Air separation pressure swing intelligent adsorption pollution discharge method, system and device combined with pressure curve

Abstract

The invention discloses an air separation pressure swing intelligent adsorption pollution discharge method, system and device combined with a pressure curve, and relates to the technical field of intelligent pollution discharge. The air separation pressure swing intelligent adsorption pollution discharge method, system and device combined with the pressure curve comprises the following steps: S1, collecting running state data in real time, and preprocessing the running state data; s2, performing pressure disturbance intensity analysis on the operation state data; s3, performing real-time analysis and feature extraction on the pressure curve; s4, performing start-stop risk prediction on the operation state data; and S5, integrating the operation state data, the pressure disturbance intensity analysis result and the start-stop risk prediction result to execute a pollution discharge control instruction. The problems that an air compressor and a pressure swing adsorption system lack linkage control based on pressure change, pollution discharge easily causes pressure fluctuation and frequent starting and stopping, energy consumption is increased, equipment abrasion is caused, and an existing control mode lacks a feedforward adjusting mechanism are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent pollution discharge, in particular to an air separation variable pressure intelligent adsorption pollution discharge method, system and device combined with a pressure curve. BACKGROUND

[0002] With the wide application of pressure swing adsorption technology in the field of air separation oxygen production, the coordination of energy consumption control and pollution discharge rhythm in system operation has become an important issue to improve operation efficiency. The air compressor, as the main gas supply equipment, has a significant coupling relationship between its start-stop frequency, loading state and adsorption tower pollution behavior. The system pressure curve, as an important data basis reflecting the dynamic changes of operation, is gradually introduced into the pollution control and energy efficiency analysis process.

[0003] For example, the invention with the announcement number CN114437846B discloses an optimization method for natural gas pressure swing adsorption denitrification based on a computer, which includes that the computer control system respectively controls the opening degree of the valve in different cycles according to a plurality of preset flow rates in a plurality of cycles in which the valve does not work; the actual flow rate of the valve is detected in the plurality of cycles; the difference between the plurality of preset flow rates and the actual flow rates in a test number cycle formed by the plurality of cycles is calculated; the valve opening degree compensation value and the adjustment value in the next test number cycle are calculated according to the plurality of differences; and the computer control system controls the opening of the valve according to the valve opening degree adjustment value in the next test number cycle. The invention can detect the frequently switched valve system in time, and optimize the valve system according to the detection result, thereby ensuring the accurate control of the valve and improving the yield and product purity of the natural gas pressure swing adsorption denitrification process.

[0004] For example, the invention with the announcement number CN114437847B discloses a natural gas pressure swing adsorption denitrification process computer control method and system. The method includes a) determining the gas inlet time of the absorption tower; b) determining the adsorption and desorption time of the first half cycle; c) determining the switching time of the first half cycle; d) determining the adsorption and desorption time of the second half cycle; e) determining the switching time of the second half cycle; and f) running the monitoring feedback stage. The natural gas pressure swing adsorption denitrification process computer control method can scientifically set the switching time of the pressure increasing and decreasing valves of a group of adsorption towers, thereby improving the production efficiency and the utilization rate of raw gas.

[0005] Most existing air separation variable pressure systems still use fixed time pollution discharge and fixed threshold triggering, which cannot be dynamically adjusted according to real-time working conditions. The pollution discharge process is easy to cause system pressure fluctuation, leading to frequent start-stop of the air compressor, increased energy consumption and increased equipment wear. The existing control means lack comprehensive analysis of pollution disturbance, operation trend and abnormal fluctuation, and it is difficult to realize stable and efficient pollution management.

[0006] In view of the above problems, it is urgent to need an air separation variable pressure intelligent adsorption blowdown method, system and device combined with pressure curve. SUMMARY

[0007] Technical problems solved In view of the deficiencies of the prior art, the present application provides an air separation variable pressure intelligent adsorption blowdown method, system and device combined with pressure curve, which solves the problem that the air compressor and the pressure swing adsorption system lack linkage control based on pressure change, blowdown easily causes pressure fluctuation and frequent start-stop, leading to increased energy consumption and equipment wear, and the existing control mode lacks a feedforward regulation mechanism.

[0008] Technical scheme To achieve the above purpose, the present application is realized by the following technical scheme: an air separation variable pressure intelligent adsorption blowdown method combined with pressure curve, comprising the following steps: S1: collecting real-time operation state data and preprocessing the operation state data; S2: analyzing the pressure disturbance intensity of the operation state data, judging the disturbance level of the current blowdown action according to the pressure disturbance intensity analysis result, and selecting the corresponding blowdown control strategy; S3: through real-time analysis and feature extraction of the pressure curve, identifying the adsorption cycle, evaluating the disturbance intensity and monitoring the abnormal fluctuation; S4: predicting the start-stop risk of the operation state data, further judging whether the blowdown behavior causes the air compressor to start-stop according to the start-stop risk prediction result, and dynamically adjusting the blowdown rhythm, opening degree and duration; S5: executing the blowdown control instruction comprehensively according to the operation state data, the pressure disturbance intensity analysis result and the start-stop risk prediction result, and feeding back the blowdown execution result and system response information to the state collection and energy efficiency evaluation, and performing parameter rolling optimization and control strategy adaptive update.

[0009] Further, the real-time acquisition of operating state data and preprocessing of operating state data are specifically as follows: the operating state data includes time variable, blowdown starting time, blowdown duration, blowdown instantaneous pressure, starting pressure, average pressure before blowdown, target pressure, baseline power, baseline operation time, optimized power, optimized operation time, baseline fluctuation standard deviation and optimized fluctuation standard deviation; the time variable is acquired in real time and is time-stamped synchronously with other operating state data to form a complete time sequence; the blowdown starting time is obtained through the trigger time stamp recorded automatically by the system when the blowdown control instruction is issued; the blowdown duration is obtained by recording the opening and closing time of the blowdown valve and calculating the time difference; the blowdown instantaneous pressure is obtained by synchronous real-time acquisition through the pressure sensor; the starting pressure is obtained by real-time acquisition through the pressure sensor at the blowdown starting time; the average pressure before blowdown is obtained by calculating the historical blowdown instantaneous pressure data collected by the pressure sensor before the blowdown starting time; the target pressure is obtained by checking the air compressor equipment manufacturing parameters and setting the calibrated target pressure according to the long-term operation experience of the system; the baseline power is obtained by extracting the power collection data in the historical operation period and calculating the average value; the baseline operation time is obtained by extracting the time record corresponding to the default blowdown strategy in the control system; the optimized power is obtained by collecting data in real time during the operation of the current control strategy through the power collection device and calculating the average value; the optimized operation time is obtained by recording the start and end time corresponding to the current control strategy; the baseline fluctuation standard deviation is obtained by calculating the sample standard deviation of the power historical sequence collected in the operation period corresponding to the default blowdown strategy; the optimized fluctuation standard deviation is obtained by calculating the sample standard deviation of the power data collected in the operation period of the current blowdown optimization strategy; the preprocessing steps include: removing outliers from the collected operating state data, identifying various data abnormal conditions including sensor value abnormality, sudden signal interference and continuous sampling interruption, and filling in missing values by interpolation according to the abnormal position, and marking the unrestorable part as missing; all variables are resampled according to the unified time step; continuous variables are normalized to the range of zero to one through maximum and minimum values; the variables for fluctuation and trend analysis are standardized; the feature values of each time period are extracted from the continuous data by setting a sliding window with fixed length and moving by steps.

