solenoid valve III, solenoid valve IV, and solenoid valve V. Among them, the intake end of the electronic flow controller I is connected to the nitrogen sampling cylinder, and the outlet end is connected to the intake ports of solenoid valve I, solenoid valve II, and solenoid valve IV. The outlet of solenoid valve I is connected to the input end of the cold mirror dew point meter to form a first gas flow pipeline. The outlet of solenoid valve II is connected to the input end of the trace oxygen analyzer to form a second gas flow pipeline. The outlet of solenoid valve IV is connected to the input end of the gas chromatograph to form a third gas flow pipeline. The output end of the cold mirror dew point meter is connected to the intake port of solenoid valve III and the input end of the trace oxygen analyzer. The output end of the trace oxygen analyzer is connected to the intake port of solenoid valve V and the input end of the gas chromatograph. The outlets of solenoid valve III and solenoid valve V are both connected to the intake end of the electronic flow controller II.

[0007] Preferably, step 2) specifically includes: When testing all components in the nitrogen sample to be measured, the programmable industrial control chip controls the flow rate of the electronic flow controller I to be 200 - 300 mL / min, the flow rate of the electronic flow controller II to be 200 - 300 mL / min, solenoid valve I and solenoid valve V are opened, and other solenoid valves are closed. The nitrogen sample to be measured is synchronously distributed to the gas chromatograph, the trace oxygen analyzer, and the cold mirror dew point meter. When testing the moisture in the nitrogen sample to be measured alone, the programmable industrial control chip controls the flow rate of the electronic flow controller I to be 300 - 350 mL / min, the flow rate of the electronic flow controller II to be 300 - 350 mL / min, solenoid valve I and solenoid valve III are opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the cold mirror dew point meter. When testing the oxygen in the nitrogen sample to be measured alone, the programmable industrial control chip controls the flow rate of the electronic flow controller I to be 10 - 100 mL / min, the flow rate of the electronic flow controller II to be 10 - 100 mL / min, solenoid valve II and solenoid valve V are opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the trace oxygen analyzer. When testing carbon monoxide and / or carbon dioxide gas in the nitrogen sample to be measured, the programmable industrial control chip controls the flow rate of the electronic flow controller I to be 10 - 50 mL / min, the flow rate of the electronic flow controller II to be 0 mL / min, solenoid valve IV is opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the gas chromatograph.

[0008] Preferably, after step 4), there is also a solenoid valve life prediction step, including: S1. Install a current sensor, a non-contact displacement sensor, and a temperature and humidity sensor on each solenoid valve body to collect electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of each solenoid valve; S2. Preprocess the collected data; S3. Build a prediction model based on LSTM, and use the prediction model to predict and give early warnings for the service life of each solenoid valve: Among them, WI is the dynamic wear index of the solenoid valve, N is the cumulative number of operations, N max is the rated life number of times, I peak is the peak value of the coil current within the current operation cycle of the solenoid valve, I initial is the peak value of the coil current in the initial state of the solenoid valve, e is a constant, equal to 2.71828, T is the real-time working environment temperature of the solenoid valve; RUL is the dynamic remaining life, WI threshold is the preset critical threshold of the solenoid valve wear index, WI current is the dynamic wear index of the solenoid valve within the current operation cycle, d( WI ) / d N is the derivative of the solenoid valve dynamic wear index with respect to the cumulative number of operations N of, is the environmental correction factor, = 1 - 0.05×( RH - 50%), RH is the environmental humidity.

[0009] Preferably, among them, in S1, the electrical parameter data includes: coil current, drive voltage, and power consumption; the mechanical parameter data includes: response time, spool displacement, and sealing pressure; the environmental parameter data includes: working temperature, humidity, and vibration intensity; the operation parameter data includes: cumulative number of operations, vibration frequency, and maintenance records.

[0010] Preferably, in S2, the electrical parameter data eliminates electromagnetic interference noise through Kalman filtering, the mechanical parameter data is smoothed and jittered by the variable-weight moving average filtering method, and the environmental parameter data is corrected through the temperature compensation correction formula, specifically including: performing adaptive Kalman filtering on the coil current data, with the process noise Q = 1e-5 and the observation noise R = 0.01; the displacement signal is smoothed and jittered by the variable-weight moving average filtering. In the starting stage of the solenoid valve core, when the displacement < 0.2 mm, a 5-point window is used. In the high-speed movement stage of the solenoid valve core, when the displacement ≥ 0.2 mm, it switches to a 3-point window; the environmental parameter data is corrected through the temperature compensation correction formula: establishing a temperature-humidity coupling correction model: P (Correction) = P (Original) × [1 + 0.003( T − 25)] × [1 − 0.02( RH − 50%)] to correct the temperature and humidity.

[0011] Preferably, the warning specifically includes: When the dynamic remaining life RUL is greater than 1000 times, the warning level is normal, the warning color identifier is green, and a health report is generated; When the dynamic remaining life RUL is greater than 500 and less than 1000 times, the warning level is attention, the warning color identifier is yellow, and the spare part procurement process is activated; When the dynamic remaining life RUL is less than 500 times, the warning level is urgent, the warning color identifier is red, and the redundant valve switching and maintenance work order are triggered.

