Pressure relief bidirectional quick cut-off method and device for gas supply system of pressure swing adsorption device

Abstract

This invention discloses a method and apparatus for bidirectional rapid cutoff of pressure relief in a pressure swing adsorption (PSA) device's gas supply system, relating to the field of gas supply pressure relief control technology. The method and apparatus include the following steps: S1, acquiring gas supply pressure relief data and preprocessing the data; S2, establishing a sliding monitoring window, assessing expansion risk based on the preprocessed gas supply pressure relief data, determining whether to trigger a pressure relief control command based on the assessment results, and outputting an expansion risk dataset; S3, reading the expansion risk dataset, comprehensively considering the expansion risk assessment value, trigger flag, and pressure relief direction to determine whether to execute pressure relief, and controlling the micro-pressure relief action response; S4, assessing the degree of pressure relief anomaly based on deviations of multiple variables during the pressure relief execution process, identifying the anomaly type, and triggering corresponding safety response strategies to achieve closed-loop control. This solves the problem that residual gas in the dead cavity after valve shut-off is prone to expansion and overpressure, leading to a secondary explosion.

Description

Technical Field

[0001] This invention relates to the field of gas supply and depressurization control technology, specifically to a method and apparatus for bidirectional rapid cutoff of depressurization in the gas supply system of a pressure swing adsorption device. Background Technology

[0002] Pressure swing adsorption (PSA) units, as a core component in gas separation and purification processes, are widely used in industrial nitrogen production, oxygen production, and natural gas processing. The safety and response efficiency of their gas supply system have a critical impact on the stability of equipment operation. In actual operation, gas supply pipelines may experience sudden pressure differential changes due to leaks, blockages, or abnormal switching. If the system response to disconnection is delayed or the direction of identification is incorrect, abnormal gas diffusion can easily occur.

[0003] Traditional pressure relief and shut-off devices are mostly based on one-way valves, which cannot achieve differential pressure direction sensing and adaptive response. Moreover, their action structure relies on electromagnetic drive and external signals, lacking the ability to respond on its own in high-risk scenarios. At the same time, after the system is shut off, there are often dead cavities before and after the valve. If residual gas continues to accumulate under closed conditions, it is easy to cause expansion and overpressure, creating a risk of local explosion.

[0004] In addition, existing devices generally lack a real-time determination mechanism for pressure change trends, making it difficult to capture abnormal signs before expansion and lacking the ability to make judgments based on the dynamic evolution of pressure difference, which limits the system's response foresight and protection accuracy.

[0005] Therefore, in order to address the above problems, there is an urgent need for a bidirectional rapid cut-off method and device for the pressure swing adsorption (PSA) gas supply system. Summary of the Invention

[0006] Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a bidirectional rapid cut-off method and device for depressurization of the gas supply system of a pressure swing adsorption device, which solves the problem that residual gas in the dead cavity after valve shut-off is prone to expansion and overpressure, thus causing a secondary explosion.

[0008] Technical solution

[0009] To achieve the above objectives, the present invention provides the following technical solution: a bidirectional rapid cut-off method for pressure relief in the gas supply system of a pressure swing adsorption (PSA) device, comprising the following steps: S1, acquiring gas supply pressure relief data in real time through multi-point sensors and controlling system status feedback, and preprocessing the gas supply pressure relief data; S2, establishing a sliding monitoring window, assessing the expansion risk based on the preprocessed gas supply pressure relief data, determining whether to trigger a pressure relief control command based on the assessment results, and outputting an expansion risk dataset; S3, reading the expansion risk dataset, comprehensively determining whether to execute pressure relief based on the expansion risk assessment value, trigger flag, and pressure relief direction, and controlling the micro-pressure relief action response; S4, assessing the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process, identifying the anomaly type, and triggering corresponding safety response strategies to achieve closed-loop control.

[0010] Furthermore, the specific steps for acquiring gas supply pressure relief data through real-time acquisition and control system status feedback via multi-point sensors, and preprocessing the gas supply pressure relief data are as follows: Gas supply pressure relief data is acquired through real-time acquisition and control system status feedback via multi-point sensors. This data includes inlet dead space pressure, outlet dead space pressure, real-time differential pressure, inlet temperature, outlet temperature, valve shut-off status, inlet dead space volume, outlet dead space volume, valve operating resistance, and heat flux through the outer wall of the dead space. The gas supply pressure relief data is locally smoothed using a sliding window averaging algorithm. Outliers in the dynamic process variables of the gas supply pressure relief data are removed using a median filtering algorithm to stabilize the data distribution. Energy input variables in the gas supply pressure relief data are processed using a time-weighted accumulation method. Finally, the gas supply pressure relief data is standardized using a range normalization algorithm to unify the numerical scale and complete the normalization of the gas supply pressure relief data.

[0011] Furthermore, the specific steps for establishing a sliding monitoring window and assessing expansion risk based on preprocessed gas supply and pressure relief data are as follows: A sliding monitoring window is constructed, and preprocessed gas supply and pressure relief data are read at fixed sampling intervals and sequentially pushed into the sliding monitoring window in chronological order. Within the monitoring window, the pressure values ​​in the dead cavity before the valve at adjacent times are calculated using a first-order difference algorithm to obtain the rate of change of the dead cavity pressure before the valve. The rate of change of the dead cavity pressure before the valve is multiplied by the volume of the dead cavity before the valve and then divided by the square of the temperature before the valve to obtain the pressure increase term. The heat flux through the outer wall of the dead cavity is incremented by one, the natural logarithm is taken, and then divided by the temperature before the valve to obtain the heat flux growth term. The square of the valve action resistance is divided by the volume of the dead cavity before the valve to obtain the structural stagnation term. The pressure increase term, heat flux growth term, and structural stagnation term are added sequentially to obtain the expansion risk assessment value.

