Gas purification real-time monitoring method, system and device and medium

Abstract

The invention discloses a gas purification real-time monitoring method, system and device and a medium, and belongs to the technical field of gas purification. The method comprises the following steps: a data acquisition stage: acquiring inlet gas flow and outlet component data; in the adsorbent state monitoring stage, the adsorbent state is monitored in real time through the monitoring point position in the adsorption barrel; in the adsorption capacity calculation stage, the residual adsorption capacity is predicted through an impurity penetration curve model on the basis of inlet flow and outlet component data, and bidirectional verification and correction are carried out on the residual adsorption capacity and an internal monitoring state; in the regeneration time prediction stage, the optimal regeneration time is dynamically predicted according to discharged gas component analysis and residual adsorption capacity; in the real-time regulation and control stage, parameters of the purifier and the regenerative heater are adjusted in real time according to the prediction result. Accurate evaluation of the state of the adsorption cylinder and intelligent optimization of the regeneration process are achieved, the utilization rate of the adsorption cylinder is remarkably increased, and the total energy consumption of regeneration is reduced.

Description

Technical Field

[0001] This invention relates to the field of gas purification technology, and in particular to a method, system, device and medium for real-time monitoring of gas purification. Background Technology

[0002] Regenerative gas purifiers (such as pressure swing adsorption and temperature swing adsorption devices) are core equipment in the field of industrial gas separation and purification. They achieve continuous gas purification through the periodic adsorption and desorption of impurities by adsorbents. The key to the operating efficiency and energy consumption of such equipment lies in the accurate judgment of the state of the adsorption cylinder (especially the remaining adsorption capacity) and the intelligent control of the regeneration process (such as heating and purging).

[0003] Currently, common adsorption cylinder status monitoring and regeneration control strategies have the following limitations: First, most systems rely on simple fixed time periods or trigger mechanisms based on a single threshold of outlet concentration to initiate regeneration. This method fails to fully utilize the dynamic adsorption capacity of the adsorbent, easily leading to "premature regeneration" resulting in low adsorbent utilization and increased energy consumption, or "late regeneration" causing the risk of substandard outlet gas. Second, the regeneration process typically employs an "open-loop" control strategy based on a fixed temperature curve or time, failing to dynamically adjust according to the actual desorption load of the adsorbent (i.e., impurity desorption rate). This often results in insufficient heating in the early stages of regeneration leading to incomplete desorption, or excessive heating and purging in the later stages of regeneration leading to significant energy waste. Furthermore, existing technologies lack effective online diagnostic capabilities for adsorbent performance degradation (such as poisoning and coking) and sensor status. The system cannot distinguish whether abnormal outlet concentration stems from adsorption saturation, adsorbent failure, or monitoring instrument malfunction, which is detrimental to preventative maintenance and operational safety. Finally, traditional methods lack self-optimization capabilities; once control parameters are set, they are relatively fixed and cannot adapt to long-term changes such as fluctuations in feed gas load and natural aging of the adsorbent, making it difficult to maintain optimal operating conditions continuously.

[0004] Therefore, there is an urgent need to develop a real-time monitoring method for gas purification that can achieve accurate status monitoring, intelligent regeneration control, real-time fault diagnosis, and self-learning optimization capabilities, in order to solve the above problems and maximize the utilization rate of the adsorption cartridge and minimize the energy consumption of the regeneration process. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, system, device and medium for real-time monitoring of gas purification.

[0006] The objective of this invention is achieved through the following technical solution: The first aspect of this invention provides: a real-time monitoring method for gas purification, used to monitor a regenerative purifier, comprising the following steps: Data acquisition phase: Inlet gas flow rate data and outlet composition data are collected by a flow sensor installed at the inlet of the adsorption cylinder and a gas chromatograph installed at the outlet of the adsorption cylinder. Adsorbent status monitoring stage: The adsorbent status is monitored in real time by setting multiple monitoring points inside the adsorption cylinder. These monitoring points are equipped with temperature sensors or component monitoring probes. Adsorption capacity calculation stage: Based on inlet gas flow data and outlet composition data, the remaining adsorption capacity of the adsorption cylinder is predicted by the impurity penetration curve model. According to the real-time monitoring of the adsorbent state, cross-comparison and mutual verification are performed to determine whether there are any abnormalities in the remaining adsorption capacity and adsorbent state. Then, the remaining adsorption capacity is corrected and / or the monitoring point abnormalities are reported based on the verification results. Regeneration time prediction stage: Based on the composition analysis of the gas discharged from the adsorption tube during the purification process and the remaining adsorption capacity, the optimal regeneration time is dynamically predicted; Real-time control stage: The remaining adsorption capacity and optimal regeneration time are transmitted to the monitoring system to adjust the operating parameters of the regenerative purifier and regeneration heater in real time, so that they are always in the current optimal working state at different working stages.

[0007] Preferably, the temperature sensors are evenly distributed along the airflow direction of the adsorption cylinder, and the component monitoring probes are located in the middle and at the outlet of the adsorption cylinder.

[0008] Preferably, the remaining adsorption capacity is predicted through the following steps: At each preset interval, the total amount of impurities entering the adsorption cylinder per unit time is calculated in real time based on the inlet gas flow data, and the impurity concentration change trend is calculated based on the outlet composition data. When the outlet impurity concentration reaches a preset percentage of the inlet impurity concentration, the adsorption capacity is determined to be close to saturation. At this point, the remaining adsorption capacity of the adsorption cylinder is predicted by the impurity penetration curve model, thereby reducing the computational resource consumption.

[0009] Preferably, the impurity penetration curve model is constructed through the following steps: The adsorption isotherm of the adsorbent for the target impurity was obtained through experimental measurement. An adsorption rate equation was established based on the adsorbent characteristics and gas flow parameters. Based on the adsorption isotherm and adsorption rate equation, an impurity breakthrough curve model is constructed, and the model parameters are dynamically corrected according to real-time monitoring data to improve prediction accuracy.