[0012] Preferably, between step S1 and step S2, there is also a step: using the multi-copy consistent hashing storage algorithm to upload the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of each solenoid valve collected to the cloud platform for storage. The electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves are split into multiple data blocks, and the split multiple data blocks are stored on different data nodes in the cluster, specifically including: Predefining the correlation and the number of redundant copies of the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves according to the configuration file; In the Hadoop cluster of the cloud platform, calculating the hash values of different data nodes through the MapReduce program; Through the hash function shown in the following formula, mapping data of any length to a fixed 128-bit hash value; Hash (M) = Md5 (x) In the formula,Md5(x) represents a hash function x represents the electrical parameters, mechanical parameters, environmental parameters, and operating parameter data of different solenoid valves for which the hash value is to be calculated; After generating a 128-bit hash value, 32 bits are extracted therefrom as the hash value, and then they are mapped to the hash ring one by one. By comparing the hash values of the electrical parameters, mechanical parameters, environmental parameters, and operating parameter data of different solenoid valves with the hash values of different data nodes in the Hadoop cluster data, the storage locations of the electrical parameters, mechanical parameters, environmental parameters, and operating parameter data of different solenoid valves in the cloud platform are determined.

[0013] Preferably, after step S3, it further includes: establishing a concept drift detection mechanism: when the prediction confidence level is < 90% for 100 consecutive times, the prediction model is optimized.

[0014] Preferably, a quick-connect fitting is provided at the outlet of the nitrogen sampling cylinder to achieve quick-connect substitution for threaded connection.

[0015] The present invention has at least the following beneficial effects: 1. The high-efficiency integrated nitrogen component analysis method of the present invention addresses the problems in the prior art of requiring multiple injections, a large sample volume, and high testing costs, and realizes the simultaneous analysis of multiple parameters in one injection, saving samples and testing costs.

[0016] 2. The preparation method of the high-efficiency integrated nitrogen component analysis method of the present invention meets different testing requirements by setting multiple detection modes.

[0017] 3. The preparation method of the high-efficiency integrated nitrogen component analysis method of the present invention predicts the service lives of multiple solenoid valves by establishing a solenoid valve life prediction model, avoiding unplanned stops during detection and affecting the detection progress.

[0018] 4. The preparation method of the high-efficiency integrated nitrogen component analysis method of the present invention realizes a prediction accuracy within ±8% and an early warning lead of more than 500 operations through spatio-temporal correlation analysis and prediction of current-displacement-environmental parameters.

[0019] Other advantages, objectives, and features of the present invention will be partially reflected by the following description and partially understood by those skilled in the art through the research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a flowchart of the high-efficiency integrated nitrogen component analysis method according to an embodiment of the present invention; Figure 2 is a schematic structural diagram of the device used in the high-efficiency integrated nitrogen component analysis method according to an embodiment of the present invention; Figure 3 It is a schematic structural diagram of an efficiently integrated nitrogen component analysis device according to another embodiment of the present invention. Specific embodiments

[0022] The present invention will be further described in detail below so that those skilled in the art can implement it with reference to the text of the specification.

[0023] It should be understood that terms such as "having", "comprising", and "including" used herein do not preclude the presence or addition of one or more other elements or combinations thereof.

[0024] It should be noted that the experimental methods described in the following embodiments are all conventional methods unless otherwise specified, and the reagents and materials can be obtained from commercial channels unless otherwise specified.

[0025] As Figure 1 shown, the present invention provides an efficiently integrated nitrogen component analysis method, including the following steps: 1) Connect the nitrogen sample to be measured to an automatic flow control and distribution device through the outlet of a nitrogen sampling steel cylinder. The automatic flow control and distribution device includes a programmable industrial control chip, a flow control module, and a plurality of solenoid valves; 2) The programmable industrial control chip controls the opening and closing of the flow control module and a plurality of solenoid valves according to a preset mode, and distributes the nitrogen sample to be measured to a trace oxygen analyzer, a gas chromatograph, and / or a cold mirror dew point meter according to the test requirements; 3) The trace oxygen analyzer, the gas chromatograph, and the cold mirror dew point meter respectively analyze and detect the oxygen, other gas components except oxygen, and moisture components in the nitrogen sample; 4) Collect and integrate the detection data of the gas chromatograph, the trace oxygen analyzer, and / or the cold mirror dew point meter to obtain the component analysis result of the nitrogen sample to be measured.

[0026] In the above technical solution, a nitrogen sampling cylinder with a pressure reducing valve (compliant with ISO 10156 standard) is adopted, and the outlet is connected to an automatic flow control distribution device through a 316L stainless steel pipeline. A dust filter (0.1μm), a pressure sensor (range 0 - 1 MPa, accuracy ±0.1% FS), and a temperature sensor (-40 - 120°C, accuracy ±0.5°C) are configured on the pipeline. Automatic flow control distribution device: Core controller: Siemens S7-1200 PLC (integrated with PID control function) Flow control module: High-precision electronic pressure controller (EPC, Brooks Instrument 5850i series EPC, which features high precision and fast response. This series of EPC can provide multiple gas input interfaces, adapt to the nitrogen measurement requirements, and the pressure control range can cover the common industrial gas pressure range, such as 0 - 100 psig (about 0 - 690 kPa), with an accuracy of up to ±0.1% of the set value). The solenoid valve is a two-position three-way direct-acting solenoid valve, such as the SMC VQ110U series. This type of solenoid valve has a simple structure and fast response speed, and is suitable for small-flow gas control. For common gas media such as nitrogen, it can reliably achieve the on-off and flow direction switching functions. Cold mirror dew point meter: Model: MBW M3000 (range -100 - +20°C Td, accuracy ±0.2°C Td) Configuration: Automatic defrost function (activated when the dew point < -60°C). Trace oxygen analyzer: Model: Servomex 4100 (range 0 - 100 ppm, accuracy ±0.5 ppm) Configuration: Automatic zero calibration (once every 24 hours) Gas chromatograph: Model: Agilent 7890B, Detector: TCD (detecting H 2 、CO、CO 2 ), + FID (detecting hydrocarbons), Chromatographic column: MS-5A molecular sieve column (separating O 2 / N 2 ), + PLOT U column (separating CO 2 ). Each instrument is connected to the PLC through the Modbus RTU / RS485 protocol. Sampling frequency: 10 Hz (real-time data), Storage interval: 1 minute. Detection data algorithm: Integrated algorithm: Oxygen content correction (temperature / pressure compensation formula), Moisture content conversion (look-up table method + linear interpolation), Gas chromatographic peak area normalization calculation. Data storage: SQLite database (supporting 3-year historical data storage).