[0010] Further, the pressure disturbance intensity analysis of the operation state data specifically comprises the following steps: obtaining time variable, blowdown starting time, blowdown duration, blowdown instantaneous pressure and average pressure before blowdown; by integrating the change rate of the blowdown instantaneous pressure within the blowdown duration, the integral time interval starts from the blowdown starting time and lasts until the blowdown starting time plus the blowdown duration. Within the interval, the derivative of the blowdown instantaneous pressure with respect to time is calculated, and the absolute value thereof is taken, representing the rate of pressure change during the blowdown process; the rate is multiplied by a correction factor, which is composed of the difference between the blowdown instantaneous pressure and the average pressure before blowdown divided by the average pressure before blowdown. The integral result represents the cumulative value of the disturbance during the entire blowdown process, and finally divided by the blowdown duration, the average disturbance intensity per unit time, i.e. the pressure disturbance intensity value, is obtained.

[0011] Further, the judgment of the disturbance level of the current blowdown action according to the pressure disturbance intensity analysis result and the selection of the corresponding blowdown control strategy specifically comprises the following steps: real-time comparison of the pressure disturbance intensity value and the disturbance intensity threshold, the disturbance intensity threshold including a first-level disturbance threshold and a second-level disturbance threshold; when the pressure disturbance intensity value is greater than or equal to the first-level disturbance threshold, the blowdown operation is executed with flow limitation, the blowdown valve opening is controlled within 30%, segmented small flow release is adopted, and the buffer time is extended, while the air supplement and buffer device is enabled, abnormal behavior is recorded and alarm is triggered, and when it occurs continuously, it is switched to a low-frequency blowdown mode and prompts manual review; when the pressure disturbance intensity value is greater than the second-level disturbance threshold and less than the first-level disturbance threshold, the blowdown is switched to a slow blowdown mode, the valve is intermittently opened with a maximum opening of no more than 60%, the discharge time is appropriately extended, and when multiple adsorption towers are blowdown, automatic peak shifting is executed, the controller records the disturbance level and feeds back to the adjustment module; when the pressure disturbance intensity value is less than or equal to the second-level disturbance threshold, the original rhythm and opening of the blowdown are maintained, and the control system does not intervene.

[0012] Further, the adsorption cycle recognition, disturbance intensity evaluation and abnormal fluctuation monitoring through real-time analysis and feature extraction of the pressure curve specifically comprise the following steps: obtaining the adsorption tower inlet and outlet pressure data and processing the pressure curve in real time, identifying the adsorption, desorption and equalization stages of the system, and providing cycle reference for the blowdown strategy; extracting the pressure change rate through sliding window; analyzing the amplitude, backfall and fluctuation characteristics of the curve before and after blowdown, extracting the deviation, slope and stability indicators, identifying the valve jam, back pressure abnormal loss of control signs; comparing the current cycle with the historical typical cycle to judge the operation deviation and assist in strategy adjustment; all analysis results are structured and output for use by the linkage adjustment and energy efficiency evaluation module to support strategy judgment and parameter optimization.

[0013] Further, the start-stop risk prediction of the operation state data is specifically as follows: obtain the blowdown starting time, starting pressure and target pressure; the calculation process of the start-stop risk prediction value includes two parts. The first part is the derivative of the starting pressure to time, that is, the instantaneous change rate of the system pressure at the blowdown starting time, multiplied by the rate weight coefficient. The second part is to calculate the difference between the starting pressure and the target pressure, and then divide the difference by the target pressure to obtain the deviation proportion of the starting pressure relative to the target pressure, and then square the proportion to represent the strength of the deviation, and then multiply by the deviation weight coefficient. Add the calculation results of the first two parts to obtain the start-stop risk prediction value.

[0014] Further, the specific steps of further judging whether the blowdown behavior triggers the start-stop of the air compressor according to the start-stop risk prediction result and dynamically adjusting the blowdown rhythm, opening degree and duration are as follows: real-time comparison of the start-stop risk prediction value and the start-stop risk threshold, the start-stop risk threshold includes a first risk threshold and a second risk threshold; when the start-stop risk prediction value is greater than or equal to the first risk threshold, the current blowdown is suspended, and after the pressure is recovered, the blowdown valve opening degree is limited to within 30%, and the air supplement and high pressure loading mode are enabled when necessary, and the multi-section low-flow discharge is adopted when it cannot be skipped and the interval buffer is set; trigger alarm and manual prompt, switch to low-frequency blowdown and protection mode when it appears continuously; when the start-stop risk prediction value is greater than the second risk threshold and less than the first risk threshold, the blowdown is adjusted to an intermittent slow blowdown mode, the valve opening degree is limited to not more than 60%, the discharge time is moderately extended, the automatic peak shifting is performed when multiple adsorption towers blowdown, the disturbance trajectory is recorded and dynamic adjustment is entered; when the start-stop risk prediction value is less than or equal to the second risk threshold, the original blowdown strategy remains unchanged, the valve is normally fully opened, the discharge time is kept default, and the control system does not intervene and trigger compensation and protection logic.

[0015] Further, the comprehensive operation state data, pressure disturbance intensity analysis results and start-stop risk prediction results execute the blowdown control instruction, and the blowdown execution results and system response information are fed back to the state acquisition and energy efficiency evaluation to perform parameter rolling optimization and control strategy adaptive update. The specific steps are as follows: obtaining the reference power, reference operation time, optimized power, optimized operation time, reference fluctuation standard deviation and optimized fluctuation standard deviation; the first part multiplies the reference power by the reference operation time, subtracts the optimized power multiplied by the optimized operation time to obtain the absolute magnitude difference value of the current energy saving benefit, and then divides the difference value by the product of the reference power and the reference operation time plus a small correction term to form a normalized ratio, which is used as the input of the hyperbolic tangent function; the second part subtracts the optimized fluctuation standard deviation from the reference fluctuation standard deviation to obtain the fluctuation improvement amount, divides the difference value by the reference fluctuation standard deviation plus a small correction term and takes the square root to obtain the fluctuation improvement factor; finally, the result of the hyperbolic tangent function part is multiplied by one plus the fluctuation improvement factor to calculate the energy saving and stability comprehensive value. The energy saving and stability comprehensive value is compared with the energy saving evaluation threshold in real time, and the energy saving evaluation threshold includes a primary energy saving threshold and a secondary energy saving threshold; when the energy saving and stability comprehensive value is greater than or equal to the primary energy saving threshold, the current parameter is marked as not recommended, the upper limit of the valve opening degree is automatically reduced, the blowdown time is shortened, and the period is appropriately extended, and if the continuous low efficiency is recovered, the previous round of high-score parameter is recovered, and the conservative mode can be switched to; when the energy saving and stability comprehensive value is greater than the secondary energy saving threshold and less than the primary energy saving threshold, the current parameter is temporarily retained, the valve opening degree adjustment step is reduced, the duration control is more precise, the trend is continuously tracked, and if the medium efficiency state lasts for two rounds, it is retained as a candidate scheme; when the energy saving and stability comprehensive value is less than or equal to the secondary energy saving threshold, the current parameter is marked as a high-optimization scheme, stored in the strategy library and preferentially called, continuously reused under normal conditions, and when the parallel control condition is met, the strategy is applied to other towers and partitions.