[0012] Furthermore, the specific steps for determining whether to trigger a pressure relief control command based on the assessment results and outputting the expansion risk dataset are as follows: The expansion risk assessment value and the expansion risk threshold are compared in real time. If the expansion risk assessment value is greater than or equal to the expansion risk threshold, the trigger flag is set to 1, and a control command is generated to drive the micro-pressure relief device to perform a pressure relief action. If the expansion risk assessment value is less than the expansion risk threshold, the trigger flag is set to 0, the control command remains empty, and the gas supply pressure relief data continues to be monitored. The gas supply pressure relief data, expansion risk assessment value, expansion risk threshold, and trigger flag are combined to form a data record, which is output as the expansion risk dataset. Simultaneously, the monitoring window is moved to the next time point to enter the next round of sampling and judgment process, achieving continuous dynamic updates.

[0013] Further, the specific steps for determining whether to perform pressure relief by reading the expansion risk dataset and combining the expansion risk assessment value, trigger flag, and pressure relief direction are as follows: Read the expansion risk dataset; when the trigger flag is 1 and the valve is in the off state, enter the pressure relief control logic; if any condition is not met, skip the pressure relief action for this cycle and return to the monitoring process; compare the dead chamber pressure before the valve with the dead chamber pressure after the valve; if the dead chamber pressure before the valve is greater than the dead chamber pressure after the valve, determine the pressure relief direction as the direction before the valve; if the dead chamber pressure after the valve is greater than the dead chamber pressure before the valve, determine the pressure relief direction as the direction after the valve; calculate the pressure relief execution judgment value based on the expansion risk dataset.

[0014] Furthermore, based on the expansion risk dataset, the specific steps for calculating the pressure relief execution judgment value are as follows: Subtract the expansion risk threshold from the expansion risk assessment value to obtain the risk response term; divide the absolute value of the real-time differential pressure by the valve action resistance plus one to obtain the differential pressure regulation term; calculate the exponential function value with the natural constant as the base, the negative of the product of the risk response term and the pressure relief regulation coefficient as the exponent, add one to the exponential function value as the denominator, and the valve shut-off state as the numerator, calculate the ratio of the numerator to the denominator to obtain the state control term; multiply the risk response term, differential pressure regulation term, and state control term sequentially to obtain the pressure relief execution judgment value.

[0015] Furthermore, the specific steps for controlling the micro-pressure relief action response are as follows: compare the pressure relief execution judgment value and the pressure relief execution threshold in real time. When the pressure relief execution judgment value is greater than or equal to the pressure relief execution threshold, send an opening signal to the corresponding pressure relief channel; otherwise, keep it closed and continue monitoring; continuously read the expansion risk assessment value and the real-time pressure difference during the pressure relief process. If both are lower than the expansion risk threshold and the minimum pressure difference threshold for three consecutive monitoring windows, it is determined that the pressure relief is completed and a pressure relief closing command is sent.

[0016] Furthermore, the specific steps for assessing the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process are as follows: Read the expansion risk assessment value, pressure relief execution judgment value, valve shut-off status, pressure relief execution threshold, and gas supply and pressure relief data within the current monitoring window to form an abnormal state assessment variable set; based on the abnormal state assessment variable set, quantify the degree of abnormality of the pressure relief action: divide the square of the difference between the expansion risk assessment value and the expansion risk threshold by the square of the inlet valve temperature to obtain the risk deviation term; calculate the square of the difference between the pressure relief execution threshold and the pressure relief execution judgment value as the numerator, calculate the absolute value of the real-time pressure difference plus 1 as the denominator, and divide the numerator by the denominator to obtain the response hysteresis term; divide the square of the valve action resistance by the inlet dead space volume to obtain the structural load term; invert the valve shut-off status to obtain the action anomaly term; add the risk deviation term, response hysteresis term, structural load term, and action anomaly term sequentially to obtain the pressure relief anomaly assessment value.

[0017] Furthermore, the specific steps for identifying anomaly types and triggering corresponding safety response strategies to achieve closed-loop control are as follows: Real-time comparison of the pressure relief anomaly assessment value and the pressure relief anomaly threshold. When the pressure relief anomaly assessment value is greater than the pressure relief anomaly threshold, the current pressure relief action is determined to be a high-level control anomaly, and the safety linkage processing flow is initiated: Based on the trigger flag constituting the pressure relief anomaly assessment value, the pressure relief execution judgment value, the expansion risk assessment value, the real-time differential pressure, the valve action resistance, and the valve shut-off status, their combination relationships are analyzed, and the anomaly types are classified as trigger non-response, pressure relief ineffective, and valve jamming, and the current time and judgment result are recorded; The corresponding linkage strategy is executed according to the anomaly type: When trigger non-response occurs, the control logic is paused and a risk alarm is issued; when pressure relief is ineffective, the backup pressure relief channel is switched; when the valve is jammed, a forced closure command is sent to the gas supply source; The anomaly type, pressure relief anomaly assessment value, response strategy, and time are combined into a complete anomaly record, written to the anomaly log table, and backed up synchronously; After the linkage response is completed, the control flow closes the current monitoring window and automatically enters the next monitoring window as the gas supply pressure relief data continues to advance, forming a continuous closed-loop operation mechanism.

[0018] The second aspect of this invention provides a bidirectional rapid cut-off device for pressure relief in a pressure swing adsorption (PSA) device's gas supply system, comprising: a gas supply and pressure relief data acquisition and preprocessing module, an expansion risk assessment and pressure relief triggering module, a pressure relief control execution logic processing module, and an anomaly recording and safety linkage response module. The gas supply and pressure relief data acquisition and preprocessing module is used to acquire and preprocess gas supply and pressure relief data in real time through multi-point sensors and control system status feedback. The expansion risk assessment and pressure relief triggering module is used to establish a sliding monitoring window, assess expansion risk based on the preprocessed gas supply and pressure relief data, determine whether to trigger a pressure relief control command based on the assessment results, and output an expansion risk dataset. The pressure relief control execution logic processing module is used to read the expansion risk dataset, comprehensively consider the expansion risk assessment value, trigger flag, and pressure relief direction to determine whether to execute pressure relief, and control the micro-pressure relief action response. The anomaly recording and safety linkage response module is used to assess the degree of pressure relief anomaly based on deviations of multiple variables during the pressure relief execution process, identify the anomaly type, and trigger corresponding safety response strategies to achieve closed-loop control.