[0027] Specific test example: Test object: Semiconductor-grade nitrogen (purity ≥ 99.999%). Select the mode. By controlling the opening and closing of multiple valves, all components in nitrogen are detected. The system automatically adjusts the flow rate to 200 sccm, and the test is completed within 3 minutes. Oxygen content detection (display: 5.2 ppm), moisture detection (display: -75°C Td), gas chromatography analysis (trace CO 2 : 0.8 ppm) is detected, and an analysis report is generated and uploaded to the LIMS system Through modular design, this implementation mode can adjust the instrument configuration and detection parameters according to the needs of different industries (such as electronics, chemical industry, food), realizing high-precision and high-efficiency nitrogen component analysis.

[0028] In another technical solution, the flow control module includes an electronic flow controller I and an electronic flow controller II, and multiple solenoid valves, namely solenoid valve I, solenoid valve II, solenoid valve III, solenoid valve IV, and solenoid valve V; Among them, the intake end of the electronic flow controller I is connected to the nitrogen sampling cylinder, and the outlet end is connected to the intake ports of solenoid valve I, solenoid valve II, and solenoid valve IV. The outlet of solenoid valve I is connected to the input end of the cold mirror dew point meter to form a first gas flow pipeline. The outlet of solenoid valve II is connected to the input end of the trace oxygen analyzer to form a second gas flow pipeline. The outlet of solenoid valve IV is connected to the input end of the gas chromatograph to form a third gas flow pipeline. The output end of the cold mirror dew point meter is connected to the intake port of solenoid valve III and the input end of the trace oxygen analyzer. The output end of the trace oxygen analyzer is connected to the intake port of solenoid valve V and the input end of the gas chromatograph. The outlets of solenoid valve III and solenoid valve V are both connected to the intake end of the electronic flow controller II.

[0029] In the above technical solution, as Figure 2As shown in the figure, nitrogen component detection in different modes is achieved through different gas flow pipelines. When performing the full-component analysis mode: The PLC controls the electronic flow controller to control the output of nitrogen with a suitable flow rate. Solenoid valve 1 and solenoid valve 5 are opened, and other solenoid valves are closed; the nitrogen sample to be measured is synchronously distributed to the gas chromatograph, the micro-oxygen analyzer, and the cold mirror dew point meter. After the detection is completed, through solenoid valve 5, the gases detected by each instrument are processed by the electronic flow controller 2 (such as recovery or discharge). When separately testing the moisture in the nitrogen sample to be measured, the PLC controls the electronic flow controller to control the output of nitrogen with a suitable flow rate. Solenoid valve 1 and solenoid valve 3 are opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the cold mirror dew point meter. After the detection is completed, through solenoid valve 3, the gas detected by the cold mirror dew point meter is processed by the electronic flow controller 2 (such as recovery or discharge). When testing carbon monoxide and / or carbon dioxide gases in the nitrogen sample to be measured, the PLC controls the electronic flow controller to control the output of nitrogen with a suitable flow rate. At the same time, solenoid valve 4 is opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the gas chromatograph.

[0030] In another technical solution, step 2) specifically includes: When testing all components in the nitrogen sample to be measured, the programmable industrial control chip controls the flow rate of the electronic flow controller 1 to be 200 - 300 mL / min, and the flow rate of the electronic flow controller 2 to be 200 - 300 mL / min. Solenoid valve 1 and solenoid valve 5 are opened, and other solenoid valves are closed; the nitrogen sample to be measured is synchronously distributed to the gas chromatograph, the micro-oxygen analyzer, and the cold mirror dew point meter; When separately testing the moisture in the nitrogen sample to be measured, the programmable industrial control chip controls the flow rate of the electronic flow controller 1 to be 300 - 350 mL / min, and the flow rate of the electronic flow controller 2 to be 300 - 350 mL / min. Solenoid valve 1 and solenoid valve 3 are opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the cold mirror dew point meter; When separately testing the oxygen in the nitrogen sample to be measured, the programmable industrial control chip controls the flow rate of the electronic flow controller 1 to be 10 - 100 mL / min, and the flow rate of the electronic flow controller 2 to be 10 - 100 mL / min. Solenoid valve 2 and solenoid valve 5 are opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the micro-oxygen analyzer; When testing carbon monoxide and / or carbon dioxide gases in the nitrogen sample to be measured, the programmable industrial control chip controls the flow rate of the electronic flow controller 1 to be 10 - 50 mL / min, and the flow rate of the electronic flow controller 2 to be 0 mL / min. Solenoid valve 4 is opened, and other solenoid valves are closed. The nitrogen sample to be measured is only distributed to the gas chromatograph.

[0031] In the above technical solution, specifically, the test mode is as follows: Mode 1: When all parameters in nitrogen are to be tested simultaneously, the flow rate of Electronic Flow Controller 1 is set to 310 mL / min, the flow rate of Electronic Flow Controller 2 is set to 300 mL / min, Solenoid Valve 1 and Solenoid Valve 5 are opened and the others are closed.