[0010] Preferably, the adsorption capacity calculation stage further includes the following steps: When the monitoring system receives abnormal information from the monitoring point, it automatically diagnoses the cause of the fault, determines whether it is a sensor malfunction or an adsorbent failure, and executes the corresponding fault repair process.

[0011] Preferably, during regeneration, a gradient control strategy is employed for the power of the regeneration heater, including the following steps: Regeneration triggering steps: When the remaining adsorption capacity is lower than the first preset threshold, a regeneration suggestion is generated; when the remaining adsorption capacity is lower than the second preset threshold, which is lower than the first preset threshold, the regeneration process is automatically executed. Regeneration preheating step: After the regeneration process is executed, the regeneration heater is started with preheating power to preheat the regeneration gas flow, while the outlet gas temperature of the adsorption cylinder is monitored; when the outlet gas temperature stabilizes and rises to the set value, the preheating is considered complete. Variable power control steps: After preheating, the composition analyzer installed on the regeneration exhaust pipe monitors the change in the concentration of desorbed impurities; when a sharp increase in the concentration of desorbed impurities is detected, the regeneration heater is controlled to operate at the first power level to provide the energy required for desorption; when the concentration of desorbed impurities is detected to reach its peak and begin to decrease, the power of the regeneration heater is switched to a second power level lower than the first power level to maintain the necessary desorption temperature. Post-regeneration optimization and termination steps: Continuously monitor the concentration of desorbed impurities. When it drops to the preset concentration, further reduce the power of the regeneration heater to the third power level and perform heat preservation and purging. Calculate the total amount of desorbed impurities per unit time in real time. When this value is lower than the preset critical value, the regeneration is determined to be complete, and the regeneration heater is turned off.

[0012] Preferably, after regeneration is completed, the following steps are also included: Regeneration effect evaluation and model self-learning steps: After regeneration is completed, record the total energy consumption, effective regeneration time and initial adsorption performance of the regenerated adsorption cartridge, and compare it with historical data. Based on the comparison results, adaptively correct the impurity breakthrough curve model parameters and regeneration control parameters for the next cycle. If the initial adsorption performance of the regenerated adsorption cartridge is lower than expected, the monitoring system will automatically generate an alarm, indicating possible adsorbent failure or equipment malfunction, and automatically adjust the regeneration control strategy.

[0013] Preferably, the first and second preset thresholds are set based on the design adsorption capacity of the adsorption cartridge, the actual impurity load of the current purified gas, and the desired adsorption cartridge utilization safety factor; the specific values ​​of the first, second, and third power levels are calculated and set in real time based on the specifications of the adsorption cartridge, the type of adsorbent, and the regeneration gas flow rate through an energy consumption optimization algorithm; the preset critical value is dynamically set based on the total amount of impurities adsorbed by the adsorption cartridge in the previous adsorption cycle.

[0014] Preferably, the regeneration effect evaluation and model self-learning steps further include the following steps: Adsorption cartridge performance recovery evaluation steps: After the regeneration cycle is completed and the adsorption cartridge is put into the next purification cycle, analyze the rate of change of impurity concentration in the outlet gas at the beginning of the purification cycle, compare it with the performance data of the benchmark adsorption cartridge, and calculate the adsorbent performance recovery coefficient. Data correlation and analysis steps: Correlate the efficiency data of this regeneration with the adsorption cylinder state data before regeneration. The adsorption cylinder state data includes the total amount of impurities adsorbed in the previous purification cycle and the predicted remaining adsorption capacity before regeneration. Model parameter adaptive correction steps: Based on the correlation analysis results, the impurity breakthrough curve model parameters and regeneration control parameters are dynamically corrected; if the energy consumption of this regeneration is higher than the historical average level or the preset energy consumption threshold, but the adsorbent performance recovery coefficient is higher than the first preset coefficient threshold, it is determined that there is room for optimization in the regeneration process. The concentration threshold or temperature set point of heating power switching in the regeneration control parameters is automatically fine-tuned to seek a better energy efficiency ratio strategy and use it for the next regeneration; if the adsorbent performance recovery coefficient after this regeneration is lower than the second preset coefficient threshold, it is determined that the adsorbent has irreversible deactivation or that there was over-adsorption in the previous purification cycle. Based on this, a deactivation warning is issued or the remaining adsorption capacity threshold for triggering regeneration in the next cycle is adjusted accordingly. Knowledge base update steps: The pre-regeneration state, regeneration process control parameters, and post-regeneration performance evaluation results of this regeneration cycle are stored as a data sample in the historical knowledge base for use in optimizing decisions for subsequent regeneration cycles.

[0015] Preferably, the adsorbent performance recovery coefficient is calculated as follows: under the same inlet impurity concentration and flow rate conditions, the ratio of the time required for the current adsorption cylinder to reach the same outlet purity standard or the total amount of impurities processed at the beginning of the purification cycle to the reference adsorption cylinder is the adsorbent performance recovery coefficient.

[0016] Preferably, in the model parameter adaptive correction step, if the monitoring system detects a continuous downward trend in the adsorbent performance recovery coefficient for a preset number of consecutive times, it will automatically trigger an advanced diagnostic alarm, indicating that the adsorbent is deactivated and needs to be replaced or requires manual diagnosis.