[0016] The second aspect of the present application provides a space variable pressure intelligent adsorption blowdown system combined with a pressure curve, comprising: a state acquisition module, a blowdown control module, a curve analysis module, a linkage adjustment module and an energy efficiency evaluation module, characterized in that: the state acquisition module is used for acquiring real-time operation state data and pre-processing the operation state data; the blowdown control module is used for analyzing the pressure disturbance intensity of the operation state data, judging the disturbance level of the current blowdown action according to the pressure disturbance intensity analysis result, and selecting a corresponding blowdown control strategy; the curve analysis module is used for identifying the adsorption cycle, evaluating the disturbance intensity and monitoring the abnormal fluctuation through real-time analysis and feature extraction of the pressure curve; the linkage adjustment module is used for predicting the start-stop risk of the operation state data, further judging whether the blowdown behavior causes the air compressor to start-stop according to the start-stop risk prediction result, and dynamically adjusting the blowdown rhythm, opening degree and duration; and the energy efficiency evaluation module is used for executing the blowdown control instruction by comprehensively considering the operation state data, the pressure disturbance intensity analysis result and the start-stop risk prediction result, and feeding back the blowdown execution result and system response information to the state acquisition and energy efficiency evaluation for parameter rolling optimization and control strategy adaptive update.

[0017] The third aspect of the present application provides an intelligent adsorption blowdown device combined with a pressure curve, comprising: a state acquisition and preprocessing unit, a disturbance intensity analysis unit, a curve feature identification unit, a start-stop risk prediction unit and a blowdown control and strategy optimization unit, characterized in that: the state acquisition and preprocessing unit is used for acquiring the data related to the adsorption tower, the air compressor and the blowdown, and completing the alignment, cleaning and standardization processing; the disturbance intensity analysis unit is used for calculating the pressure disturbance intensity during blowdown and determining the disturbance level to trigger the corresponding control strategy; the curve feature identification unit is used for analyzing the pressure curve, identifying the adsorption stage and extracting the key change indicators, and monitoring the abnormal fluctuation; the start-stop risk prediction unit is used for evaluating the influence of blowdown on the operation of the air compressor, generating a risk score and adjusting the blowdown parameters; and the blowdown control and strategy optimization unit is used for executing the blowdown control and feeding back the operation result, updating the control parameters and optimizing the strategy.

[0018] Advantages The present application has the following advantages: (1) The present application unifies the acquisition and preprocessing of operation state data, establishes a standardized time series data structure, solves the problem of disordered data sources and variable alignment difficulty in traditional systems, provides a unified and calculable input basis for subsequent analysis modules, and improves the accuracy and availability of control strategy response.

[0019] (2), the present application realizes quantitative division of disturbance level by designing pressure disturbance intensity analysis formula, and realizes corresponding control strategy based on threshold selection, solves the problem that the existing pollution control cannot identify disturbance intensity and the single response mode, and significantly enhances the dynamic adjustment ability and anti-disturbance performance of the system.

[0020] (3), the present application introduces the cycle identification and fluctuation monitoring mechanism based on pressure curve, extracts curve characteristic index through sliding window, identifies the current adsorption stage and abnormal fluctuation behavior, solves the problem that the traditional pollution control cannot identify the running cycle and cannot identify abnormal fluctuation in advance, enhances the coupling of pollution control and system state and the feedforward regulation ability.

[0021] (4), the present application realizes closed-loop adjustment and strategy optimization of control parameters by starting and stopping risk prediction and energy-saving stability evaluation index, linkage disturbance intensity, running trend and energy efficiency performance, solves the problem that the existing pollution control has no feedback mechanism and the parameters cannot be self-updated, and significantly improves the energy-saving level of the system and the self-adaptive ability of the operation strategy.

[0022] Of course, any product implementing the present application does not necessarily need to achieve all the advantages described above at the same time. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The present application is a process flow chart of intelligent adsorption pollution control of air separation variable pressure combined with pressure curve; Figure 2 The present application is a system structure diagram of intelligent adsorption pollution control of air separation variable pressure combined with pressure curve; Figure 3 The present application is an energy-saving stability comprehensive value distribution diagram; Figure 4 The present application is a pressure curve comparison diagram under typical pollution disturbance condition. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0025] Please refer to Figures 1-4The embodiment of the application provides a technical scheme: an air separation transformer intelligent adsorption blowdown method combined with a pressure curve, comprising the following steps: S1: collecting running state data in real time, and preprocessing the running state data; S2: performing pressure disturbance intensity analysis on the running state data, judging the disturbance level of the current blowdown action according to the pressure disturbance intensity analysis result, and selecting a corresponding blowdown control strategy; S3: performing adsorption cycle identification, disturbance intensity evaluation and abnormal fluctuation monitoring through real-time analysis and feature extraction of the pressure curve; S4: performing start-stop risk prediction on the running state data, further judging whether the blowdown behavior causes the air compressor to start and stop according to the start-stop risk prediction result, and dynamically adjusting the blowdown rhythm, opening degree and duration; S5: executing the blowdown control instruction in combination with the running state data, the pressure disturbance intensity analysis result and the start-stop risk prediction result, and feeding back the blowdown execution result and system response information to the state collection and energy efficiency evaluation to perform parameter rolling optimization and control strategy adaptive update.

[0026] Specifically, the running state data is collected in real time, and the running state data is preprocessed in the following specific steps: the running state data includes a time variable, a blowdown starting time, a blowdown duration, a blowdown instantaneous pressure, a starting pressure, an average pressure before blowdown, a target pressure, a reference power, a reference running time, an optimized power and an optimized running time. These data constitute the key input of system operation, and provide support for subsequent disturbance identification, risk assessment and energy efficiency determination. The time variable is time-stamped by a collection device to form a structured time series; the blowdown starting time is determined by the time stamp recorded by the system when the control instruction is triggered; the blowdown duration is calculated by the opening and closing time of the blowdown valve; the blowdown instantaneous pressure is collected by a pressure sensor in real time; the starting pressure is the pressure value measured by the system at the blowdown starting time; the average pressure before blowdown is calculated by backtracking the historical data of a short period before blowdown; the target pressure is set according to the air compressor parameters and experience; the reference power and the reference running time are calculated by the historical default strategy period; and the optimized power and the optimized running time are obtained by real-time sampling and statistical results during the execution of the current strategy. The preprocessing steps include: outlier rejection, interpolation and marker processing of missing data; all variables are uniformly resampled and aligned by time step; continuous variables are normalized to the [0, 1] interval, and trend variables are standardized; the average value, volatility and other statistical characteristics of each time period are extracted from the continuous data by setting a fixed length and sliding time window, to provide input support for subsequent disturbance analysis and strategy optimization.

[0027] In this embodiment, this step provides high-quality and highly comparable basic data support for subsequent disturbance intensity analysis, cycle identification, start-stop prediction, energy efficiency evaluation and other modules of the system by constructing a structured and standardized state data sequence. By real-time acquisition and preprocessing of key operating variables, the accuracy and responsiveness of data-driven control strategies are improved, which helps to achieve intelligent linkage adjustment and energy-saving optimization control based on pressure changes.