[0019] Beneficial effects

[0020] The present invention has the following beneficial effects:

[0021] (1) The pressure swing adsorption device gas supply system pressure relief bidirectional rapid cut-off method and device can accurately identify the gas thermal expansion trend in the dead cavity by comprehensively considering multiple gas supply and pressure relief data such as the pressure change rate of the dead cavity in front of the valve, the volume of the dead cavity in front of the valve, the temperature in front of the valve and the heat flux of the outer wall of the dead cavity. It can detect the expansion risk in advance before the pressure difference is abnormal, and effectively avoid the dead cavity rupture or secondary explosion caused by local overpressure.

[0022] (2) The pressure swing adsorption device gas supply system pressure relief bidirectional rapid cut-off method and device, through dynamic analysis of combined expansion risk assessment value, real-time differential pressure, valve action resistance and valve cut-off status, can accurately determine the triggering conditions of pressure relief control action, improve the response stability under actual working conditions, and avoid pressure relief failure or frequent shaking caused by misjudgment or delayed judgment.

[0023] (3) The pressure swing adsorption device gas supply system pressure relief bidirectional rapid cut-off method and device, by continuously comparing the expansion risk assessment value, pressure relief execution judgment value, pressure relief execution threshold, real-time pressure difference, valve action resistance and valve cut-off status, forms a pressure relief abnormality assessment value, which can systematically identify control abnormalities such as trigger non-response, pressure relief ineffectiveness or valve jamming, and ensure the integrity and safety of the pressure relief process.

[0024] (4) The pressure swing adsorption device gas supply system pressure relief bidirectional rapid cut-off method and device, by constructing the linkage relationship between the pressure relief execution judgment value and the expansion risk assessment value, turns the pressure relief action from passive response to active judgment, establishes a collaborative triggering model between pressure difference drive, structural stagnation and cut-off state, and breaks through the limitation of poor adaptability of traditional single-factor control strategy to complex working conditions. Attached Figure Description

[0025] Figure 1 Flowchart of a bidirectional rapid cut-off method for depressurizing the gas supply system of a pressure swing adsorption (PSA) device;

[0026] Figure 2 Structural diagram of a bidirectional rapid shut-off device for depressurization of the gas supply system of a pressure swing adsorption unit;

[0027] Figure 3 This is a diagram illustrating the expansion risk assessment values ​​and trigger indicators.

[0028] Figure 4 The chart shows the fluctuation trend of the judgment value for pressure relief. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0030] Please see Figures 1-4 This invention provides a technical solution: a method and apparatus for bidirectional rapid cut-off of pressure relief in a pressure swing adsorption (PSA) device's gas supply system, comprising the following steps: S1, acquiring gas supply pressure relief data in real time through multi-point sensors and controlling system status feedback, and preprocessing the gas supply pressure relief data; S2, establishing a sliding monitoring window, assessing the expansion risk based on the preprocessed gas supply pressure relief data, determining whether to trigger a pressure relief control command based on the assessment result, and outputting an expansion risk dataset; S3, reading the expansion risk dataset, comprehensively determining whether to execute pressure relief based on the expansion risk assessment value, trigger flag, and pressure relief direction, and controlling the micro-pressure relief action response; S4, assessing the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process, identifying the anomaly type, and triggering corresponding safety response strategies to achieve closed-loop control.

[0031] Specifically, the following steps are taken to acquire gas supply pressure relief data and preprocess the data by real-time acquisition and control system status feedback through multi-point sensors: Gas supply pressure relief data is acquired through real-time acquisition and control system status feedback via multi-point sensors. This data includes the pre-valve dead chamber pressure, post-valve dead chamber pressure, real-time differential pressure, pre-valve temperature, post-valve temperature, valve shut-off status, pre-valve dead chamber volume, post-valve dead chamber volume, valve operating resistance, and heat flux through the outer wall of the dead chamber. Specifically, the pre-valve and post-valve dead chamber pressures are acquired by installing high-precision pressure sensors; the real-time differential pressure is obtained by measuring the pressures before and after the valve and calculating the difference; the pre-valve and post-valve temperatures are acquired by deploying thermistors; and the data is obtained by reading valve control signals. The system acquires the valve shut-off status using the controller status signal, with valve shut-off recorded as 1 and valve not shut-off recorded as 0. It obtains the pre-valve dead cavity volume and post-valve dead cavity volume through structural parameter settings and factory calibration. It acquires the valve's actuation resistance by monitoring the actuator torque sensor signal. It acquires the heat flux through the dead cavity outer wall using a heat flux sensor. It performs local smoothing on the gas supply and pressure relief data using a sliding window averaging algorithm. It removes outliers from the dynamic process variables in the gas supply and pressure relief data using a median filtering algorithm to stabilize the data distribution. It processes energy input variables in the gas supply and pressure relief data using a time-weighted accumulation method. Finally, it standardizes the gas supply and pressure relief data using a range normalization algorithm to unify the numerical scale and complete the normalization of the gas supply and pressure relief data.

[0032] In this implementation plan, by performing multi-point real-time acquisition and processing on gas supply and pressure relief data such as inlet dead space pressure, outlet dead space pressure, real-time differential pressure, inlet temperature, outlet temperature, valve shut-off status, inlet dead space volume, outlet dead space volume, valve actuation resistance, and heat flux through the outer wall of the dead space, the stability, continuity, and comparability of the gas supply and pressure relief data are improved. This provides a high-quality, low-noise, and dimensionless input basis for subsequent risk identification and control decisions.

[0033] Specifically, the steps for establishing a sliding monitoring window and assessing expansion risk based on preprocessed gas supply and pressure relief data are as follows: A sliding monitoring window is constructed, and preprocessed gas supply and pressure relief data are read at fixed sampling intervals and sequentially pushed into the sliding monitoring window in chronological order to continuously assess the local dynamic state of the dead space region; within the monitoring window, the dead space pressure values ​​before the valve at adjacent time points are calculated using a first-order difference algorithm to obtain the rate of change of the dead space pressure before the valve, which serves as a sensitive indicator reflecting the pressure change trend; the rate of change of the dead space pressure before the valve is multiplied by the volume of the dead space before the valve and then divided by the valve... The square of the initial temperature yields the pressure rise term, reflecting the rate of increase in volume-pressure coupling under unit thermal conditions. The natural logarithm of the heat flux through the outer wall of the dead cavity is added to, and then divided by the initial temperature, yields the heat flux growth term, reflecting the influence of structural boundary heat input on gas expansion. The square of the valve action resistance is divided by the volume of the dead cavity before the valve, yielding the structural hysteresis term, which characterizes the hysteresis factor formed by the structural motion characteristics on the pressure relief response. The pressure rise term, heat flux growth term, and structural hysteresis term are added sequentially to obtain the expansion risk assessment value, which serves as the key input for subsequent pressure relief trigger determination.