[0017] A second aspect of the present invention provides: a real-time gas purification monitoring system for implementing any of the above-described real-time gas purification monitoring methods, comprising: The data acquisition module is used to collect inlet gas flow data and outlet composition data through a flow sensor installed at the inlet of the adsorption cylinder and a gas chromatograph installed at the outlet of the adsorption cylinder. The adsorbent state monitoring module is used to monitor the state of the adsorbent in real time through multiple monitoring points set inside the adsorption cylinder. The monitoring points are equipped with temperature sensors or component monitoring probes. The adsorption capacity calculation module is used to predict the remaining adsorption capacity of the adsorption cylinder based on the inlet gas flow data and outlet composition data, using an impurity penetration curve model. It also performs cross-comparison and mutual verification based on the real-time monitored adsorbent state to determine whether there are any abnormalities in the remaining adsorption capacity and adsorbent state. Then, it corrects the remaining adsorption capacity and / or reports abnormalities at monitoring points based on the verification results. The regeneration time prediction module is used to dynamically predict the optimal regeneration time based on the composition analysis of the gas discharged from the adsorption cartridge during the purification process and the remaining adsorption capacity. The real-time control module transmits the remaining adsorption capacity and optimal regeneration time to the monitoring system, and adjusts the operating parameters of the regenerative purifier and regeneration heater in real time to ensure that they are always in the current optimal working state at different working stages.

[0018] A third aspect of the present invention provides: a gas purification real-time monitoring device, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement any of the above-described gas purification real-time monitoring methods.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when loaded and executed by a processor, implement any of the above-described gas purification real-time monitoring methods.

[0020] The fifth aspect of the present invention provides: a computer program product containing instructions, which, when run on a terminal, causes the terminal to execute any of the above-described gas purification real-time monitoring methods.

[0021] The beneficial effects of this invention are: 1) Through high-frequency data acquisition and a precise impurity penetration curve model, the system can predict the actual remaining adsorption capacity of the adsorption cartridge in real time and dynamically, fundamentally changing the traditional extensive mode of regeneration based on fixed time or simple threshold triggering. Combined with bidirectional verification from internal monitoring points, the system can utilize the adsorption cartridge to near its theoretical adsorption capacity while ensuring safety, thereby significantly extending the single adsorption cycle and reducing unnecessary regeneration times. This maximizes the effective utilization rate of the adsorbent and directly reduces the adsorbent lifespan reduction and spare parts replacement costs caused by frequent regeneration.

[0022] 2) Through a gradient power control strategy based on real-time feedback of desorbed impurity concentration, the system can automatically identify different stages of the desorption process (rapid desorption period, gradual desorption period, and purging period) and dynamically match the most suitable heating power. This avoids the energy waste that occurs in the later stages of desorption in traditional fixed-power heating methods. Simultaneously, combined with self-learning optimization, the system can continuously find the optimal energy consumption control curve under specific operating conditions, thereby significantly reducing the total power consumption of the regeneration heater and achieving significant energy savings.

[0023] 3) The introduction of an internal state monitoring and two-way verification mechanism for the adsorption cylinder enables the system not only to predict capacity but also to diagnose anomalies in real time. When there is a significant deviation between the model prediction and the sensor inversion value, the system can automatically determine whether it is a sensor malfunction or adsorbent performance degradation and promptly report an early warning. This capability transforms traditional "reactive maintenance" into "predictive maintenance," effectively avoiding gas purity accidents caused by sudden adsorbent failure or instrument malfunction, and greatly improving the operational safety and reliability of the entire purification system.

[0024] 4) After each regeneration, the system automatically assesses the degree of adsorbent performance recovery and correlates it with historical data. This allows for adaptive correction of key predictive model parameters (such as the mass transfer coefficient) and control parameters (such as the power switching threshold). This enables the system to automatically adapt to fluctuations in feed gas load, environmental changes, and the natural aging of the adsorbent over time, ensuring that the control strategy remains at or near its current optimal state. This achieves performance self-optimization throughout the entire lifecycle and solves the problem of performance degradation over time in traditional fixed-parameter systems.

[0025] 5) By integrating model calculation, status assessment, fault diagnosis, and optimized control into an automated monitoring system, this invention significantly reduces reliance on operator experience. The system can automatically provide regeneration suggestions, execute optimized regeneration procedures, and offer clear diagnostic information in case of anomalies. This not only reduces the workload of personnel but also minimizes operational risks caused by human error or improper operation, making the management and control of high-precision gas purification processes more intelligent and convenient. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention. Figure 2 A flowchart for dynamic prediction of optimal regeneration time; Figure 3 Flowchart of the power gradient control strategy for the regenerative heater; Figure 4 This is a flowchart for evaluating the regeneration effect and the model's self-learning process. Detailed Implementation

[0027] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] See Figures 1-4 The first aspect of this invention provides: a method for real-time monitoring of gas purification, used to monitor a regenerative purifier, comprising the following steps: Data acquisition phase: Inlet gas flow rate data and outlet composition data are collected by a flow sensor installed at the inlet of the adsorption cylinder and a gas chromatograph installed at the outlet of the adsorption cylinder. Adsorbent status monitoring stage: The adsorbent status is monitored in real time by setting multiple monitoring points inside the adsorption cylinder. These monitoring points are equipped with temperature sensors or component monitoring probes. Adsorption capacity calculation stage: Based on inlet gas flow data and outlet composition data, the remaining adsorption capacity of the adsorption cylinder is predicted by the impurity penetration curve model. According to the real-time monitoring of the adsorbent state, cross-comparison and mutual verification are performed to determine whether there are any abnormalities in the remaining adsorption capacity and adsorbent state. Then, the remaining adsorption capacity is corrected and / or the monitoring point abnormalities are reported based on the verification results. Regeneration time prediction stage: Based on the composition analysis of the gas discharged from the adsorption tube during the purification process and the remaining adsorption capacity, the optimal regeneration time is dynamically predicted; Real-time control stage: The remaining adsorption capacity and optimal regeneration time are transmitted to the monitoring system to adjust the operating parameters of the regenerative purifier and regeneration heater in real time, so that they are always in the current optimal working state at different working stages.