[0028] Specifically, the specific steps of pressure disturbance intensity analysis on operating state data are as follows: obtaining time variable, blowdown start time, blowdown duration, blowdown instantaneous pressure and average pressure before blowdown to constitute the minimum input set required for disturbance evaluation. These variables can accurately depict the dynamic influence of blowdown behavior on system pressure. By integrating the change rate of blowdown instantaneous pressure within the blowdown duration, the pressure fluctuation of the system during the blowdown process is quantified. The integral interval starts from the blowdown start time and continues to the blowdown start time plus the blowdown duration, covering the entire blowdown process. Within this time interval, the derivative of the blowdown instantaneous pressure with respect to time is calculated, and its absolute value is taken as the instantaneous disturbance rate of the system; the disturbance rate is multiplied by a correction factor composed of the ratio of the difference between the blowdown instantaneous pressure and the average pressure before blowdown to the average pressure, which is used to reflect the deviation degree of the disturbance amplitude from the steady-state pressure baseline. Finally, the corrected rate value is integrated within the time interval to obtain the disturbance cumulative amount during the entire blowdown period, and then divided by the blowdown duration to obtain the average disturbance intensity of the system per unit time, defined as the pressure disturbance intensity value. This value can be used to distinguish the blowdown disturbance level and trigger the basis for strategy linkage control.

[0029] The specific calculation method of the pressure disturbance intensity value is as follows:

[0030] In the formula, represents the pressure disturbance intensity value, represents the time variable, represents the blowdown start time, represents the blowdown duration, represents the blowdown instantaneous pressure, represents the average pressure before blowdown.

[0031] In this embodiment, this step quantifies the dynamic fluctuation amplitude and change rate of the system pressure during the blowdown process by constructing a pressure disturbance intensity value, effectively reflecting the actual impact of the blowdown behavior on the system stability. This value not only has the characteristics of high sensitivity and timely response, but also can be used as an important basis for disturbance level judgment and control strategy selection, providing reliable support for the linkage adjustment of blowdown rhythm, valve opening degree and other parameters, and helping to improve the controllability of blowdown behavior and the stability of system operation.

[0032] Specifically, according to the pressure disturbance intensity analysis result, the disturbance level of the current blowdown action is judged, and the corresponding blowdown control strategy is selected. The specific steps are as follows: real-time comparison of the pressure disturbance intensity value and the set disturbance intensity threshold, the disturbance intensity threshold includes a first disturbance threshold and a second disturbance threshold, which correspond to high, medium and low three disturbance level division standards respectively. This judgment mechanism can realize rapid response and hierarchical management of pressure fluctuation caused by blowdown. When the pressure disturbance intensity value is greater than or equal to the first disturbance threshold, it is identified as a high disturbance state, and the blowdown operation is executed in a flow limiting manner, with the maximum opening degree of the blowdown valve controlled within 30%, and a segmented small flow release mode is adopted to reduce the instantaneous pressure impact and prolong the exhaust buffer time, and the air supplement device and buffer module are coordinated to improve the system back pressure stability; abnormal blowdown behavior will be recorded and an alarm event will be triggered, and if high disturbance situation occurs continuously for many times, it will automatically switch to a low-frequency blowdown safety mode and prompt manual review intervention. When the pressure disturbance intensity value is between the first and second thresholds, it is determined as a medium disturbance state, and the blowdown strategy is adjusted to a slow blowdown mode, the blowdown valve is opened intermittently, and the maximum opening degree is not more than 60%, and the emission duration is appropriately lengthened to reduce the disturbance rate; if multiple adsorption towers are to be blowdown, automatic peak shifting is performed to avoid disturbance superposition, and the disturbance level information is fed back to the adjustment module. If the pressure disturbance intensity value is less than or equal to the second disturbance threshold, it is determined that the current blowdown behavior has less impact on the pressure, and the original blowdown rhythm and valve opening degree remain unchanged, and the control system does not trigger any intervention measures, maintaining an efficient and stable emission state.

[0033] In this embodiment, this step realizes dynamic identification and response adjustment of the system pressure fluctuation caused by blowdown behavior by establishing a disturbance intensity hierarchical control mechanism. According to the disturbance intensity value, different blowdown control strategies are matched to ensure timely flow limiting and buffering in high disturbance state, optimized rhythm and peak shifting emission in medium disturbance state, and maintaining the original strategy unchanged in low disturbance state. This mechanism significantly improves the flexibility and accuracy of blowdown control, effectively suppresses the system pressure instability problem caused by blowdown, and provides strategy support for stable operation and energy saving control of the system.

[0034] Specifically, by real-time analysis and feature extraction of the pressure curve, the adsorption cycle recognition, disturbance intensity evaluation and abnormal fluctuation monitoring are carried out, and the specific steps are as follows: obtaining the adsorption tower inlet and outlet pressure data, and performing real-time processing on the pressure curve to identify the current running stage of the system, including adsorption, desorption and pressure equalization process, to provide cycle position information support for subsequent pollution control strategy. By setting a fixed length sliding window to extract the pressure change rate, the local change trend is captured to form the key input of disturbance analysis. Further analyze the change amplitude, falling slope and fluctuation intensity of the pressure curve before and after the blowdown, extract the deviation value, curve slope and pressure stable interval and other feature indexes, to judge whether the system has abnormal fluctuation behavior, such as valve jam, back pressure abnormality and other potential fault signs. Then compare and analyze the pressure curve of the current cycle with multiple historical typical cycles to identify the degree of deviation of the running state, and assist the system strategy adaptive correction. All analysis results will be output in the form of structured data, and uniformly transmitted to the linkage adjustment module and energy efficiency evaluation module to support the dynamic determination of control strategy and the rolling optimization of parameters.

[0035] As shown in Figure 4 The pressure curve comparison chart under the typical blowdown disturbance situation provided by the present embodiment is shown in the figure. The high disturbance curve rises rapidly to a maximum pressure of about 7.59 at t=15, and then falls to a minimum of about 6.38 at t=48, with a large overall fluctuation amplitude, showing a severe disturbance characteristic. The medium disturbance curve reaches a peak of about 7.40 at t=15, and then gradually falls to about 6.60 at t=48, with a relatively mild fluctuation degree. In contrast, the low disturbance curve maintains a relatively stable pressure level, with a maximum value of about 7.22 and a minimum value of about 6.81, showing high system stability. The differences in the trends of the three curves near the t=20 blowdown start time reveal the dynamic influence of different disturbance levels on the system operation. The multiple inflection points and turning positions in the figure directly show the pressure fluctuation response caused by the blowdown behavior, providing a basis for disturbance intensity identification and strategy classification.

[0036] In the present embodiment, this step accurately identifies the current adsorption cycle stage of the system, evaluates the disturbance intensity caused by the blowdown behavior, and timely detects potential abnormal fluctuations, effectively improving the timeliness and accuracy of the control strategy. By converting the curve trend features into structured data results, the linkage support of blowdown rhythm, energy-saving control and fault warning is realized, providing key basis for system intelligent adjustment and stable operation.