[0034] The specific formula for calculating the inflation risk assessment value is as follows:

[0035] ;

[0036] In the formula, R represents the expansion risk assessment value, D represents the rate of change of pressure in the dead cavity before the valve, V represents the volume of the dead cavity before the valve, T represents the temperature before the valve, H represents the heat flux of the outer wall of the dead cavity, and M represents the valve actuation resistance.

[0037] In this embodiment, Table 1 is a data table of expansion risk assessment values, listing the key physical parameters of the gas supply and depressurization data under five monitoring windows and the corresponding expansion risk assessment results. This is used to quantify the impact of the dead cavity expansion trend on the depressurization trigger judgment under different operating conditions. Specifically, in window 1, the dead cavity pressure change rate before the valve is 2500, the dead cavity volume before the valve is 0.0050, the temperature before the valve is 298, the heat flux through the dead cavity outer wall is 35, and the valve operating resistance is 0.7, resulting in a calculated expansion risk assessment value of 98.01. In window 2, the dead cavity pressure change rate before the valve is 3000, the dead cavity volume before the valve is 0.0045, the temperature before the valve is 303, the heat flux through the dead cavity outer wall is 40, and the valve operating resistance is 0.8, resulting in an expansion risk assessment value of 142.24. In window 3, the dead cavity pressure change rate before the valve is 2700, the dead cavity volume before the valve is 0.0048, and the temperature before the valve is... In window 1, the dead cavity outer wall heat flux is 38, the valve actuation resistance is 0.75, and the corresponding expansion risk assessment value is 117.20. In window 2, the dead cavity pressure change rate is 2600, the dead cavity volume is 0.0052, the valve temperature is 300, the dead cavity outer wall heat flux is 42, and the valve actuation resistance is 0.65, resulting in an expansion risk assessment value of 81.26. In window 3, the dead cavity pressure change rate is 2400, the dead cavity volume is 0.0049, the valve temperature is 295, the dead cavity outer wall heat flux is 36, the valve actuation resistance is 0.72, and the expansion risk assessment value is 105.81.

[0038] Table 1. Data Table of Inflation Risk Assessment Values

[0039]

[0040] like Figure 3 The diagram illustrates the expansion risk assessment value and trigger flag, visually demonstrating the changes in these values ​​across five consecutive monitoring windows. In window 1, the expansion risk assessment value is slightly below the expansion risk threshold, and the trigger flag is 0. In windows 2 and 3, the expansion risk assessment values ​​reach 142.24 and 117.20 respectively, significantly exceeding the expansion risk threshold, and the trigger flags are both 1. In window 4, the expansion risk assessment value is 81.26, below the expansion risk threshold, and the corresponding trigger flag is 0. In window 5, the assessment value rises again to 105.81, and the trigger flag is reset to 1. The overall trend reflects the accurate driving capability of the expansion risk assessment value in triggering the pressure relief control command, effectively demonstrating the dynamic response characteristics of the judgment mechanism.

[0041] In this implementation plan, by constructing a sliding monitoring window and building pressure rise, heat flux growth, and structural stagnation terms based on gas supply and pressure relief data such as the rate of change of dead cavity pressure before the valve, the volume of dead cavity before the valve, the temperature before the valve, the heat flux of the outer wall of the dead cavity, and the valve action resistance, the core change features reflecting the expansion trend can be extracted in real time, effectively forming an expansion risk assessment value, and providing a judgment basis with dynamic sensitivity and physical rationality for subsequent pressure relief execution.

[0042] Specifically, the steps for determining whether to trigger a pressure relief control command based on the assessment results and outputting an expansion risk dataset are as follows: Real-time comparison of the expansion risk assessment value and the expansion risk threshold; if the expansion risk assessment value is greater than or equal to the expansion risk threshold, it indicates that there is a gas expansion risk in the current dead space, the trigger flag is set to 1, and a control command is generated to drive the micro-pressure relief device to perform a pressure relief action. The control command serves as an execution signal for changes in the gas supply path status; if the expansion risk assessment value is less than the expansion risk threshold, it indicates that the current operating condition has not reached the pressure relief condition, the trigger flag is set to 0, the control command remains empty, and the gas supply pressure relief data at the current moment continues to be monitored; the gas supply pressure relief data, expansion risk assessment value, expansion risk threshold, and trigger flag of the current round are combined to form a complete data record, which is output as an expansion risk dataset. The expansion risk dataset is used to support subsequent pressure relief action execution and control anomaly judgment; simultaneously, the sliding monitoring window is moved to the next moment position to maintain the continuity of the time sequence, entering the next round of gas supply pressure relief data sampling and judgment process, realizing a dynamic update mechanism based on continuous sampling.

[0043] In this implementation plan, by comparing the expansion risk assessment value with the expansion risk threshold in real time, and combining the dynamic assignment of the trigger flag with the time sequence integration of gas supply and pressure relief data, a complete expansion risk dataset can be continuously output. This establishes a judgment mechanism that drives control commands based on expansion trends, ensuring that the pressure relief action is accurately triggered based on gas supply and pressure relief data such as the rate of change of dead chamber pressure before the valve, the volume of dead chamber before the valve, the temperature before the valve, the heat flux of the outer wall of the dead chamber, and the valve action resistance. This enables the orderly sliding of the monitoring window and the closed-loop advancement of the judgment process.