[0029] In this embodiment, a regenerative nitrogen deoxygenation and purification device is used as an example. The core adsorption cylinder of this device is filled with a copper-based adsorbent to remove trace amounts of oxygen (O2) from nitrogen. Implementing this invention requires the following components: a data acquisition unit, including a high-precision vortex flow meter (such as E+H's Promass 83 series) installed at the inlet of the adsorption cylinder, a miniature gas chromatograph (such as Siemens' Maxum edition II, equipped with a PDHID detector) installed at the outlet of the adsorption cylinder, and a monitoring network installed inside the adsorption cylinder.

[0030] Monitoring network: Multiple PT100 platinum resistance temperature sensors are evenly arranged along the airflow direction in the adsorption cylinder (four are used as an example in this embodiment, but the number is not limited and can be increased or decreased), and a laser oxygen analysis probe (such as the Lox series of SST Sensing) is embedded in the middle and outlet of the adsorption cylinder for real-time measurement of local oxygen concentration.

[0031] Regeneration unit: includes an electrically heated regeneration heater, a regeneration gas flow control valve, and a laser gas analyzer installed on the regeneration exhaust pipe (used to monitor the concentration of O2 and H2O in the desorbed gas).

[0032] Control and Processing Unit: Employs an architecture where an industrial PC and a PLC work together. The industrial PC runs advanced algorithms (adsorption model, diagnostic logic, self-learning algorithm), while the PLC handles high-speed data acquisition, interlocking control, and power regulation. Data exchange between the two occurs via Ethernet.

[0033] like Figure 1 As shown, real-time monitoring includes the following steps: 1. Data Acquisition Phase: The flow meter collects the inlet nitrogen volumetric flow rate at a preset frequency and sends the data to the PLC. The outlet gas chromatograph measures the O2 concentration in the outlet gas through multiple sampling-analysis cycles, and the data is uploaded to the industrial PC via the Modbus TCP protocol. Four temperature sensors and two oxygen analysis probes inside the adsorption cartridge transmit data to the PLC.

[0034] 2. Adsorbent Status Monitoring Stage: The system displays and records the axial temperature profile of the adsorption cylinder in real time. During the adsorption stage, as the adsorption front advances, a distinct "temperature peak" will be observed moving from the inlet to the outlet. The oxygen probe in the middle is used for early warning. When the adsorption front reaches the middle, the oxygen concentration at that point will begin to rise slowly earlier than at the outlet. The oxygen probe data at the outlet is cross-checked with the chromatograph data to ensure the reliability of the outlet concentration monitoring.

[0035] 3. Adsorption Capacity Calculation Stage: Model Construction and Initialization: Before system commissioning, the adsorption isotherm of O2 by the copper-based adsorbent is experimentally determined. Based on the adsorption bed parameters (bed height, diameter, adsorbent particle size, bulk density) and design conditions (flow rate, pressure, temperature), an initial O2 breakthrough curve model is established using adsorption mass transfer kinetics and pre-loaded into an industrial PC. Real-time Prediction: The system calculates the total cumulative adsorbed O2 every second. When the O2 concentration measured by the outlet chromatograph continuously rises from the background value (e.g., <10 ppb) to 5% (i.e., 5 ppm) of the inlet concentration (assumed to be 100 ppm), a precise prediction is triggered. The model combines the current inlet flow rate, cumulative load, and outlet concentration change trend to calculate and output the remaining adsorption capacity (unit: grams of O2) in real time.

[0036] Two-way verification and diagnosis: Verification: Simultaneously, the system uses temperature data from various temperature sensors to back-calculate the current adsorption capacity using an adsorption heat balance model. The model-predicted remaining capacity is then compared with the adsorption capacity calculated from the temperature.

[0037] Diagnosis and Handling: Set a deviation threshold of 15%. If the deviation is within the threshold, adopt the model prediction. If the model predicts a significantly larger remaining capacity than the calculated temperature value, and the central oxygen probe reading is normal, a malfunction of the outlet chromatograph or outlet oxygen probe may be suspected, and the system will report "Outlet Monitoring Abnormality." If the model prediction is significantly smaller than the calculated temperature value, and all temperature points show weak adsorption exothermic effects, a severe decline in adsorbent activity is highly suspected, and the system will report "Adsorbent Performance Abnormality," and recommend early regeneration or maintenance.

[0038] 4. Regeneration Time Prediction Stage: When the remaining adsorption capacity drops to 20% of the design capacity (safety threshold), the system generates a regeneration suggestion on the operation interface; when it drops to 10%, if the equipment is in automatic mode, the regeneration program will be automatically queued and started. When predicting the regeneration time, not only the current remaining adsorption capacity (desorption load) is considered, but also the efficiency of the previous regeneration and the regeneration temperature curve set for this time are referenced. The system will provide an estimated time range.

[0039] Real-time control phase: Gradient power control during regeneration: Preheating steps: After regeneration is started, the heater operates at 30% power, and the purge gas (pure nitrogen) passes through the adsorption cylinder at a low flow rate. Preheating is complete when the outlet temperature of the adsorption cylinder stabilizes at the preset preheating temperature.

[0040] Variable power main regeneration: Switch to the set regeneration temperature. The regeneration exhaust analyzer monitors the O2 concentration. When the O2 concentration begins to rise sharply, it indicates that desorption has entered its peak period, and the heater switches to 100% power to quickly provide energy. When the O2 concentration reaches its peak and begins to decline steadily, it indicates that most of the O2 has been desorbed, and the heating power switches to 60% to maintain the bed temperature.