[0037] Specifically, the running state data is subjected to start-stop risk prediction, and the specific steps are as follows: three key parameters of blowdown starting time, starting pressure and target pressure are obtained to construct a start-stop risk score. The calculation process of the start-stop risk prediction value contains two core parts for evaluating the influence degree of blowdown behavior on the running stability of the air compressor. The first part reflects the intensity of the current pressure change trend by calculating the derivative of the system pressure to time at the blowdown starting time, i.e. the instantaneous change rate, and multiplying a set rate weight coefficient to quantify its contribution to the system start-stop risk. The rate weight coefficient is obtained by statistical analysis of the correlation between the first-order differential change amplitude of the pressure sensor output signal and the start-stop frequency of the air compressor, and the value range is generally between 0.3 and 0.7. The second part obtains the pressure deviation proportion by calculating the difference between the starting pressure and the target pressure and dividing the difference by the target pressure, and then squares the proportion to enhance the sensitivity of the deviation intensity, and then multiplies a deviation weight coefficient to form an influence factor of the system pressure deviation state. The deviation weight coefficient is obtained by statistical analysis of the correlation between the deviation amplitude between the starting pressure before blowdown and the target pressure and the abnormal start-stop record of the air compressor, and the value range is generally between 0.5 and 0.9. Finally, the calculation results of the above two parts are added to obtain a comprehensive start-stop risk prediction value, which provides a decision basis for whether the system executes blowdown, whether the valve control is adjusted and other subsequent strategies.

[0038] The specific calculation method of the start-stop risk prediction value is as follows:

[0039] In the formula, represents the start-stop risk prediction value, represents the rate weight coefficient, represents the deviation weight coefficient, represents the starting pressure, represents the target pressure.

[0040] In the embodiment, the step comprehensively evaluates the pressure change trend at the blowdown starting time and the deviation degree of the system target pressure by constructing a start-stop risk prediction value, quantifies the influence level of the blowdown behavior on the start-stop risk of the air compressor. The value as an important basis for linkage adjustment can realize the early prediction of potential start-stop impact, support the system to dynamically adjust the blowdown rhythm, valve opening degree and execution logic when facing instability risk, and improve the synergy of running stability and energy consumption control.

[0041] Specifically, according to the start-stop risk prediction result, it is further judged whether the blowdown behavior causes the air compressor to start and stop, and the blowdown rhythm, opening degree and duration are dynamically adjusted. The specific steps are as follows: comparing the real-time start-stop risk prediction value with the set start-stop risk threshold value, the threshold value is divided into two levels of first-level risk threshold value and second-level risk threshold value, which is used to realize hierarchical response control. When the start-stop risk prediction value is greater than or equal to the first-level risk threshold value, it is identified as a high-risk state, and the current blowdown operation needs to be suspended immediately, and then executed after the system pressure is restored to stable, the blowdown valve opening degree is limited to within 30%, and the blowdown valve is closed if necessary to avoid system fluctuation, and the air supply device is started and the air compressor is kept in a high-pressure loading state. If the blowdown cannot be skipped, it is changed to intermittent small flow discharge in multiple sections and a buffer time is set. The behavior is marked as a high-risk event, and an alarm prompt is automatically triggered, and the low-frequency blowdown protection mode is switched to in the case of continuous triggering. If the start-stop risk prediction value is between the first-level and second-level threshold values, it is determined as a medium-risk state, the blowdown mode is adjusted to intermittent slow blowdown, the maximum valve opening degree is limited to not more than 60%, and the discharge time is moderately extended to reduce the disturbance intensity. If multiple adsorption towers blowdown, staggered peak processing is automatically performed to avoid disturbance superposition, and the disturbance trajectory is recorded and enters the dynamic adjustment state, and the pressure recovery condition is continuously observed. If the start-stop risk prediction value is less than or equal to the second-level risk threshold value, it is determined as a low-risk state, the blowdown rhythm and control logic remain unchanged, the valve is normally fully opened, the discharge time is not adjusted, the control system does not trigger intervention and does not enable the protection logic, and the blowdown is efficiently and stably performed.

[0042] In the embodiment, by comparing the start-stop risk prediction value with the preset threshold value in real time, the dynamic identification and hierarchical response of the blowdown behavior that may cause the air compressor to start and stop are realized. According to the risk level, the blowdown rhythm, valve opening degree and discharge time are automatically adjusted. In the high-risk state, the flow is actively suspended and limited, in the medium-risk state, the rhythm is optimized and staggered execution is performed, and in the low-risk state, the original strategy remains unchanged. The mechanism effectively avoids the frequent start and stop of the air compressor caused by the blowdown operation, and improves the stability of the system operation and the safety of the blowdown control.

[0043] In particular, the integrated operation state data, pressure disturbance intensity analysis results and start-stop risk prediction results are used to execute the blowdown control instruction, and the blowdown execution results and system response information are fed back to the state acquisition and energy efficiency evaluation module for parameter rolling optimization and control strategy adaptive update, and the specific steps are as follows: first, the key performance indicators of the reference power, reference operation time, optimized power, optimized operation time, reference fluctuation standard deviation and optimized fluctuation standard deviation are obtained, which are used to quantify the energy saving effect and operation stability of the control strategy. The first part multiplies the reference power and the reference operation time, subtracts the product of the optimized power and the optimized operation time, obtains the energy saving benefit of the current strategy compared with the default strategy, and then divides the product of the reference power and the operation time by a small correction term to construct a normalized index as the input of the hyperbolic tangent function, which is used to control the nonlinear gain. The small correction term is used to prevent the denominator from being zero or close to zero, which causes calculation abnormality, and guarantees the stability of the formula, which is set as a constant, such as Even smaller, adjusted according to the numerical range of the system. The second part subtracts the optimized fluctuation standard deviation from the reference fluctuation standard deviation to obtain the system fluctuation improvement amount, which is further normalized and square rooted to form a fluctuation improvement factor. The two are combined to calculate an energy saving and stability comprehensive value, which is used to measure the comprehensive performance of the current strategy in energy saving and stability. The real-time value is compared with the energy saving evaluation threshold, and the energy saving evaluation threshold includes a first-level and a second-level energy saving threshold. When the energy saving and stability comprehensive value is greater than or equal to the first threshold, it is determined that the system is in a low efficiency state, and the current parameter set is marked as not recommended. The upper limit of the blowdown valve opening is automatically tightened, the blowdown time is shortened, and the blowdown period is appropriately extended. If the low efficiency state persists, the previous parameter set with better performance is restored, and the conservative strategy mode can be entered. When the energy saving and stability comprehensive value is between the second threshold and the first threshold, it is determined that the system is in a medium efficiency state, and the current parameter set is temporarily retained. At the same time, the valve opening adjustment step is reduced, the time control logic is refined, and the subsequent score changes are continuously tracked. If the medium efficiency state persists for two cycles, the parameter set is retained as a candidate solution. When the energy saving and stability comprehensive value is less than or equal to the second energy saving threshold, it is determined that the current strategy has good energy saving effect and operation stability, and the parameter set is marked as a high priority solution and stored in the strategy library. It is set as the priority call configuration and can be reused continuously under the premise of stable system operation. When there are multiple tower control capabilities, the strategy can also be applied to other towers and partitions to improve overall operation efficiency.