[0044] Specifically, the steps for determining whether to perform pressure relief by reading the expansion risk dataset and combining the expansion risk assessment value, trigger flag, and pressure relief direction are as follows: Read the expansion risk dataset, which consists of gas supply pressure relief data, expansion risk assessment value, expansion risk threshold, and trigger flag; when the trigger flag is 1 and the valve cut-off status is 1, enter the pressure relief control logic. The valve cut-off status is extracted from the gas supply pressure relief data, indicating that the micro-pressure relief device is in an executable state; if any condition is not met, skip the pressure relief action of this cycle and return to the monitoring process to continue receiving the next round of gas supply pressure relief data; compare the dead chamber pressure before the valve with... The pressure in the dead space after the valve is compared. Both the pressure in the dead space before and after the valve are real-time physical quantities in the gas supply and pressure relief data. If the pressure in the dead space before the valve is greater than the pressure in the dead space after the valve, the pressure relief direction is determined to be the direction before the valve. At this time, the pressure gradient is released from the gas supply source side to the outside. If the pressure in the dead space after the valve is greater than the pressure in the dead space before the valve, the pressure relief direction is determined to be the direction after the valve. At this time, the residual gas inside the structure is preferentially released to the downstream path. Based on the expansion risk dataset, combined with the expansion risk assessment value, real-time pressure difference, valve action resistance, valve cut-off status and pressure relief adjustment coefficient, the pressure relief execution judgment value is calculated for subsequent control command generation and pressure relief action decision.

[0045] In this implementation plan, by reading the expansion risk dataset and combining it with gas supply and pressure relief data such as trigger flags, valve shut-off status, pre-valve dead space pressure, post-valve dead space pressure, real-time differential pressure, valve action resistance, and pressure relief adjustment coefficient, the execution conditions and direction of the pressure relief action can be clearly defined. A linkage judgment mechanism based on expansion trend and pressure distribution can be established to ensure that the pressure relief control logic is accurately activated when the trigger requirements are met, and automatically reverts to the monitoring process when the conditions are not met, thereby improving the response consistency and execution accuracy of the control chain.

[0046] Specifically, based on the expansion risk dataset, the specific steps for calculating the pressure relief execution judgment value are as follows: Subtract the expansion risk threshold from the expansion risk assessment value to obtain the risk response term; divide the absolute value of the real-time differential pressure by the valve action resistance plus one to obtain the differential pressure adjustment term. Both the real-time differential pressure and the valve action resistance are process variables in the gas supply and pressure relief data, used to reflect the corrective effect of the current pressure relief resistance on the pressure release capability; calculate the exponential function value of the product of the risk response term and the pressure relief adjustment coefficient, using the natural constant as the base and the negative of the product as the exponential function value, adding one to the exponential function value as the denominator, and using the valve cut-off state as the numerator, and calculate the numerator and denominator... The mother ratio is used to obtain the state control term, which reflects the probability that the pressure relief action will be effectively responded to under the current control command. Among them, a nonlinear relationship model is established between the historical expansion risk assessment value and the actual pressure relief action response data through an exponential fitting algorithm. The slope parameter is extracted from the model to obtain the pressure relief adjustment coefficient, which has a value range of [0,1]. The risk response term, differential pressure adjustment term and state control term are multiplied in sequence to obtain the pressure relief execution judgment value. The pressure relief execution judgment value serves as the core quantitative basis for determining whether to issue a pressure relief command. It is used to compare with the pressure relief execution threshold in the future to guide the pressure relief device to perform the action.

[0047] The specific formula for calculating the pressure relief execution judgment value is as follows:

[0048] ;

[0049] In the formula, E represents the pressure relief execution judgment value, and R represents the expansion risk assessment value. Indicates the inflation risk threshold. The real-time differential pressure is represented by M, the valve operating resistance is represented by F, the valve is closed, and the pressure relief adjustment coefficient is represented by k.

[0050] In this embodiment, Table 2 is a data table of pressure relief execution judgment values, listing the expansion risk assessment values ​​and key control parameters under 45 monitoring windows. This data is used to analyze the calculation process of the pressure relief execution judgment values ​​and their corresponding control states under different operating conditions. Specifically, in monitoring window C01, the expansion risk assessment value is 103.2, the expansion risk threshold is 100, the real-time differential pressure is 12.8, the valve action resistance is 0.70, the valve shut-off state is 1, and the calculated pressure relief execution judgment value is 12.10. In monitoring window C02, the expansion risk assessment value is 104.5, the real-time differential pressure is 12.0, the valve action resistance is 0.75, the valve shut-off state is 1, and the pressure relief execution judgment value is 15.47. In C03, the expansion risk assessment value is 103. 9. Real-time differential pressure is 12.4, valve operating resistance is 0.75, valve shut-off status is 1, and pressure relief execution judgment value is 13.84; In C04, the expansion risk assessment value is 104.0, real-time differential pressure is 13.7, valve operating resistance is 0.78, valve shut-off status is 1, and pressure relief execution judgment value is 15.46; In C05, the expansion risk assessment value is 103.7, real-time differential pressure is 12.8, valve operating resistance is 0.70, valve shut-off status is 1, and pressure relief execution judgment value is 14.10.

[0051] Table 2 Data Table of Pressure Relief Execution Judgment Values

[0052]

[0053] like Figure 4 The graph shows the fluctuation trend of the pressure relief execution judgment value, illustrating its dynamic changes across multiple monitoring windows. The triggering state at each moment is visually labeled in conjunction with the pressure relief execution threshold. The vertical axis represents the pressure relief execution judgment value, the horizontal axis represents the monitoring window number, and the red dashed line represents the pressure relief execution threshold, used to distinguish between triggered and non-triggered states. In windows C02, C04, and C05, the pressure relief execution judgment value exceeds the pressure relief execution threshold, the trigger flag is 1, and the pressure relief control process begins. In windows C01 and C03, the pressure relief execution judgment value does not reach the pressure relief execution threshold, the trigger flag is 0, and the monitoring state is maintained. This graph verifies the judgment logic and response accuracy of the pressure relief control system under different monitoring cycles.

