An intelligent operating condition adjustment system for air separation equipment based on multi-point temperature and pressure monitoring

Abstract

An embodiment of the present invention provides an intelligent operating condition adjustment system for an air separation unit based on multi-point temperature and pressure monitoring, relating to the field of air separation technology and capable of realizing automated operating condition adjustment of a purification system. The system comprises: multiple internal temperature sensors; multiple internal pressure sensors; a regeneration data collection device that collects internal temperature and pressure data and performs preprocessing; a reconstruction device that reconstructs the three-dimensional temperature and pressure fields of the adsorber; a feature engineering device that extracts the three-dimensional temperature and pressure tomography matrices of the adsorber; a regeneration operating condition adjustment model that calculates the three-dimensional temperature and pressure tomography matrices and the current operating conditions and outputs operating condition adjustment instructions; a main controller that issues an operating condition adjustment signal; and an actuator that adjusts the adsorber operating condition based on the operating condition adjustment signal.

Description

Technical Field

[0001] The present invention relates to the technical field of air separation, and in particular to an intelligent operating condition adjustment system for air separation equipment based on multi-point temperature and pressure monitoring. Background Art

[0002] Air separation equipment, commonly referred to as air separation plants (ASUs), liquefies and distills air, ultimately separating it into oxygen, nitrogen, and other useful gases. These plants typically include a purification system to remove substances such as water, carbon dioxide, and hydrocarbons from the air output from the pre-cooling system. The purification system typically consists of two adsorbers: one in the purification phase and the other in the regeneration phase, which can include multiple regeneration conditions. Existing technology typically relies on manual experience to adjust operating conditions, which cannot guarantee the purification and regeneration performance of the purification system, thus affecting the final air separation quality. Summary of the Invention

[0003] In view of this, an embodiment of the present invention provides an intelligent operating condition adjustment system for an air separation plant based on multi-point temperature and pressure monitoring, which can realize automatic operating condition adjustment of the purification system.

[0004] To achieve the above objectives, the present invention provides an intelligent operating condition adjustment system for an air separation plant based on multi-point temperature and pressure monitoring.

[0005] The air separation equipment of the embodiment of the present invention includes: a purification system, the purification system includes: two adsorbers and an electric heater, each adsorber is in a purification stage or a regeneration stage; the intelligent operating condition adjustment system of the air separation equipment includes: a plurality of internal temperature sensors arranged inside the two adsorbers; a plurality of internal pressure sensors fixed inside the two adsorbers; a regeneration data collection device, which collects the internal temperature data detected in real time by the internal temperature sensors and the internal pressure data detected in real time by the internal pressure sensors in the adsorbers during the regeneration stage and performs preprocessing; a reconstruction device, which reconstructs the three-dimensional temperature field of the adsorber according to the internal temperature data after preprocessing, and reconstructs the three-dimensional temperature field of the adsorber according to the preprocessed internal temperature data. The three-dimensional pressure field of the adsorber is reconstructed based on the internal pressure data after the operation; a feature engineering device is used to extract the three-dimensional temperature tomography matrix of the adsorber from the three-dimensional temperature field, and to extract the three-dimensional pressure tomography matrix of the adsorber from the three-dimensional pressure field; a regeneration working condition adjustment model is a pre-trained machine learning model, which calculates the input three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix as well as the pre-recorded current working condition, and outputs an instruction to maintain the current working condition or an instruction to adjust the working condition; a main controller sends a working condition adjustment signal after receiving the working condition adjustment instruction; an actuator adjusts the working condition of the adsorber in the regeneration stage according to the received working condition adjustment signal.