[0041] Optimization Termination: When the oxygen concentration drops to 5% of the peak value, the later stage of purging begins. The heating power is reduced to 20%. The system calculates the average desorption rate over the past 5 minutes. When this rate falls below the critical value, regeneration is considered complete, the heater is turned off, and the bed is cooled again using cold purging gas.

[0042] In some embodiments, the temperature sensors are evenly distributed along the airflow direction of the adsorption cylinder, and the component monitoring probe is located in the middle and at the outlet of the adsorption cylinder.

[0043] In some embodiments, the remaining adsorption capacity is predicted by the following steps: At each preset interval, the total amount of impurities entering the adsorption cylinder per unit time is calculated in real time based on the inlet gas flow data, and the impurity concentration change trend is calculated based on the outlet composition data. When the outlet impurity concentration reaches a preset percentage of the inlet impurity concentration, the adsorption capacity is determined to be close to saturation. At this point, the remaining adsorption capacity of the adsorption cylinder is predicted by the impurity penetration curve model, thereby reducing the computational resource consumption.

[0044] In this embodiment, as Figure 2 As shown, the dynamic prediction of the optimal regeneration time in the regeneration time prediction stage includes the following steps: Basic time estimation steps: Based on the remaining adsorption capacity and the preset regeneration process conditions, the baseline regeneration time (T_base) required to completely desorb the impurities corresponding to the remaining adsorption capacity under the preset regeneration process conditions is calculated using the desorption kinetic model. Influencing factor correction steps: Obtain real-time operating condition parameters and equipment status parameters that may affect the actual regeneration efficiency, and dynamically correct the baseline regeneration time (T_base) based on a preset correction algorithm or correction coefficient library to obtain the preliminary predicted regeneration time (T_pred1); The real-time operating condition parameters and equipment status parameters include at least: the actual available heating power of the current regeneration heater, the real-time flow rate and inlet temperature of the regeneration gas (purge gas), and the current average bed temperature of the adsorption cylinder; Process feedback adjustment steps: After the regeneration program is actually started and running, monitor the change rate of impurity concentration (desorption rate) in the regeneration exhaust gas in real time; compare the actual initial desorption rate monitored with the predicted initial desorption rate predicted by the desorption kinetic model under the same conditions; if the actual initial desorption rate is significantly lower than the predicted initial desorption rate, extend the preliminary predicted regeneration time (T_pred1) based on the rate deviation ratio to generate the updated predicted regeneration time (T_pred2). Safety and optimization boundary constraint steps: Set a minimum time value (T_min) and a maximum time value (T_max) for the updated predicted regeneration time (T_pred2); the minimum time value (T_min) is the shortest time required to ensure basic regeneration effect, and the maximum time value (T_max) is the longest allowable time set based on energy consumption economy; the final "optimal regeneration time" output value is constrained within the interval [T_min, T_max]. Continuous rolling prediction step: During the regeneration process, the average desorption rate and remaining impurity load within the previous preset time interval are used as inputs. The process feedback adjustment step is repeatedly executed to predict and update the remaining required regeneration time until regeneration is completed.

[0045] In the step of correcting influencing factors, the correction logic for the baseline regeneration time (T_base) specifically includes the following steps: when the real-time flow rate of the regenerated gas is lower than the design value or its inlet temperature is lower than the design temperature, the baseline regeneration time is extended accordingly; when the actual available heating power of the regeneration heater is lower than the rated power due to the decline in the aging of the equipment, the baseline regeneration time is extended accordingly; when the current average bed temperature of the adsorption cylinder is higher than the ambient temperature, the baseline regeneration time in the preheating stage can be shortened accordingly.

[0046] In the process feedback adjustment step, the threshold for determining whether the actual initial desorption rate is "significantly lower" than the predicted initial desorption rate is: the proportion of the actual initial desorption rate being lower than the predicted initial desorption rate exceeds 15%.

[0047] In the safety and optimization boundary constraint step, the minimum time value (T_min) is determined based on the technical data and / or historical successful regeneration data provided by the adsorbent manufacturer; the maximum time value (T_max) is dynamically determined based on the economic analysis model between the energy consumption cost per unit time of regeneration and the opportunity cost of early regeneration of the adsorption cylinder, and can be adjusted at any time.

[0048] The optimal regeneration time obtained through dynamic prediction will serve as a reference for the overall time frame of the gradient power control strategy in the real-time regulation phase. It will be used to initially allocate the estimated time proportion of each power phase and will be dynamically adjusted in the continuous rolling prediction step.

[0049] In some embodiments, the impurity penetration curve model is constructed through the following steps: The adsorption isotherm of the adsorbent for the target impurity was obtained through experimental measurement. An adsorption rate equation was established based on the adsorbent characteristics and gas flow parameters. Based on the adsorption isotherm and adsorption rate equation, an impurity breakthrough curve model is constructed, and the model parameters are dynamically corrected according to real-time monitoring data to improve prediction accuracy.

[0050] In some embodiments, the adsorption capacity calculation stage further includes the following steps: When the monitoring system receives abnormal information from the monitoring point, it automatically diagnoses the cause of the fault, determines whether it is a sensor malfunction or an adsorbent failure, and executes the corresponding fault repair process.