[0054] In this implementation plan, by jointly modeling gas supply and pressure relief data such as expansion risk assessment value, expansion risk threshold, real-time differential pressure, valve action resistance, pressure relief adjustment coefficient and valve shut-off status, a multi-factor relationship between risk response item, differential pressure adjustment item and state control item is constructed. This can quantify the trigger intensity of pressure relief action and form a pressure relief execution judgment value with dynamic response capability, which is used to accurately determine whether the actual conditions for performing pressure relief are met at the current moment.

[0055] Specifically, the steps for controlling the micro-pressure relief action response are as follows: The pressure relief execution judgment value and the pressure relief execution threshold are compared in real time. When the pressure relief execution judgment value is greater than or equal to the pressure relief execution threshold, an opening signal is sent to the corresponding pressure relief channel to drive the micro-pressure relief device to respond to the current operating condition; otherwise, it remains closed, and the gas supply pressure relief data continues to be monitored until the triggering condition is met before re-judging. During the pressure relief process, the expansion risk assessment value and real-time pressure difference are continuously read. The expansion risk assessment value reflects the expansion trend inside the dead cavity, and the real-time pressure difference reflects the degree of pressure release during the pressure relief process. If the expansion risk assessment value is lower than the expansion risk threshold for three consecutive monitoring windows, and the real-time pressure difference is lower than the minimum pressure difference threshold for three consecutive monitoring windows, it is determined that the pressure relief process has reached the target release state. A pressure relief closing command is sent to terminate this round of pressure relief action and return to the data monitoring process.

[0056] In this implementation plan, by continuously comparing the pressure relief execution judgment value with the pressure relief execution threshold, and combining the expansion risk assessment value with the continuous change trend of real-time differential pressure in multiple monitoring windows, it is possible to achieve precise triggering and automatic closing control of micro-pressure relief actions. This ensures that the pressure relief process terminates in a timely manner under the conditions of valve shut-off, sufficient differential pressure release, and relief of expansion risk, thereby improving the closed-loop reliability of pressure relief behavior and the condition matching of response rhythm.

[0057] Specifically, the steps for assessing the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process are as follows: Read the expansion risk assessment value, pressure relief execution judgment value, valve shut-off status, pressure relief execution threshold, and gas supply pressure relief data within the current monitoring window to form an abnormal state assessment variable set, which serves as the computational basis for anomaly identification; Based on the abnormal state assessment variable set, quantify the degree of anomaly in the pressure relief action: Divide the square of the difference between the expansion risk assessment value and the expansion risk threshold by the square of the inlet valve temperature to obtain the risk deviation term. The inlet valve temperature is derived from the gas supply pressure relief data and reflects the risk fluctuation amplitude under hot conditions; Calculate the square of the difference between the pressure relief execution threshold and the pressure relief execution judgment value as the numerator, and calculate the absolute value of the real-time pressure difference plus 1 as the denominator. The differential pressure, also derived from the gas supply and pressure relief data, describes the driving intensity of the pressure relief process. Dividing the numerator by the denominator yields the response lag term, which describes the delay between the pressure relief determination and the execution action. The square of the valve action resistance, divided by the dead space volume before the valve, serves as the structural load term, reflecting the degree of interference of mechanical resistance on the pressure relief behavior during valve opening. The valve shut-off state is inverted as the action anomaly term. The valve shut-off state is obtained in real time from the gas supply and pressure relief data, and a state value of 0 indicates that the pressure relief command has not been executed. The risk deviation term, response lag term, structural load term, and action anomaly term are added sequentially to obtain the pressure relief anomaly assessment value, which characterizes the overall deviation of the pressure relief action under the current monitoring window, providing a quantitative basis for the judgment of subsequent response strategies.

[0058] The specific formula for calculating the abnormal pressure relief assessment value is as follows:

[0059] ;

[0060] In the formula, S represents the abnormal pressure relief assessment value, and R represents the expansion risk assessment value. Here, T represents the expansion risk threshold, T represents the inlet temperature of the valve, and E represents the pressure relief execution judgment value. Indicates the pressure relief execution threshold. The real-time differential pressure is represented by M, the valve operating resistance is represented by V, the dead chamber volume before the valve is represented by F, and the valve is in the off state.

[0061] In this implementation plan, by extracting gas supply and pressure relief data such as expansion risk assessment value, pressure relief execution judgment value, pressure relief execution threshold, real-time differential pressure, valve action resistance, valve shut-off status, valve inlet temperature and valve inlet dead cavity volume, a deviation structure of risk deviation item, response lag item, structural load item and action anomaly item is constructed. This structure can comprehensively measure the degree of anomaly during the pressure relief execution process and form a pressure relief anomaly assessment value with identification capability, providing criterion support for subsequent determination of control anomaly type and activation of safety response process.

[0062] Specifically, the steps for identifying anomaly types and triggering corresponding safety response strategies to achieve closed-loop control are as follows: Real-time comparison of the pressure relief anomaly assessment value and the pressure relief anomaly threshold. When the pressure relief anomaly assessment value exceeds the pressure relief anomaly threshold, the current pressure relief action is determined to be a high-level control anomaly, and the safety linkage processing flow is initiated: Based on the trigger flag constituting the pressure relief anomaly assessment value, the pressure relief execution judgment value, the expansion risk assessment value, the real-time differential pressure, the valve action resistance, and the valve shut-off status, their combination relationships are analyzed, and the anomaly types are classified as trigger non-response, pressure relief ineffectiveness, and valve jamming, and the current time and judgment result are recorded; Corresponding linkage strategies are executed according to the anomaly type: When trigger non-response occurs, the control logic is paused and a risk alarm is issued; when pressure relief is ineffective, the backup pressure relief channel is switched; when the valve is jammed, a forced closure command is sent to the gas supply source; The anomaly type, pressure relief anomaly assessment value, response strategy, and time are combined into a complete anomaly record, written to the anomaly log table, and backed up synchronously; After the linkage response is completed, the control flow closes the current monitoring window and automatically enters the next monitoring window as the gas supply pressure relief data continues to advance, forming a continuous closed-loop operation mechanism.