[0032] When all parameters in nitrogen need to be tested, after setting the flow rate parameters of Electronic Flow Controller 1 and Electronic Flow Controller 2 respectively, Solenoid Valve 1 and Solenoid Valve 5 are opened and the others are closed. At this time, nitrogen flows out from the nitrogen sampling cylinder, is distributed by Electronic Flow Controller 1. Part of the gas enters the cold mirror dew point meter through the first gas flow pipeline to detect the moisture content; the other part enters the subsequent analytical instrument through Solenoid Valve 5, and finally the simultaneous detection of multiple parameters is completed. The detection data is displayed on the display screens of each instrument in real time, or the data can be transmitted to the corresponding file storage location through the data transmission interface; Mode 2: When testing moisture alone, the flow rate of Electronic Flow Controller 1 is 300 mL / min, the flow rate of Electronic Flow Controller 2 is 300 mL / min, Solenoid Valve 1 and Solenoid Valve 3 are opened and the others are closed.

[0033] When only the moisture content in nitrogen needs to be detected, after setting the flow rate parameters of Electronic Flow Controller 1 and Electronic Flow Controller 2 respectively, Solenoid Valve 1 and Solenoid Valve 3 are opened and the others are closed. Nitrogen enters the first gas flow pipeline through Electronic Flow Controller 1, and after the moisture detection is completed in the cold mirror dew point meter, the gas flows into Electronic Flow Controller 2 and is discharged through Solenoid Valve 3 to avoid interference of gas residue on subsequent detections. The detection result is directly displayed on the cold mirror dew point meter.

[0034] Mode 3: When testing oxygen alone, the flow rate of Electronic Flow Controller 1 is 100 mL / min, the flow rate of Electronic Flow Controller 2 is 100 mL / min, Solenoid Valve 2 and Solenoid Valve 5 are opened and the others are closed.

[0035] When only the oxygen content in nitrogen needs to be detected, after setting the flow rate parameters of Electronic Flow Controller 1 and Electronic Flow Controller 2 respectively, Solenoid Valve 2 and Solenoid Valve 5 are opened and the others are closed. Nitrogen enters the second gas flow pipeline through Electronic Flow Controller 1, and the oxygen content is detected by the trace oxygen analyzer. The detected gas flows into Electronic Flow Controller 2 and is discharged through Solenoid Valve 5. The operator can read the oxygen content data from the display screen of the trace oxygen analyzer.

[0036] Mode 4: When testing CO / CO 2 , the flow rate of Electronic Flow Controller 1 is 20 mL / min, the flow rate of Electronic Flow Controller 2 is 0 mL / min, Solenoid Valve 4 is opened and the others are closed.

[0037] When only the contents of carbon monoxide and carbon dioxide in nitrogen need to be detected, after setting the flow parameters of the first electronic flow controller and the second electronic flow controller respectively, the fourth solenoid valve is opened and the others are closed. Nitrogen enters the third gas flow pipeline through the first electronic flow controller, and the carbon monoxide and carbon dioxide are detected and analyzed by the gas chromatograph. The detected gas is discharged from the gas outlet of the gas chromatograph, and the detection results are processed and displayed through the supporting software of the gas chromatograph.

[0038] After completing all tests, close the nitrogen sampling cylinder and stop the gas supply. Close all solenoid valves to prevent gas backflow and leakage. Clean and calibrate the analytical instrument to prepare for the next test; regularly check whether the connecting pipelines are aging or damaged, and replace the damaged components in time to ensure that the equipment is always in good operating condition.

[0039] In another technical solution, after step 4), there is also a solenoid valve life prediction step, including: S1. Install a current sensor, configure a non-contact displacement sensor and a temperature and humidity sensor on each solenoid valve body to collect the electrical parameters, mechanical parameters, environmental parameters and operation parameter data of each solenoid valve; S2. Preprocess the collected data; S3. Build a prediction model based on LSTM, and predict and give early warnings about the life of each solenoid valve through the prediction model: Among them, WI is the dynamic wear index of the solenoid valve, N is the cumulative number of operations, N max is the rated life number of times, I peak is the peak value of the coil current within the current operation cycle of the solenoid valve, I initial is the peak value of the coil current in the initial state of the solenoid valve, e is a constant, equal to 2.71828, T is the real-time working environment temperature of the solenoid valve; RUL is the dynamic remaining life, WI threshold is the preset critical threshold of the solenoid valve wear index, WI current is the dynamic wear index of the solenoid valve within the current operation cycle, d( WI ) / d N is the derivative of the solenoid valve dynamic wear index with respect to the cumulative number of operations N of, is the environmental correction factor, = 1 - 0.05×( RH - 50%), RH where