[0044] The specific calculation method of the energy saving and stability comprehensive value is as follows:

[0045] In the formula, E represents the energy saving and stability comprehensive value, P represents the reference power, T represents the reference operation time, P represents the optimized power, T represents the optimized operation time, Indicates a minimal correction term. Indicates the standard deviation of the benchmark fluctuation. This represents the optimized standard deviation of fluctuation.

[0046] Table 1 shows the comprehensive energy-saving and stable value data provided in the embodiments of this application. In this embodiment, the baseline power of Scheme 1 is set to 100, the baseline operating time is set to 2.0, the optimized power is set to 88, the optimized operating time is set to 2.0, the baseline fluctuation standard deviation is set to 2.2, and the optimized fluctuation standard deviation is set to 3.5; the baseline power of Scheme 2 is set to 100, the baseline operating time is set to 2.0, the optimized power is set to 84, the optimized operating time is set to 2.0, the baseline fluctuation standard deviation is set to 3.1, and the optimized fluctuation standard deviation is set to 3.8; the baseline power of Scheme 3 .... The baseline power is set to 2.0, the optimized power is set to 82, the optimized running time is set to 2.0, the baseline fluctuation standard deviation is set to 2.4, and the optimized fluctuation standard deviation is set to 3.9. For Scheme 4, the baseline power is set to 100, the baseline running time is set to 2.0, the optimized power is set to 77, the optimized running time is set to 2.0, the baseline fluctuation standard deviation is set to 3.5, and the optimized fluctuation standard deviation is set to 4.0. For Scheme 5, the baseline power is set to 100, the baseline running time is set to 2.0, the optimized power is set to 70, the optimized running time is set to 2.0, the baseline fluctuation standard deviation is set to 3.9, and the optimized fluctuation standard deviation is set to 4.2.

[0047] Table 1. Comprehensive Energy Saving and Stability Value Data Table

[0048] like Figure 3 The figure shows the distribution of the comprehensive energy-saving stability value provided in the embodiments of this application. According to the data in the image and table, the set primary energy-saving threshold is 0.4, and the secondary energy-saving threshold is 0.2. The comprehensive energy-saving stability value fluctuates significantly among the five schemes, ranging from 0.3442 to 0.5645. Scheme 3 has the highest comprehensive value at 0.5645, exceeding the primary energy-saving threshold, indicating that this scheme has low system energy efficiency and high volatility under the pollution control strategy. Scheme 1's Ec value also exceeds the primary threshold at 0.4913, similarly falling within the low-efficiency range. Schemes 2, 4, and 5 have Ec values ​​of 0.3442, 0.3550, and 0.3714, respectively, falling between the primary and secondary thresholds, reflecting moderate energy-saving and stability performance. This figure visually demonstrates the differences in energy efficiency among different control schemes during system operation, providing a quantitative basis for the dynamic optimization and selection of pollution control strategies.

[0049] In this embodiment, the step is to evaluate the current pollution control strategy in terms of energy consumption and system fluctuation by calculating the energy-saving stability comprehensive value, and to realize dynamic evaluation and adaptive update of the strategy based on hierarchical threshold. According to the evaluation results, the high, medium and low efficient strategy states are automatically identified, and then the pollution parameters are adjusted, the control logic is optimized, and the high quality strategy is rolled over and reused across towers. This mechanism significantly enhances the closed-loop optimization capability of the control strategy, and improves the overall energy-saving level and operation stability of the system.

[0050] As shown in Figure 2 The structure schematic diagram of the air separation variable pressure intelligent adsorption pollution control system combined with the pressure curve provided by the embodiment is shown in the figure. The air separation variable pressure intelligent adsorption pollution control system combined with the pressure curve provided by the embodiment applies the air separation variable pressure intelligent adsorption pollution control method combined with the pressure curve, which includes a state acquisition module, a pollution control module, a curve analysis module, a linkage adjustment module and an energy efficiency evaluation module. The state acquisition module is used to acquire multi-dimensional state data in real time during system operation, including pressure, valve state, pollution timing, and complete data cleaning, alignment and normalization preprocessing operations to ensure the accuracy and timeliness of subsequent analysis; the pollution control module is used to analyze the pressure disturbance intensity based on the operation data, construct the disturbance intensity value and compare it with the set threshold, determine the pollution level and match the corresponding control strategy, and realize hierarchical response adjustment; the curve analysis module is used to analyze the pressure curve trend in real time, identify that the system is in the adsorption, desorption and pressure equalization stages, extract the pressure change rate, fluctuation characteristics and other key indicators, monitor abnormal behavior and assist in strategy adjustment; the linkage adjustment module is used to predict the start-stop risk, evaluate the potential impact of pollution action on the start-stop stability of the air compressor, automatically adjust the pollution rhythm, valve opening degree and execution mode according to the risk level, and prevent system fluctuation amplification; the energy efficiency evaluation module is used to integrate system energy consumption data and operation fluctuation, calculate the energy-saving stability comprehensive value, determine the strategy execution effect, and feed back the results to the acquisition and adjustment module to support parameter rolling optimization and strategy adaptive update, forming a complete closed-loop control logic.

[0051] In this embodiment, the system structure realizes the whole-process closed-loop control from data acquisition, disturbance identification, risk prediction to strategy evaluation and optimization through modular design with clear division of labor. The modules work cooperatively to ensure that the pollution behavior realizes energy-saving goal while ensuring system stability. The state acquisition and preprocessing module provides high-quality basic data, the curve analysis and linkage adjustment module realizes dynamic identification and accurate response, the pollution control module realizes hierarchical strategy matching, and the energy efficiency evaluation module is responsible for continuous optimization and intelligent update, which improves the intelligent level and control precision of system operation as a whole.

[0052] The intelligent adsorption pollution control system provided by the embodiment of the application is combined with the pressure curve, and the intelligent adsorption pollution control method combined with the pressure curve comprises a state acquisition and preprocessing unit, a disturbance intensity analysis unit, a curve feature recognition unit, a start-stop risk prediction unit and a pollution control and strategy optimization unit. The state acquisition and preprocessing unit is used to acquire multiple types of operation data from the inlet and outlet of the adsorption tower, the outlet of the air compressor and the pollution discharge execution path, and perform time synchronization, abnormal value elimination and normalization conversion, so as to provide input information with a unified structure for subsequent analysis; the disturbance intensity analysis unit calculates the disturbance intensity value in real time according to the pressure change rate and amplitude fluctuation in the pollution discharge process, and divides the disturbance level by comparing with a threshold level, to trigger the corresponding control logic; the curve feature recognition unit performs real-time analysis based on the continuous pressure curve, identifies the current adsorption cycle stage, extracts key indicators such as the slope, fluctuation, deviation and other key indicators, and detects abnormal fluctuation behavior in a timely manner; the start-stop risk prediction unit outputs the start-stop risk score by jointly modeling the pressure trend and target pressure deviation at the start of pollution discharge, and dynamically adjusts the pollution discharge rhythm, valve opening degree and execution mode according to the score result; the pollution control and strategy optimization unit executes the corresponding control strategy after receiving the analysis unit result, records the system response in real time, and feeds back the execution result to the parameter optimization mechanism, to realize continuous adjustment of the control parameters and rolling update of the strategy library, and to build a data-driven adaptive pollution control process.