[0063] In this implementation plan, by comparing the pressure relief anomaly assessment value with the pressure relief anomaly threshold in real time, and combining the gas supply and pressure relief data such as trigger flags, pressure relief execution judgment values, expansion risk assessment values, real-time differential pressure, valve action resistance, and valve shut-off status, the combination relationship between variables can be analyzed. This allows for the accurate identification of control anomaly types during the pressure relief process, which are dynamically classified as trigger non-response, pressure relief ineffectiveness, or valve jamming. The plan also matches and executes linkage strategies such as pause control, channel switching, or forced valve closure. After establishing a complete anomaly record, the monitoring window is automatically advanced, realizing a closed-loop safety control mechanism based on the continuous evolution of the pressure relief process.

[0064] like Figure 2 As shown, the second aspect of the present invention provides a bidirectional rapid cut-off device for pressure relief in the gas supply system of a pressure swing adsorption (PSA) device, comprising: a gas supply and pressure relief data acquisition and preprocessing module, an expansion risk assessment and pressure relief triggering module, a pressure relief control execution logic processing module, and an anomaly recording and safety linkage response module. The gas supply and pressure relief data acquisition and preprocessing module is used to acquire and preprocess gas supply and pressure relief data in real time through multi-point sensors and control system status feedback. The expansion risk assessment and pressure relief triggering module is used to establish a sliding monitoring window, assess the expansion risk based on the preprocessed gas supply and pressure relief data, determine whether to trigger a pressure relief control command based on the assessment result, and output an expansion risk dataset. The pressure relief control execution logic processing module is used to read the expansion risk dataset, comprehensively assess the expansion risk assessment value, trigger flag, and pressure relief direction to determine whether to execute pressure relief, and control the micro-pressure relief action response. The anomaly recording and safety linkage response module is used to assess the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process, identify the anomaly type, and trigger corresponding safety response strategies to achieve closed-loop control.

[0065] In this implementation plan, by collecting and preprocessing gas supply and pressure relief data such as inlet dead space pressure, outlet dead space pressure, real-time differential pressure, inlet temperature, outlet temperature, valve shut-off status, inlet dead space volume, outlet dead space volume, valve action resistance, and heat flux through the outer wall of the dead space, a sliding monitoring window is constructed to dynamically assess expansion risk, determine pressure relief trigger conditions, and generate an expansion risk dataset. The micro-pressure relief action response is controlled by combining the pressure relief direction and trigger status. The deviation between the expansion risk assessment value, pressure relief execution judgment value, pressure relief execution threshold, and abnormal pressure relief assessment value and the monitored variables is compared in real time. Abnormal types such as trigger failure, ineffective pressure relief, and valve jamming are classified and identified, and corresponding safety response strategies are executed to achieve closed-loop control of the entire process based on the continuous evolution of gas supply and pressure relief data.

[0066] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0067] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for rapid bidirectional pressure relief and shut-off of the gas supply system of a pressure swing adsorption (PSA) device, characterized in that, Includes the following steps: S1 acquires gas supply and pressure relief data in real time by collecting and controlling the system status feedback through multiple sensors, and preprocesses the gas supply and pressure relief data. It then standardizes the gas supply and pressure relief data by using a range normalization algorithm to unify the numerical scale and complete the normalization of the gas supply and pressure relief data. S2, establish a sliding monitoring window, assess the expansion risk based on the preprocessed gas supply and pressure relief data, determine whether to trigger the pressure relief control command based on the assessment results, and output the expansion risk dataset. Based on the expansion risk dataset, calculate the pressure relief execution judgment value. The specific steps for establishing a sliding monitoring window and assessing expansion risk based on preprocessed gas supply depressurization data are as follows: A sliding monitoring window is constructed, and the pre-processed gas supply and pressure relief data are read at fixed sampling intervals and pushed into the sliding monitoring window in chronological order. Within the monitoring window, the valve inlet dead space pressure value at adjacent time points is calculated using a first-order difference algorithm to obtain the valve inlet dead space pressure change rate. Calculate the expansion risk assessment value based on the pre-processed gas supply depressurization data: ； In the formula, R represents the expansion risk assessment value, D represents the rate of change of dead cavity pressure before the valve, V represents the volume of dead cavity before the valve, T represents the temperature before the valve, H represents the heat flux of the outer wall of the dead cavity, and M represents the valve action resistance. The specific steps for calculating the pressure relief execution decision value based on the expansion risk dataset are as follows: ； In the formula, E represents the pressure relief execution judgment value, and R represents the expansion risk assessment value. Indicates the inflation risk threshold. M represents the real-time differential pressure, F represents the valve operating resistance, and k represents the valve shut-off state. S3 reads the expansion risk dataset, comprehensively considers the expansion risk assessment value, trigger flag, and pressure relief direction to determine whether to perform pressure relief, and controls the micro pressure relief device to respond. It compares the pressure relief execution judgment value and the pressure relief execution threshold in real time. When the pressure relief execution judgment value is greater than or equal to the pressure relief execution threshold, it sends an opening signal to the corresponding pressure relief channel; otherwise, it remains closed and continues monitoring. S4 assesses the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process, identifies the anomaly type, and triggers corresponding safety response strategies to achieve closed-loop control.

2. The method for bidirectional rapid cut-off of pressure relief in the gas supply system of a pressure swing adsorption device according to claim 1, characterized in that: The specific steps for acquiring gas supply depressurization data in real time through multi-point sensors and controlling system status feedback, and for preprocessing the gas supply depressurization data, are as follows: By collecting and controlling the system status feedback in real time through multiple sensors, gas supply and pressure relief data are obtained. The gas supply and pressure relief data includes the dead cavity pressure before the valve, the dead cavity pressure after the valve, the real-time pressure difference, the temperature before the valve, the temperature after the valve, the valve shut-off status, the volume of the dead cavity before the valve, the volume of the dead cavity after the valve, the valve action resistance, and the heat flux of the outer wall of the dead cavity. The gas supply and pressure relief data are locally smoothed using a sliding window averaging algorithm; outliers in the dynamic process variables of the gas supply and pressure relief data are removed using a median filtering algorithm to stabilize the data distribution; and energy input variables in the gas supply and pressure relief data are processed using a time-weighted cumulative method.