[0051] In some embodiments, such as Figure 3 As shown, during regeneration, a gradient control strategy is employed for the power of the regeneration heater, including the following steps: Regeneration triggering steps: When the remaining adsorption capacity is lower than the first preset threshold, a regeneration suggestion is generated; when the remaining adsorption capacity is lower than the second preset threshold, which is lower than the first preset threshold, the regeneration process is automatically executed. Regeneration preheating step: After the regeneration process is executed, the regeneration heater is started with preheating power to preheat the regeneration gas flow, while the outlet gas temperature of the adsorption cylinder is monitored; when the outlet gas temperature stabilizes and rises to the set value, the preheating is considered complete. Variable power control steps: After preheating, the composition analyzer installed on the regeneration exhaust pipe monitors the change in the concentration of desorbed impurities; when a sharp increase in the concentration of desorbed impurities is detected, the regeneration heater is controlled to operate at the first power level to provide the energy required for desorption; when the concentration of desorbed impurities is detected to reach its peak and begin to decrease, the power of the regeneration heater is switched to a second power level lower than the first power level to maintain the necessary desorption temperature. Post-regeneration optimization and termination steps: Continuously monitor the concentration of desorbed impurities. When it drops to the preset concentration, further reduce the power of the regeneration heater to the third power level and perform heat preservation and purging. Calculate the total amount of desorbed impurities per unit time in real time. When this value is lower than the preset critical value, the regeneration is determined to be complete, and the regeneration heater is turned off.

[0052] In some embodiments, after regeneration is complete, the following steps are also included: Regeneration effect evaluation and model self-learning steps: After regeneration is completed, record the total energy consumption, effective regeneration time and initial adsorption performance of the regenerated adsorption cartridge, and compare it with historical data. Based on the comparison results, adaptively correct the impurity breakthrough curve model parameters and regeneration control parameters for the next cycle. If the initial adsorption performance of the regenerated adsorption cartridge is lower than expected, the monitoring system will automatically generate an alarm, indicating possible adsorbent failure or equipment malfunction, and automatically adjust the regeneration control strategy.

[0053] In this embodiment, after regeneration is completed, the system records the total energy consumption, effective high-temperature time, and purge gas consumption for this regeneration.

[0054] At the start of the next adsorption cycle, the system pays special attention to the outlet O2 concentration in the initial stage (e.g., the first 30 minutes). This concentration is compared to the "baseline performance curve" (data when the adsorbent is new or in good condition), and the performance recovery coefficient η (the ratio of time required to reach the same purity) is calculated. If η > 0.95, the recovery is considered good; if 0.8 < η < 0.95, a slight degradation is considered; if η < 0.8, a performance alarm is issued.

[0055] Model calibration: The industrial PC performs correlation analysis on the "pre-regeneration load," "regeneration energy consumption," and "post-regeneration η." If the regeneration energy consumption is high but η is good, the self-learning algorithm will attempt to fine-tune the regeneration control parameters. For example, it might reduce the main regeneration stage power from 100% to 95%, or slightly advance the O2 concentration threshold for switching to 60% power, to explore more energy-efficient strategies and apply them to the next regeneration. If η remains low, the system will gradually increase the remaining capacity threshold that triggers regeneration (e.g., from 10% to 15%) to provide a larger safety margin for adsorbents with declining performance and prompt "Adsorbent activity has decreased; replacement is recommended." All process data is stored in a historical knowledge base for long-term trend analysis and model optimization.

[0056] In some embodiments, the first preset threshold and the second preset threshold are set according to the design adsorption capacity of the adsorption cartridge, the actual impurity load of the current purified gas, and the expected adsorption cartridge utilization safety factor; the specific values ​​of the first power level, the second power level, and the third power level are calculated and set in real time based on the specifications of the adsorption cartridge, the type of adsorbent, and the regeneration gas flow rate through an energy consumption optimization algorithm; the preset critical value is dynamically set according to the total amount of impurities adsorbed by the adsorption cartridge in the previous adsorption cycle.

[0057] In some embodiments, such as Figure 4 As shown, the regeneration effect evaluation and model self-learning steps also include the following steps: Adsorption cartridge performance recovery evaluation steps: After the regeneration cycle is completed and the adsorption cartridge is put into the next purification cycle, analyze the rate of change of impurity concentration in the outlet gas at the beginning of the purification cycle, compare it with the performance data of the benchmark adsorption cartridge, and calculate the adsorbent performance recovery coefficient. Data correlation and analysis steps: Correlate the efficiency data of this regeneration with the adsorption cylinder state data before regeneration. The adsorption cylinder state data includes the total amount of impurities adsorbed in the previous purification cycle and the predicted remaining adsorption capacity before regeneration. Model parameter adaptive correction steps: Based on the correlation analysis results, the impurity breakthrough curve model parameters and regeneration control parameters are dynamically corrected; if the energy consumption of this regeneration is higher than the historical average level or the preset energy consumption threshold, but the adsorbent performance recovery coefficient is higher than the first preset coefficient threshold, it is determined that there is room for optimization in the regeneration process. The concentration threshold or temperature set point of heating power switching in the regeneration control parameters is automatically fine-tuned to seek a better energy efficiency ratio strategy and use it for the next regeneration; if the adsorbent performance recovery coefficient after this regeneration is lower than the second preset coefficient threshold, it is determined that the adsorbent has irreversible deactivation or that there was over-adsorption in the previous purification cycle. Based on this, a deactivation warning is issued or the remaining adsorption capacity threshold for triggering regeneration in the next cycle is adjusted accordingly. Knowledge base update steps: The pre-regeneration state, regeneration process control parameters, and post-regeneration performance evaluation results of this regeneration cycle are stored as a data sample in the historical knowledge base for use in optimizing decisions for subsequent regeneration cycles.

[0058] In some embodiments, the adsorbent performance recovery coefficient is calculated as follows: under the same inlet impurity concentration and flow rate conditions, the ratio of the time required for the current adsorption cartridge to reach the same outlet purity standard or the total amount of impurities processed at the beginning of the purification cycle to the reference adsorption cartridge is the adsorbent performance recovery coefficient.

[0059] In some embodiments, during the model parameter adaptive correction step, if the monitoring system detects a continuous downward trend in the adsorbent performance recovery coefficient for a preset number of consecutive times, an advanced diagnostic alarm will be automatically triggered, indicating that the adsorbent is deactivated and needs to be replaced or requires manual diagnosis.