[0053] In the embodiment, the structure divides the functions into five core units, and builds a logical and efficient intelligent pollution control process. The units form a closed loop among data acquisition, disturbance identification, pressure analysis, risk prediction and control execution, to ensure that the system can realize high-precision data-driven decision-making during operation. Through joint determination of the disturbance intensity and the start-stop risk, the system can not only cope with different pollution disturbances, but also continuously optimize the control parameters, to realize dynamic adjustment of the pollution behavior and adaptive achievement of the energy saving goal.

[0054] It should be noted that, in this document, the relationship terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device.

[0055] The preferred embodiments of the application disclosed above are only to facilitate the elucidation of the application. The preferred embodiments do not describe all the details of the application and limit the application to the specific embodiments described. Obviously, many modifications and variations can be made in light of the teachings above. The description is chosen and described in order to best explain the principles of the application and its practical application to thereby enable others skilled in the art to best utilize the application and get the best results from the application. The application is only limited by the claims and their full scope and equivalents.

Claims

1. A smart adsorption-based wastewater discharge method using pressure curves in conjunction with air separation pressure swing analysis, characterized in that: The method comprises the following steps: S1: real-time acquisition of running state data and preprocessing of the running state data; S2: pressure disturbance intensity analysis of the running state data, judgment of the disturbance level of the current pollution action according to the pressure disturbance intensity analysis result, and selection of the corresponding pollution control strategy; S3: adsorption cycle identification, disturbance intensity evaluation and abnormal fluctuation monitoring through real-time analysis and feature extraction of the pressure curve; S4: start-stop risk prediction of the running state data, further judgment of whether the pollution behavior causes the air compressor to start-stop according to the start-stop risk prediction result, dynamic adjustment of the pollution rhythm, opening degree and duration; S5: execution of the pollution control instruction in combination with the running state data, the pressure disturbance intensity analysis result and the start-stop risk prediction result, and feedback of the pollution execution result and system response information to the state acquisition and energy efficiency evaluation for parameter rolling optimization and adaptive updating of the control strategy.

2. The pressure curve integrated air separation pressure swing intelligent adsorption purging method according to claim 1, characterized in that: The real-time acquisition of the running state data and the preprocessing of the running state data specifically comprises the following steps: The running state data comprises time variable, pollution starting time, pollution duration, pollution instantaneous pressure, starting pressure, average pressure before pollution, target pressure, baseline power, baseline running time, optimized power, optimized running time, baseline fluctuation standard deviation and optimized fluctuation standard deviation; The time variable is acquired in real time and is time-stamped synchronously with other running state data to form a complete time sequence; the pollution starting time is obtained through the trigger time stamp recorded by the system automatically when the pollution control instruction is issued; the pollution duration is obtained by recording the opening and closing times of the pollution valve and calculating the time difference; the pollution instantaneous pressure is obtained by synchronous real-time acquisition by the pressure sensor; the starting pressure is obtained by real-time acquisition by the pressure sensor at the pollution starting time; the average pressure before pollution is obtained by calculating the historical pollution instantaneous pressure data continuously acquired by the pressure sensor before the pollution starting time; the target pressure is obtained by checking the air compressor equipment manufacturing parameters and setting the calibrated target pressure according to the long-term running experience of the system; the baseline power is obtained by extracting the power acquisition data in the historical running period and calculating the average value; the baseline running time is obtained by extracting the time record corresponding to the default pollution strategy in the period; the optimized power is obtained by real-time acquisition of the data by the power acquisition device during the running of the current control strategy and calculation of the average value; the optimized running time is obtained by counting the start and end time records corresponding to the current control strategy; the baseline fluctuation standard deviation is obtained by calculating the sample standard deviation of the power historical sequence acquired in the running period corresponding to the default pollution strategy; and the optimized fluctuation standard deviation is obtained by calculating the sample standard deviation of the power data acquired in the running period of the current pollution optimization strategy. The preprocessing step includes: removing outliers from the collected operating state data, identifying various data abnormal conditions including sensor value abnormalities, sudden signal interference and continuous sampling interruption, and filling in missing values by interpolation according to the abnormal position, and marking the missing part as missing; all variables are resampled according to a unified time step; continuous variables are normalized to the range of zero to one through maximum and minimum values; variables for fluctuation and trend analysis are standardized; by setting a sliding window with a fixed length and moving in steps, feature values of each time period are extracted from continuous data.

3. The intelligent pressure swing adsorption method of claim 1, wherein: The specific steps of the pressure disturbance intensity analysis on the operating state data are as follows: Obtain the time variable, the blowdown start time, the blowdown duration, the blowdown instantaneous pressure and the average pressure before blowdown; By integrating the change rate of the blowdown instantaneous pressure within the blowdown duration, the time interval of the integral starts from the blowdown start time and lasts until the blowdown start time plus the blowdown duration. Within this interval, the derivative of the blowdown instantaneous pressure with respect to time is calculated, and its absolute value is taken, which represents the rate of pressure change during the blowdown process; this rate is multiplied by a correction factor, which is composed of the difference between the blowdown instantaneous pressure and the average pressure before blowdown divided by the average pressure before blowdown. The integral result represents the cumulative value of the disturbance during the entire blowdown process, and finally divided by the blowdown duration, the average disturbance intensity per unit time is obtained, which is the pressure disturbance intensity value.

4. The pressure curve integrated air separation pressure swing intelligent adsorption purging method according to claim 1, characterized in that: The specific steps of judging the disturbance level of the current blowdown action according to the pressure disturbance intensity analysis result and selecting the corresponding blowdown control strategy are as follows: Real-time comparison of the pressure disturbance intensity value and the disturbance intensity threshold, the disturbance intensity threshold includes a first disturbance threshold and a second disturbance threshold; When the pressure disturbance intensity value is greater than or equal to the first disturbance threshold, the blowdown operation is limited, the blowdown valve opening is controlled within 30%, a segmented small flow release is adopted, and the buffer time is extended, at the same time, the air supplement and buffer device are enabled, the abnormal behavior is recorded and the alarm is triggered, and when it appears continuously, it is switched to the low-frequency blowdown mode and prompts manual review; When the pressure disturbance intensity value is greater than the second disturbance threshold and less than the first disturbance threshold, the blowdown is switched to the slow blowdown mode, the valve is intermittently opened and the maximum opening is not more than 60%, the emission time is appropriately extended, the automatic peak shifting is executed when multiple adsorption towers blowdown, and the controller records the disturbance level and feeds back to the adjustment module; When the pressure disturbance intensity value is less than or equal to the second disturbance threshold, the original rhythm and opening of the blowdown are maintained, and the control system does not intervene.

5. The pressure curve integrated air separation pressure swing intelligent adsorption purging method according to claim 1, characterized in that: The specific steps of identifying the adsorption cycle, evaluating the disturbance intensity and monitoring the abnormal fluctuation by real-time analysis and feature extraction of the pressure curve are as follows: The pressure data at the inlet and outlet of the adsorption tower are acquired and the pressure curve is processed in real time to identify the adsorption, desorption and equalization stages of the system, to provide a cycle reference for the blowdown strategy; the pressure change rate is extracted through a sliding window; the amplitude, backfall and fluctuation characteristics of the curve before and after blowdown are analyzed to extract the deviation, slope and stability indicators, to identify the valve sticking, back pressure abnormal loss of control signs; the current cycle is compared with the historical typical cycle to judge the operation deviation and assist the strategy adjustment; all analysis results are structured and output for the use of the linkage adjustment and energy efficiency evaluation module to support the strategy judgment and parameter optimization.