3. The method for bidirectional rapid cut-off of pressure relief in the gas supply system of a pressure swing adsorption device according to claim 1, characterized in that: The specific steps for determining whether to trigger a pressure relief control command based on the evaluation results and outputting an expansion risk dataset are as follows: The system compares the expansion risk assessment value and the expansion risk threshold in real time. When the expansion risk assessment value is greater than or equal to the expansion risk threshold, a trigger flag will be set to 1, generating a control command to drive the micro pressure relief device to perform a pressure relief action. When the expansion risk assessment value is less than the expansion risk threshold, a trigger flag will be set to 0, the control command will remain empty, and the gas supply pressure relief data will continue to be monitored. The gas supply depressurization data, expansion risk assessment value, expansion risk threshold and trigger flag are combined into a data record and output as an expansion risk dataset; at the same time, the monitoring window is moved to the next time position to enter the next round of sampling and judgment process, so as to achieve continuous dynamic updates.

4. The method for bidirectional rapid cut-off of pressure relief in the gas supply system of a pressure swing adsorption device according to claim 1, characterized in that: The specific steps for reading the expansion risk dataset and determining whether to perform pressure relief based on the comprehensive expansion risk assessment value, trigger flag, and pressure relief direction are as follows: Read the expansion risk dataset. When the trigger flag is 1 and the valve is closed, enter the pressure relief control logic. If either condition is not met, skip the pressure relief action for this cycle and return to the monitoring process. Compare the dead chamber pressure before the valve with the dead chamber pressure after the valve. If the dead chamber pressure before the valve is greater than the dead chamber pressure after the valve, the direction of pressure relief is determined to be the direction before the valve; if the dead chamber pressure after the valve is greater than the dead chamber pressure before the valve, the direction of pressure relief is determined to be the direction after the valve.

5. The method for bidirectional rapid cut-off of pressure relief in the gas supply system of a pressure swing adsorption device according to claim 1, characterized in that: The specific steps for controlling the micro-decompression response are as follows: During the depressurization process, the expansion risk assessment value and real-time differential pressure are continuously read. If both are lower than the expansion risk threshold and the minimum differential pressure threshold for three consecutive monitoring windows, the depressurization is determined to be complete, and a depressurization shutdown command is sent.

6. The method for bidirectional rapid cut-off of pressure relief in the gas supply system of a pressure swing adsorption device according to claim 1, characterized in that: The specific steps for assessing the degree of abnormality in pressure relief based on the deviation of multiple variables during the pressure relief execution process are as follows: Read the expansion risk assessment value, pressure relief execution judgment value, valve shut-off status, pressure relief execution threshold and gas supply pressure relief data in the current monitoring window to form an abnormal state assessment variable set; Based on the abnormal state assessment variable set, the degree of abnormality of the pressure relief action is quantified: the square of the difference between the expansion risk assessment value and the expansion risk threshold is divided by the square of the inlet temperature of the valve to obtain the risk deviation term. The numerator is the square of the difference between the pressure relief execution threshold and the pressure relief execution judgment value. The denominator is the absolute value of the real-time pressure difference plus 1. The numerator is divided by the denominator to obtain the response hysteresis term. The structural load term is obtained by dividing the square of the valve action resistance by the dead cavity volume in front of the valve. The abnormal action term is obtained by reversing the valve shut-off state. The risk deviation term, response hysteresis term, structural load term and abnormal action term are added in sequence to obtain the pressure relief abnormality assessment value.

7. The method for bidirectional rapid cut-off of pressure relief in the gas supply system of a pressure swing adsorption device according to claim 1, characterized in that: The specific steps for identifying anomaly types and triggering corresponding safety response strategies to achieve closed-loop control are as follows: The system compares the pressure relief anomaly assessment value and the pressure relief anomaly threshold in real time. When the pressure relief anomaly assessment value exceeds the pressure relief anomaly threshold, the current pressure relief action is determined to be a high-level control anomaly, and the safety linkage processing flow is initiated. Based on the trigger flag constituting the pressure relief anomaly assessment value, the pressure relief execution judgment value, the expansion risk assessment value, the real-time differential pressure, the valve action resistance, and the valve shut-off status, the system analyzes their combination relationships, classifying the anomaly type as trigger non-response, pressure relief ineffective, and valve jamming, and records the current time and judgment result. The system executes corresponding linkage strategies according to the anomaly type: When trigger non-response occurs, the control logic is paused and a risk alarm is issued; when pressure relief is ineffective, the backup pressure relief channel is switched; when the valve is jammed, a forced closure command is sent to the gas supply source. Combine the anomaly type, pressure relief anomaly assessment value, response strategy, and time into a complete anomaly record, write it to the anomaly log table, and back it up synchronously. After the linkage response is completed, the control process closes the current monitoring window and automatically enters the next monitoring window as the gas supply and depressurization data continues to advance, forming a continuous closed-loop operation mechanism.

8. A bidirectional rapid pressure relief shut-off device for a pressure swing adsorption (PSA) device gas supply system, employing the bidirectional rapid pressure relief shut-off method for a PSA device gas supply system as described in any one of claims 1-7, characterized in that: include: The system includes a gas supply and depressurization data acquisition and preprocessing module, an expansion risk assessment and depressurization triggering module, a depressurization control execution logic processing module, and an anomaly recording and safety linkage response module, among which: The gas supply and pressure relief data acquisition and preprocessing module is used to acquire gas supply and pressure relief data in real time through multi-point sensors and control system status feedback, and to preprocess the gas supply and pressure relief data. The expansion risk assessment and pressure relief triggering module is used to establish a sliding monitoring window, assess the expansion risk based on the pre-processed gas supply pressure relief data, determine whether to trigger the pressure relief control command based on the assessment results, and output the expansion risk dataset. The pressure relief control execution logic processing module is used to read the expansion risk dataset, comprehensively consider the expansion risk assessment value, trigger flag and pressure relief direction to determine whether to perform pressure relief, and control the micro pressure relief action response. The anomaly recording and safety linkage response module is used to assess the degree of pressure relief anomaly based on the deviation of multiple variables during the pressure relief execution process, identify the anomaly type, and trigger the corresponding safety response strategy to achieve closed-loop control.

Images