[0040] is the ambient humidity. I peak is the peak coil current (unit: mA or A) within the current action cycle of the solenoid valve. Physical meaning: Reflects the coil aging state (increase in resistance leads to an increase in current demand) and changes in the spool movement resistance. Measurement method: Captures the instantaneous current maximum value of each action through a high-precision current sensor. I initial is the peak coil current (benchmark value under the same working conditions) in the initial state of the solenoid valve. Physical meaning: Provides a normalized benchmark for current changes, eliminates the influence of individual differences. Acquisition method: The average current of the first 10 actions after a new valve completes the running-in period (such as 100 standard actions). e is a constant equal to 2.71828. Function: Constructs a temperature accelerated aging model to quantify the non-linear effect of ambient temperature on wear. T is the real-time working environment temperature of the solenoid valve (unit: °C). Physical meaning: An increase in temperature will accelerate degradation processes such as material fatigue and lubricant failure. Benchmark temperature: 25 °C (reference value for normal temperature working conditions). For every 1 °C increase in temperature, the wear index increases by approximately 1.01 times. Engineering significance of the solenoid valve dynamic wear index: Comprehensively evaluates wear through multi-dimensional coupling effects: Mechanical wear: Through N / N max quantifies the linear effect of the cumulative number of actions; Electrical aging: Through I peak / I initial reflects the degradation of the coil and magnetic circuit system. Environmental effect: Through e 0.01（T-25） characterizes the non-linear acceleration effect of temperature on material properties. WI threshold is the preset critical threshold of the solenoid valve wear index (dimensionless). Physical meaning: When the wear index of the solenoid valve reaches this threshold, it is determined that its life has ended (maintenance or replacement is required). Determination method: Set based on experimental data, manufacturer specifications, or historical failure statistics. For example, through an accelerated life test, it is found that when WI = 1.2, the spool seal failure probability exceeds 90%, then set WI threshold = 1.2. WI currentis the dynamic wear index of the solenoid valve within the current action cycle (dimensionless). Physical meaning: reflects the real-time comprehensive wear state of the solenoid valve (mechanical, electrical, environmental effects). Calculation method: see WI for the calculation formula, d( WI ) / d N is the derivative of the dynamic wear index of the solenoid valve with respect to the cumulative number of actions N (unit: times⁻¹). Physical meaning: characterizes the change rate of the wear index caused by each action, that is, the degradation rate. Calculation method: fit the WI-N curve through a sliding window (such as the last 100 actions) is the environmental correction factor (dimensionless), used to quantify the impact of the environment on the lifespan. Physical meaning: the correction of environmental factors such as humidity and pollutants on the degradation rate. = 1 - 0.05×( RH - 50%), RH is the environmental humidity.

[0041] According to the maintenance record (such as after replacing the sealing ring), automatically reset WI threshold to a new reference value to avoid cumulative errors. Use a sliding window to calculate d( WI ) / d N in real time, capture the non-linear change of the wear rate (such as low rate in the initial running-in period and accelerated failure in the later stage), and incorporate environmental parameters such as humidity and pollutants into the lifespan prediction through to solve the limitation of traditional models relying only on mechanical parameters. Application case Lifespan prediction of the solenoid valve in the nitrogen detection system of a chemical plant: 1. Equipment and parameter setting: Solenoid valve model: FESTO MHJ9-5 / 2G-1 / 8-B, application scenario: opening and closing valve in the nitrogen transmission pipeline.

[0042] Preset key parameters: Initial current peak I initial = 12.5 mA.

[0043] Rated lifespan times N max = 100,000 times.

[0044] Wear threshold WI threshold = 1.2.

[0045] Environmental temperature T = 38℃, humidity RH = 60%.

[0046] 2. Data collection and preprocessing: Sensor type: Hall current sensor, installation location: in series with the drive circuit, sampling frequency: 1000 Hz during operation; Sensor type: laser displacement sensor, installation location: 5 mm directly above the valve core, sampling frequency: 200 Hz during the action phase; Sensor type: temperature and humidity sensor, installation location: on the valve body surface, sampling frequency: once every 5 seconds; Signal processing: Current data: Apply adaptive Kalman filtering (Q = 1×10 -5 , R = 0.01) to eliminate electromagnetic interference, and the peak detection error of the filtered current waveform is <0.2 mA.

[0047] Displacement data: Use variable-weight moving average (5-point window in the startup phase, 3-point window in the high-speed phase), and the displacement jitter suppression rate is 85%.

[0048] Environmental compensation: P Correction = P Original × [1 + 0.003(38 - 25)] × [1 - 0.02(60% - 50%)] = 1.03 × 0.98 = 1.018.

[0049] 3. Feature extraction and calculation of dynamic wear index: Time-domain features: Current rise time t 10-90% = 2.8 ms (initial value 2.5 ms), displacement overshoot δ = 0.12 mm (allowable upper limit 0.15 mm).

[0050] Frequency-domain features: The energy ratio in the 2 - 4 kHz frequency band after wavelet packet decomposition is 18.7% (initial value 15.2%).

[0051] Dynamic wear index: WI current = (82,000 / 100,000) × (14.8 / 12.5) × e 0.01（38-25） = 0.82 × 1.184 × 1.139 ≈ 1.11.

[0052] 4. Construction of LSTM prediction model: Model architecture: See Table 1 below: Table 1. Model architecture of the LSTM prediction model Hierarchy Parameter Configuration Function Description Input Layer 30 Action Cycle Sequences 15 - Dimensional Features × 30 Steps Bidirectional LSTM 64 Units, return_sequences = True Capture Temporal Bidirectional Dependencies Attention Weights Focus on the Last 5 Cycles Enhance Recently Degraded Features Fully Connected Layer 32 Units, ReLU Activation Feature Fusion Output Layer 2 Branches: RUL + Failure Probability Multi - Task Prediction Model training: Training data: 60,000 action records (including 12 fault cases) Hyperparameters: Learning rate 0.001, batch_size = 64, epochs = 100 Performance indicators: MAE = 86 operations (<10% of the rated life) Fault classification accuracy rate: 93.5%

[0053] 5. Life prediction: Calculation of degradation rate: d( WI ) / d N =(1.11 - 1.08) / 100 = 0.0003 operations −1 .

[0054] Calculation of environmental correction factor: = 1 - 0.05×(60% - 50%) = 0.95

[0055] Calculation of dynamic remaining life: RUL =[(1.2 - 1.11) / 0.0003]×0.95 = 285 operations

[0056] 6. Economic benefit analysis: See Table 2 below Table 2. Comparison results of economic benefit analysis between this embodiment and the traditional method Indicator Traditional Maintenance Scheme Scheme of This Embodiment Improvement Effect Annual Average Unplanned Shutdown 5.2 Times 1.1 Times ↓ 79% Spare Part Inventory Cost $15,000 $9,800 ↓ 35% Average Valve Life 92,000 Times 108,000 Times ↑ 17% Among them, the traditional maintenance plan usually adopts time - based maintenance (TBM). Through multi - source sensor data fusion, dynamic wear index calculation and hybrid drive model, this embodiment realizes accurate life prediction of nitrogen solenoid valves. Compared with traditional regular maintenance, the maintenance cost is reduced by more than 35%, and 3 unplanned shutdown events are successfully avoided (the loss of a single shutdown is about $50,000). Verification shows that the prediction error of the solenoid valve life under complex working conditions (high temperature, high humidity) of this scheme is stable within ±10%.