[0060] A second aspect of the present invention provides: a real-time gas purification monitoring system for implementing any of the above-described real-time gas purification monitoring methods, comprising: The data acquisition module is used to collect inlet gas flow data and outlet composition data through a flow sensor installed at the inlet of the adsorption cylinder and a gas chromatograph installed at the outlet of the adsorption cylinder. The adsorbent state monitoring module is used to monitor the state of the adsorbent in real time through multiple monitoring points set inside the adsorption cylinder. The monitoring points are equipped with temperature sensors or component monitoring probes. The adsorption capacity calculation module is used to predict the remaining adsorption capacity of the adsorption cylinder based on the inlet gas flow data and outlet composition data, using an impurity penetration curve model. It also performs cross-comparison and mutual verification based on the real-time monitored adsorbent state to determine whether there are any abnormalities in the remaining adsorption capacity and adsorbent state. Then, it corrects the remaining adsorption capacity and / or reports abnormalities at monitoring points based on the verification results. The regeneration time prediction module is used to dynamically predict the optimal regeneration time based on the composition analysis of the gas discharged from the adsorption cartridge during the purification process and the remaining adsorption capacity. The real-time control module transmits the remaining adsorption capacity and optimal regeneration time to the monitoring system, and adjusts the operating parameters of the regenerative purifier and regeneration heater in real time to ensure that they are always in the current optimal working state at different working stages.

[0061] A third aspect of the present invention provides: a gas purification real-time monitoring device, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement any of the above-described gas purification real-time monitoring methods.

[0062] A fourth aspect of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when loaded and executed by a processor, implement any of the above-described gas purification real-time monitoring methods.

[0063] The fifth aspect of the present invention provides: a computer program product containing instructions, which, when run on a terminal, causes the terminal to execute any of the above-described gas purification real-time monitoring methods.

[0064] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for real-time monitoring of gas purification for monitoring a regenerative purifier, characterized by: Includes the following steps: Data acquisition phase: Inlet gas flow rate data and outlet composition data are collected by a flow sensor installed at the inlet of the adsorption cylinder and a gas chromatograph installed at the outlet of the adsorption cylinder. Adsorbent status monitoring stage: The adsorbent status is monitored in real time by setting multiple monitoring points inside the adsorption cylinder. These monitoring points are equipped with temperature sensors or component monitoring probes. Adsorption capacity calculation stage: Based on inlet gas flow data and outlet composition data, the remaining adsorption capacity of the adsorption cylinder is predicted by the impurity penetration curve model. According to the real-time monitoring of the adsorbent state, cross-comparison and mutual verification are performed to determine whether there are any abnormalities in the remaining adsorption capacity and adsorbent state. Then, the remaining adsorption capacity is corrected and / or the monitoring point abnormalities are reported based on the verification results. Regeneration time prediction stage: Based on the composition analysis of the gas discharged from the adsorption tube during the purification process and the remaining adsorption capacity, the optimal regeneration time is dynamically predicted; Real-time control stage: The remaining adsorption capacity and optimal regeneration time are transmitted to the monitoring system to adjust the operating parameters of the regenerative purifier and regeneration heater in real time, so that they are always in the current optimal working state at different working stages.

2. The method of claim 1, wherein: The temperature sensors are evenly distributed along the airflow direction of the adsorption cylinder, and the component monitoring probes are located in the middle and at the outlet of the adsorption cylinder.

3. The method of claim 1, wherein: The remaining adsorption capacity is predicted through the following steps: At each preset interval, the total amount of impurities entering the adsorption cylinder per unit time is calculated in real time based on the inlet gas flow data, and the impurity concentration change trend is calculated based on the outlet composition data. When the outlet impurity concentration reaches a preset percentage of the inlet impurity concentration, the adsorption capacity is determined to be close to saturation. At this point, the remaining adsorption capacity of the adsorption cylinder is predicted by the impurity penetration curve model, thereby reducing the computational resource consumption.

4. The gas purification real-time monitoring method according to claim 3, characterized in that: The impurity penetration curve model is constructed through the following steps: The adsorption isotherm of the adsorbent for the target impurity was obtained through experimental measurement. An adsorption rate equation was established based on the adsorbent characteristics and gas flow parameters. Based on the adsorption isotherm and adsorption rate equation, an impurity breakthrough curve model is constructed, and the model parameters are dynamically corrected according to real-time monitoring data to improve prediction accuracy.

5. The method of claim 1, wherein: The adsorption capacity calculation stage also includes the following steps: When the monitoring system receives abnormal information from the monitoring point, it automatically diagnoses the cause of the fault, determines whether it is a sensor malfunction or an adsorbent failure, and executes the corresponding fault repair process.

6. The method of claim 1, wherein: During regeneration, a gradient control strategy is employed for the power of the regeneration heater, including the following steps: Regeneration triggering steps: When the remaining adsorption capacity is lower than the first preset threshold, a regeneration suggestion is generated; when the remaining adsorption capacity is lower than the second preset threshold, which is lower than the first preset threshold, the regeneration process is automatically executed. Regeneration preheating step: After the regeneration process is executed, the regeneration heater is started with preheating power to preheat the regeneration gas flow, while the outlet gas temperature of the adsorption cylinder is monitored; when the outlet gas temperature stabilizes and rises to the set value, the preheating is considered complete. Variable power control steps: After preheating, the composition analyzer installed on the regeneration exhaust pipe monitors the change in the concentration of desorbed impurities; when a sharp increase in the concentration of desorbed impurities is detected, the regeneration heater is controlled to operate at the first power level to provide the energy required for desorption; when the concentration of desorbed impurities is detected to reach its peak and begin to decrease, the power of the regeneration heater is switched to a second power level lower than the first power level to maintain the necessary desorption temperature. Post-regeneration optimization and termination steps: Continuously monitor the concentration of desorbed impurities. When it drops to the preset concentration, further reduce the power of the regeneration heater to the third power level and perform heat preservation and purging. Calculate the total amount of desorbed impurities per unit time in real time. When this value is lower than the preset critical value, the regeneration is determined to be complete, and the regeneration heater is turned off.