6. The pressure curve integrated air separation pressure swing intelligent adsorptive purging method according to claim 1, characterized in that: The specific steps of the start-stop risk prediction based on the operation state data are as follows: The blowdown starting time, starting pressure and target pressure are acquired; The calculation process of the start-stop risk prediction value includes two parts. The first part is the derivative of the starting pressure with respect to time, that is, the instantaneous change rate of the system pressure at the blowdown starting moment, multiplied by the rate weight coefficient. The second part is to calculate the difference between the starting pressure and the target pressure, and then divide the difference by the target pressure to obtain the deviation proportion of the starting pressure with respect to the target pressure, and then square the proportion to represent the strength of the deviation, and then multiply by the deviation weight coefficient. The calculation results of the first two parts are added to obtain the start-stop risk prediction value.

7. The pressure curve integrated air separation pressure swing intelligent adsorptive purge process of claim 1, wherein: The specific steps of further judging whether the blowdown behavior causes the air compressor to start-stop according to the start-stop risk prediction result, and dynamically adjusting the blowdown rhythm, opening degree and duration are as follows: The start-stop risk prediction value is compared with the start-stop risk threshold in real time, and the start-stop risk threshold includes a first risk threshold and a second risk threshold; When the start-stop risk prediction value is greater than or equal to the first risk threshold, the current blowdown is suspended, and after the pressure is restored, the blowdown valve opening degree is limited to within 30%, and the air supplement and high pressure loading mode are enabled when necessary, and the multi-segment low flow discharge is adopted when it cannot be skipped and the interval buffer is set; an alarm and manual prompting are triggered, and when it occurs continuously, it is switched to a low frequency blowdown and protection mode; When the start-stop risk prediction value is greater than the second risk threshold and less than the first risk threshold, the blowdown is adjusted to an intermittent slow blowdown mode, the valve opening degree is limited to not more than 60%, the discharge time is moderately extended, the blowdown of multiple adsorption towers is automatically staggered, the disturbance trajectory is recorded and dynamic adjustment is entered; When the start-stop risk prediction value is less than or equal to the second risk threshold, the original blowdown strategy remains unchanged, the valve is normally fully opened, the discharge duration is maintained by default, and the control system does not intervene and trigger the compensation and protection logic.

8. The pressure curve integrated air separation pressure swing intelligent adsorptive purge process of claim 1, wherein: The specific steps of executing the blowdown control instruction based on the comprehensive operation state data, pressure disturbance intensity analysis result and start-stop risk prediction result, and feeding back the blowdown execution result and system response information to the state acquisition and energy efficiency evaluation for parameter rolling optimization and control strategy adaptive update are as follows: The reference power, reference operation time, optimized power, optimized operation time, reference fluctuation standard deviation and optimized fluctuation standard deviation are acquired; The first part multiplies the benchmark power by the benchmark running time, subtracts the optimized power multiplied by the optimized running time, to obtain an absolute magnitude difference of the current energy saving benefit, and then divides the difference by the product of the benchmark power and the benchmark running time plus a minimum correction term to form a normalized ratio, which is used as the input of the hyperbolic tangent function; The second part subtracts the optimized fluctuation standard deviation from the benchmark fluctuation standard deviation to obtain a fluctuation improvement amount, divides the difference by the benchmark fluctuation standard deviation plus a minimum correction term, and takes the square root to obtain a fluctuation improvement factor; Finally, the result of the hyperbolic tangent function part is multiplied by one plus the fluctuation improvement factor to calculate the energy saving stability comprehensive value.

9. Real-time comparison of the energy saving stability comprehensive value and the energy saving evaluation threshold, the energy saving evaluation threshold including a first-level energy saving threshold and a second-level energy saving threshold; When the energy saving stability comprehensive value is greater than or equal to the first-level energy saving threshold, the current parameter is marked as not recommended, the upper limit of the valve opening degree is automatically reduced, the blowdown time is shortened, and the cycle is appropriately extended. If it is continuously inefficient, the previous round of high-score parameter is restored, and it can be switched to a conservative mode; When the energy saving stability comprehensive value is greater than the second-level energy saving threshold and less than the first-level energy saving threshold, the current parameter is temporarily retained, the valve opening degree adjustment step is reduced, the duration control is more precise, the trend is continuously tracked, and if the medium-efficiency state lasts for two rounds, it is retained as a candidate solution; When the energy saving stability comprehensive value is less than or equal to the second-level energy saving threshold, the current parameter is marked as a high-optimization solution, stored in the strategy library and preferentially called, continuously reused under normal conditions, and synchronized to other towers and partitions when parallel control conditions are met.

10. The intelligent adsorption pollution control system with air separation and pressure change, applying the intelligent adsorption pollution control method with air separation and pressure change of any one of claims 1-8, comprising: The state acquisition module, the blowdown control module, the curve analysis module, the linkage adjustment module, and the energy efficiency evaluation module are characterized in that: The state acquisition module is configured to acquire running state data in real time and pre-process the running state data; The blowdown control module is configured to analyze the pressure disturbance intensity based on the running state data, determine the disturbance level of the current blowdown action based on the pressure disturbance intensity analysis result, and select a corresponding blowdown control strategy; The curve analysis module is configured to identify the adsorption cycle, evaluate the disturbance intensity, and monitor the abnormal fluctuation by analyzing and extracting the characteristics of the pressure curve in real time; The linkage adjustment module is configured to predict the start-stop risk based on the running state data, further determine whether the blowdown action triggers the start-stop of the air compressor based on the start-stop risk prediction result, and dynamically adjust the blowdown rhythm, opening degree, and duration; The energy efficiency evaluation module is configured to execute the blowdown control instruction based on the running state data, the pressure disturbance intensity analysis result, and the start-stop risk prediction result, and feed back the blowdown execution result and system response information to the state acquisition and energy efficiency evaluation to perform parameter rolling optimization and adaptive update of the control strategy. The air separation variable pressure intelligent adsorption blowdown device combined with the pressure curve is applied to the air separation variable pressure intelligent adsorption blowdown method combined with the pressure curve according to any one of claims 1-8, and includes a state acquisition and preprocessing unit, a disturbance intensity analysis unit, a curve characteristic identification unit, a start-stop risk prediction unit, and a blowdown control and strategy optimization unit, which are characterized in that: The state acquisition and preprocessing unit is used for acquiring adsorption tower, air compressor and pollution discharge related data, and completing alignment, cleaning and standardization processing; The disturbance intensity analysis unit is used for calculating the pressure disturbance intensity during pollution discharge, and determining the disturbance level to trigger the corresponding control strategy; The curve feature recognition unit is used for analyzing the pressure curve, identifying the adsorption stage and extracting the key change index, and monitoring abnormal fluctuations; The start-stop risk prediction unit is used for evaluating the influence of pollution discharge on the operation of the air compressor, generating a risk score and adjusting the pollution discharge parameters; The pollution discharge control and strategy optimization unit is used for executing pollution discharge control and feeding back the operation results, updating the control parameters and optimizing the strategy.

Images