[0057] In another technical solution, in S1, the electrical parameter data includes: coil current, drive voltage and power consumption; the mechanical parameter data includes: response time, spool displacement and sealing pressure; the environmental parameter data includes: working temperature, humidity and vibration intensity; the operation parameter data includes: cumulative number of operations, vibration frequency and maintenance records

[0058] In the above - mentioned technical solution, the specific acquisition methods of the electrical parameter data, mechanical parameter data and environmental parameter data are shown in Table 3 below Table 3. Specific acquisition methods of electrical parameter data, mechanical parameter data and environmental parameter data In another technical solution, in S2, the electrical parameter data eliminates electromagnetic interference noise through Kalman filtering, the mechanical parameter data is smoothed and jittered by the variable-weight moving average filtering method, and the environmental parameter data is corrected through the temperature compensation correction formula, which specifically includes: performing adaptive Kalman filtering on the coil current data, with the process noise Q = 1e-5 and the observation noise R = 0.01; the displacement signal is smoothed and jittered by the variable-weight moving average filtering. In the starting stage of the solenoid valve core, when the displacement <0.2 mm, a 5-point window is used. In the high-speed movement stage of the solenoid valve core, when the displacement ≥0.2 mm, it switches to a 3-point window; the environmental parameter data is corrected through the temperature compensation correction formula: establishing a temperature-humidity coupling correction model: P Correction = P Original × [1 + 0.003( T −25)] × [1 − 0.02( RH −50%)] to correct the temperature and humidity.

[0059] In the above technical solution, through Kalman filtering, the peak noise suppression can be reduced from the original data fluctuation of ±2.1 mA to ±0.3 mA after filtering, effectively retaining the characteristics of the current rising edge (the rising time measurement error <0.1 ms). By using the variable-weight moving average filtering for smoothing and jittering, the jitter suppression rate in the starting stage is 82%, and the response delay in the high-speed stage is <0.5 ms. By correcting through the temperature compensation correction formula, the drift error of the pressure sensor caused by temperature of 0.12 MPa is eliminated, and the deviation of the sealing performance evaluation affected by humidity is corrected by 8%.

[0060] In another technical solution, the early warning specifically includes: When the dynamic remaining life RUL is greater than 1000 times, the early warning level is normal, the early warning color identification is green, and a health report is generated; When the dynamic remaining life RUL is greater than 500 and less than 1000 times, the early warning level is attention, the early warning color identification is yellow, and the spare part procurement process is activated; When the dynamic remaining life RUL is less than 500 times, the early warning level is urgent, the early warning color identification is red, and the redundant valve switching and maintenance work order are triggered.

[0061] In the above technical solution, as the calculation result of the above example, RUL = 285, less than 500 times, the early warning level is urgent, the early warning color identification is red, and the redundant valve switching (designing a redundant path beside each solenoid valve for redundant switching) and maintenance work order are triggered: such as solenoid valve two - 2025 - 087.

[0062] In another technical solution, between step S1 and step S2, there is also a step: using a multi-copy consistent hashing storage algorithm to upload the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of each solenoid valve collected to the cloud platform for storage. The electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves are split into multiple data blocks, and the split multiple data blocks are stored on different data nodes in the cluster. Specifically, it includes: Predefine the correlation and the number of redundant copies of the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves according to the configuration file; In the Hadoop cluster of the cloud platform, calculate the hash values of different data nodes through the MapReduce program; Through the hash function shown in the following formula, map data of any length to a fixed 128-bit hash value; Hash (M) = Md5 (x) In the formula, Md5(x) represents the hash function, x represents the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves for which the hash value is to be calculated; After generating the 128-bit hash value, extract 32 bits from it as the hash value, and then map them one by one to the hash ring. Determine the storage location of the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves on the cloud platform by comparing the hash values of the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves with the hash values of different data nodes in the Hadoop cluster data.

[0063] In the above technical solution, during the daily operation of the solenoid valve, a large amount of operation data is bound to be generated. If the local calculation method is adopted and only the processed data is stored, it is difficult to ensure the accuracy of the monitoring results of the solenoid valve operation status. By introducing the cloud platform and using the multi-copy consistent hashing storage algorithm to store the data uploaded to the cloud platform, the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves uploaded are split into multiple data blocks, and the split multiple data blocks are stored on different data nodes in the cluster. This can not only meet the storage needs of a large amount of solenoid valve big data, but also facilitate the identification of the electrical parameters, mechanical parameters, environmental parameters, and operation parameter data of different solenoid valves. At the same time, the monitoring data can be processed in parallel to ensure the real-time nature of the monitoring results. Among them, when the stored monitoring data is used for prediction, it can be preprocessed, such as using the EEMD method for data denoising processing, and introducing a convolutional neural network to identify the online monitoring data to ensure the integrity and consistency of multi-dimensional data.

[0064] In another technical solution, after step S3, it further includes: establishing a concept drift detection mechanism: when the prediction confidence level is < 90% for 100 consecutive times, optimize the prediction model.