7. The method of real-time monitoring of gas purification according to claim 6, characterized by: After regeneration is complete, the following steps are also included: Regeneration effect evaluation and model self-learning steps: After regeneration is completed, record the total energy consumption, effective regeneration time and initial adsorption performance of the regenerated adsorption cartridge, and compare it with historical data. Based on the comparison results, adaptively correct the impurity breakthrough curve model parameters and regeneration control parameters for the next cycle. If the initial adsorption performance of the regenerated adsorption cartridge is lower than expected, the monitoring system will automatically generate an alarm, indicating possible adsorbent failure or equipment malfunction, and automatically adjust the regeneration control strategy.

8. The method of claim 6, wherein: The first and second preset thresholds are set based on the design adsorption capacity of the adsorption cartridge, the actual impurity load of the current purified gas, and the desired adsorption cartridge utilization safety factor. The specific values ​​of the first, second, and third power levels are calculated and set in real time based on the specifications of the adsorption cartridge, the type of adsorbent, and the regeneration gas flow rate through an energy consumption optimization algorithm. The preset critical value is dynamically set based on the total amount of impurities adsorbed by the adsorption cartridge in the previous adsorption cycle.

9. The method of claim 7, wherein: The regeneration effect evaluation and model self-learning steps also include the following steps: Adsorption cartridge performance recovery evaluation steps: After the regeneration cycle is completed and the adsorption cartridge is put into the next purification cycle, analyze the rate of change of impurity concentration in the outlet gas at the beginning of the purification cycle, compare it with the performance data of the benchmark adsorption cartridge, and calculate the adsorbent performance recovery coefficient. Data correlation and analysis steps: Correlate the efficiency data of this regeneration with the adsorption cylinder state data before regeneration. The adsorption cylinder state data includes the total amount of impurities adsorbed in the previous purification cycle and the predicted remaining adsorption capacity before regeneration. Model parameter adaptive correction steps: Based on the correlation analysis results, the impurity breakthrough curve model parameters and regeneration control parameters are dynamically corrected; if the energy consumption of this regeneration is higher than the historical average level or the preset energy consumption threshold, but the adsorbent performance recovery coefficient is higher than the first preset coefficient threshold, it is determined that there is room for optimization in the regeneration process. The concentration threshold or temperature set point of heating power switching in the regeneration control parameters is automatically fine-tuned to seek a better energy efficiency ratio strategy and use it for the next regeneration; if the adsorbent performance recovery coefficient after this regeneration is lower than the second preset coefficient threshold, it is determined that the adsorbent has irreversible deactivation or that there was over-adsorption in the previous purification cycle. Based on this, a deactivation warning is issued or the remaining adsorption capacity threshold for triggering regeneration in the next cycle is adjusted accordingly. Knowledge base update steps: The pre-regeneration state, regeneration process control parameters, and post-regeneration performance evaluation results of this regeneration cycle are stored as a data sample in the historical knowledge base for use in optimizing decisions for subsequent regeneration cycles.

10. The method of claim 9, wherein: The adsorbent performance recovery coefficient is calculated as follows: Under the same inlet impurity concentration and flow rate conditions, the ratio of the time required for the current adsorption cylinder to reach the same outlet purity standard or the total amount of impurities processed at the beginning of the purification cycle to the reference adsorption cylinder is the adsorbent performance recovery coefficient.

11. The method of claim 9, wherein: In the aforementioned model parameter adaptive correction step, if the monitoring system detects a continuous downward trend in the adsorbent performance recovery coefficient for a preset number of consecutive times, it will automatically trigger an advanced diagnostic alarm, indicating that the adsorbent is deactivated and needs to be replaced or requires manual diagnosis.

12. A system for real-time monitoring of gas purification, characterized by: A method for implementing real-time monitoring of gas purification as described in any one of claims 1-11, comprising: The data acquisition module is used to collect inlet gas flow data and outlet composition data through a flow sensor installed at the inlet of the adsorption cylinder and a gas chromatograph installed at the outlet of the adsorption cylinder. The adsorbent state monitoring module is used to monitor the state of the adsorbent in real time through multiple monitoring points set inside the adsorption cylinder. The monitoring points are equipped with temperature sensors or component monitoring probes. The adsorption capacity calculation module is used to predict the remaining adsorption capacity of the adsorption cylinder based on the inlet gas flow data and outlet composition data, using an impurity penetration curve model. It also performs cross-comparison and mutual verification based on the real-time monitored adsorbent state to determine whether there are any abnormalities in the remaining adsorption capacity and adsorbent state. Then, it corrects the remaining adsorption capacity and / or reports abnormalities at monitoring points based on the verification results. The regeneration time prediction module is used to dynamically predict the optimal regeneration time based on the composition analysis of the gas discharged from the adsorption cartridge during the purification process and the remaining adsorption capacity. The real-time control module transmits the remaining adsorption capacity and optimal regeneration time to the monitoring system, and adjusts the operating parameters of the regenerative purifier and regeneration heater in real time to ensure that they are always in the current optimal working state at different working stages.

13. A real-time monitoring device for gas purification, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that: When the processor executes the computer program, it implements the real-time monitoring method for gas purification as described in any one of claims 1-11.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which, when loaded and executed by a processor, implement the real-time gas purification monitoring method as described in any one of claims 1-11.

15. A computer program product comprising instructions, characterized in that: When the computer program product is run on a terminal, the terminal performs the gas purification real-time monitoring method as described in any one of claims 1-11.

Images