[0065] In the above technical solution, an online self-optimizing system is formed, and an edge-cloud collaborative architecture is deployed: the edge side executes feature extraction and model lightweight inference (accelerated by TensorRT), the cloud side implements model incremental learning, injects new data daily to update the network weights, and establishes a concept drift detection mechanism: when the prediction confidence level is < 90% for 100 consecutive times, automatically start the transfer learning module, retain the underlying feature extraction network, and fine-tune the top-level regressor. Additionally, the prediction system verifies and forms a physical closed-loop: implement virtual-real combination verification: Digital Twin layer: establish a valve body degradation simulation model in MATLAB / Simulink and inject 30% adversarial samples; Physical verification layer: conduct helium mass spectrometry leak detection tests and SEM microscopic morphology analysis on the valve bodies with predicted critical service lives; establish an error compensation mechanism: when MAE > 10%, start the feature re-selection process and dynamically adjust the feature weights based on SHAP value analysis; for cases where the prediction deviation > 20%, trigger the expert diagnosis mode and generate correction coefficients by integrating maintenance historical data; closed-loop optimization system design: the edge-cloud collaborative architecture realizes the continuous evolution of the prediction model, with the accuracy improved by more than 35% compared to traditional fixed models; virtual experiment verification mechanism: through the dual verification of digital twin and physical detection, ensure the robustness of the prediction system under actual working conditions.

[0066] In another technical solution, a quick-connect fitting is provided at the outlet of the nitrogen sampling cylinder to achieve quick-connect replacement of the threaded connection.

[0067] In the above technical solution, as Figure 3 shown, by modifying the valve outlet of the nitrogen sampling cylinder, replacing the threaded connection with a quick-connect fitting, and selecting a quick-connect fitting made of 316L stainless steel, which has good sealing performance, can withstand a certain pressure, and has stable chemical properties and will not contaminate the sample gas. To achieve the quick-connect effect, when simultaneously detecting oxygen, CO, CO 2 simultaneously, the system can be improved as Figure 3 follows. Connect a quick-connect fitting to the gas chromatograph as well to make it match the quick-connect fitting on the nitrogen sampling cylinder, and connect the outlet of the quantitative loop of the gas chromatograph to the input end of the micro-oxygen analyzer through a fixed pipeline, so as to simultaneously detect oxygen, CO, CO 2 simultaneously through one pipeline.

[0068] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the embodiments. It can be fully applied to various fields suitable for the present invention. For those skilled in the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to the specific details and the embodiments shown and described herein.

Claims

1. An efficient and integrated nitrogen composition analysis method, characterized in that: The following steps are involved: 1) Connecting the nitrogen sample to be tested to an automatic flow control distribution device through the outlet of the nitrogen sampling cylinder, wherein the automatic flow control distribution device includes a programmable industrial control chip, a flow control module and a plurality of solenoid valves; 2) The programmable industrial control chip controls the flow control module and the switches of multiple solenoid valves according to a preset mode, and distributes the nitrogen sample to be tested to the trace oxygen analyzer, gas chromatograph and / or chilled mirror dew point meter according to the test requirements; 3) The trace oxygen analyzer, gas chromatograph and chilled mirror dew point meter are used to analyze and detect the oxygen, other gas components except oxygen, and water components in the nitrogen sample respectively; 4) Collect and integrate the test data of the gas chromatograph, trace oxygen analyzer and / or chilled mirror dew point meter to obtain the composition analysis results of the nitrogen sample to be tested.

2. The efficient integrated nitrogen composition analysis method according to claim 1, characterized in that: The flow control module includes an electronic flow controller 1 and an electronic flow controller 2, and a plurality of solenoid valves, a solenoid valve 1, a solenoid valve 2, a solenoid valve 3, a solenoid valve 4 and a solenoid valve 5; Among them, the air inlet end of the electronic flow controller one is connected with the nitrogen sampling cylinder, and the air outlet end is connected with the air inlets of solenoid valve one, solenoid valve two and solenoid valve four. The air outlet of solenoid valve one is connected with the input end of the cold mirror dew point meter to form a first gas flow pipeline, the air outlet of solenoid valve two is connected with the input end of the trace oxygen analyzer to form a second gas flow pipeline, and the air outlet of solenoid valve four is connected with the input end of the gas chromatograph to form a third gas flow pipeline; the output end of the cold mirror dew point meter is connected with the air inlet of solenoid valve three and the input end of the trace oxygen analyzer, the output end of the trace oxygen analyzer is connected with the air inlet of solenoid valve five and the input end of the gas chromatograph, and the air outlets of solenoid valve three and solenoid valve five are connected with the air inlet end of the electronic flow controller two.

3. The efficient integrated nitrogen composition analysis method according to claim 2, characterized in that: Step 2) specifically includes: When all components in the nitrogen sample to be tested are tested, the programmable industrial control chip controls the flow rate of electronic flow controller 1 to be 200-300 mL / min, the flow rate of electronic flow controller 2 to be 200-300 mL / min, solenoid valve 1 and solenoid valve 5 are opened, and other solenoid valves are closed; the nitrogen sample to be tested is synchronously distributed to the gas chromatograph, trace oxygen analyzer and chilled mirror dew point meter; When the moisture in the nitrogen sample to be tested is tested separately, the programmable industrial control chip controls the flow rate of the electronic flow controller 1 to be 300-350 mL / min, the flow rate of the electronic flow controller 2 to be 300-350 mL / min, the solenoid valve 1 and the solenoid valve 3 are opened, and the other solenoid valves are closed, so that the nitrogen sample to be tested is only distributed to the cold mirror dew point meter; When the oxygen in the nitrogen sample to be tested is tested separately, the programmable industrial control chip controls the flow rate of the electronic flow controller 1 to be 10-100 mL / min, the flow rate of the electronic flow controller 2 to be 10-100 mL / min, the solenoid valve 2 and the solenoid valve 5 are opened, and the other solenoid valves are closed, so that the nitrogen sample to be tested is only distributed to the trace oxygen analyzer; When testing carbon monoxide and / or carbon dioxide gas in the nitrogen sample to be tested, the programmable industrial control chip controls the flow rate of electronic flow controller 1 to be 10-50 mL / min, the flow rate of electronic flow controller 2 to be 0 mL / min, solenoid valve 4 is opened, and other solenoid valves are closed, so that the nitrogen sample to be tested is only distributed to the gas chromatograph.

4. The efficient integrated nitrogen composition analysis method according to claim 1, characterized in that: Step 4) also includes a solenoid valve life prediction step, including: S1. Install a current sensor, a non-contact displacement sensor and a temperature and humidity sensor on each solenoid valve body to collect electrical parameters, mechanical parameters, environmental parameters and operating parameter data of each solenoid valve; S2, preprocessing the collected data; S3. Build a prediction model based on LSTM, and use the prediction model to predict and warn the life of each solenoid valve: in, WI is the dynamic wear index of the solenoid valve, N is the cumulative number of actions, N max is the rated life times, I peak is the peak value of the coil current in the current action cycle of the solenoid valve, I initial is the peak value of the coil current in the initial state of the solenoid valve, e is a constant, equal to 2.71828, T It is the real-time working environment temperature of the solenoid valve; RUL is the dynamic remaining life, WI threshold is the preset critical threshold of the solenoid valve wear index, WI current is the dynamic wear index of the solenoid valve in the current action cycle, d( WI ) / d N The dynamic wear index of the solenoid valve versus the cumulative number of operations N The derivative of is the environmental correction factor, =1-0.05×( RH -50%) RH is the ambient humidity.

5. The efficient integrated nitrogen composition analysis method according to claim 4, characterized in that: In S1, the electrical parameter data include: coil current, drive voltage and power consumption; the mechanical parameter data include: response time, valve core displacement and sealing pressure; the environmental parameter data include: working temperature, humidity and vibration intensity; the operating parameter data include: cumulative number of actions, vibration frequency and maintenance records.

6. The efficient integrated nitrogen composition analysis method according to claim 5, characterized in that: In S2, the electrical parameter data is filtered by Kalman filtering to eliminate electromagnetic interference noise, the mechanical parameter data is smoothed by variable weight sliding average filtering, and the environmental parameter data is corrected by the temperature compensation correction formula, including: adaptive Kalman filtering of the coil current data, process noise Q=1e-5, observation noise R=0.01; the displacement signal is smoothed by variable weight sliding average filtering. In the start-up stage of the solenoid valve core, the displacement is <0.2 mm, and a 5-point window is used. In the high-speed movement stage of the solenoid valve core, the displacement is ≥0.2 mm, and it is switched to a 3-point window; the environmental parameter data is corrected by the temperature compensation correction formula: Establish a temperature-humidity coupling correction model: P (Correction) = P (original)×[1+0.003( T −25)]×[1−0.02( RH −50%)] to correct for temperature and humidity.

7. The efficient integrated nitrogen composition analysis method according to claim 6, characterized in that: The warnings include: When the dynamic remaining life RUL When it is greater than 1,000 times, the warning level is normal, the warning color is green, and a health report is generated; When the dynamic remaining life RUL When the number is greater than 500 and less than 1000 times, the warning level is attention, the warning color is yellow, and the spare parts procurement process is activated; When the dynamic remaining life RUL When it is less than 500 times, the warning level is urgent, the warning color is red, and the redundant valve switching and maintenance work order are triggered.

8. The efficient integrated nitrogen composition analysis method according to claim 7, characterized in that: Between step S1 and step S2, there is also a step of uploading the collected electrical parameters, mechanical parameters, environmental parameters and operating parameter data of each solenoid valve to the cloud platform for storage using a multi-copy consistent hash storage algorithm, splitting the electrical parameters, mechanical parameters, environmental parameters and operating parameter data of different solenoid valves into multiple data blocks, and storing the split multiple data blocks on different data nodes in the cluster, specifically including: Predefine the correlation and redundant copy number of electrical parameters, mechanical parameters, environmental parameters and operating parameter data of different solenoid valves according to the configuration file; In the Hadoop cluster of the cloud platform, the hash values ​​of different data nodes are calculated through the MapReduce program; Through the hash function shown in the following formula, data of any length is mapped to a fixed 128-bit hash value; Hash (M) = Md5 (x) In the formula, Md5(x) represents a hash function, x Electrical parameters, mechanical parameters, environmental parameters and operating parameter data of different solenoid valves for which hash values ​​are to be calculated; After generating a 128-bit hash value, 32 bits are extracted from it as the hash value, and then mapped one by one to the hash ring. By monitoring the hash values ​​of the electrical parameters, mechanical parameters, environmental parameters and operating parameter data of different solenoid valves and comparing them with the hash values ​​of different data nodes of the Hadoop cluster data, the locations where the electrical parameters, mechanical parameters, environmental parameters and operating parameter data of different solenoid valves are stored on the cloud platform are determined.

9. The efficient integrated nitrogen composition analysis method according to claim 8, characterized in that: After step S3, the method further includes: establishing a concept drift detection mechanism: when the prediction confidence is less than 90% for 100 consecutive times, optimizing the prediction model.

10. The highly efficient integrated nitrogen composition analysis method according to claim 9, characterized in that: A quick-connect connector is provided on the outlet of the nitrogen sampling cylinder to replace the threaded connection with a quick-connect connection.

Images