[0006] Preferably, the intelligent operating condition adjustment system of the air separation equipment further includes: an inlet temperature sensor arranged at the adsorber inlet, an outlet temperature sensor and an outlet pressure sensor arranged at the adsorber outlet, a sensor data processing device communicating with the inlet temperature sensor, the outlet temperature sensor and the outlet pressure sensor, and a control confirmation device communicating with the sensor data processing device; after receiving the operating condition adjustment instruction, the main controller sends an operating condition adjustment confirmation request to the control confirmation device, and the operating condition adjustment confirmation request carries a current operating condition identifier; the control confirmation device receives the operating condition adjustment confirmation request, determines whether the sensor data transmitted by the inlet temperature sensor, the outlet temperature sensor and the outlet pressure sensor meet pre-configured necessary adjustment conditions for the current operating condition, and returns the judgment result to the main controller; the main controller receives the judgment result; if the judgment result indicates that the necessary adjustment conditions for the current operating condition are met, it sends an operating condition adjustment signal; if the judgment result indicates that the necessary adjustment conditions for the current operating condition are not met, it sends a manual intervention signal to the management unit.

[0007] Preferably, after the actuator adjusts the operating condition of the adsorber in the regeneration phase according to the operating condition adjustment signal, the purification system operates according to pre-configured default operating parameters of the adjusted operating condition; the default operating parameters include: the heating temperature of the electric heater and the blowing flow rate of the contaminated nitrogen gas from the cold box.

[0008] Preferably, after adjusting the operating condition of the adsorber in the regeneration phase according to the operating condition adjustment signal, the actuator returns an adjustment completion notification to the main controller, and the main controller returns an adjustment success notification to the regeneration operating condition adjustment model, and the regeneration operating condition adjustment model locally updates and records the current operating condition.

[0009] Preferably, the intelligent operating condition adjustment system of the air separation equipment further comprises: a working parameter regression model communicating with the regeneration operating condition adjustment model, wherein the working parameter regression model is a pre-trained machine learning regression model; the regeneration operating condition adjustment model calculates the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix as well as the current operating condition, and if the calculation result matches the abnormal operating mode pre-configured for the current operating condition, a working parameter calculation request is sent to the working parameter regression model; the working parameter calculation request carries the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix, the current operating condition, and the matched abnormal operating mode; the working parameter regression model outputs the target operating parameter for responding to the abnormal operating mode according to the received working parameter calculation request, and sends the working parameter update instruction carrying the target operating parameter to the main controller; the main controller receives the working parameter calculation request and sends the working parameter update instruction carrying the target operating parameter to the main controller; After receiving the working parameter update instruction, a parameter adjustment signal carrying the target working parameters is sent to the actuator; the target working parameters include: the heating temperature of the electric heater and the blowing flow rate of the contaminated nitrogen gas; the actuator operates the purification system according to the target working parameters in the parameter adjustment signal, and returns a change completion notification to the main controller after the operation is completed; after receiving the change completion notification, the main controller returns a change success notification to the working parameter regression model; after receiving the change success notification, the working parameter regression model sends a working parameter update notification carrying the target working parameters to the regeneration working condition adjustment model; the regeneration working condition adjustment model calculates the current three-dimensional temperature tomography matrix and three-dimensional pressure tomography matrix, the current working condition and the target working parameters to perform working condition adjustment judgment after the working parameters are updated.

[0010] Preferably, the intelligent operating condition adjustment system of the air separation equipment also includes: a plurality of humidity sensors fixed inside the adsorber; a plurality of adsorbent detection devices installed at different positions near the adsorber outlet; a purification data collection device, which collects the internal temperature data detected in real time by the internal temperature sensor in the adsorber during the purification stage, the internal humidity data detected in real time by the humidity sensor, and the adsorbent front position data collected by the adsorbent detection device and performs preprocessing; a feature extraction device, which splices the preprocessed internal temperature data, the internal humidity data and the adsorbent front position data into purification features; a purification decision model, which is a pre-trained machine learning model, calculates the input purification features and outputs a default working signal or an emergency switching signal; the emergency switching signal carries the remaining regeneration time of the adsorber in the regeneration stage, which is used to indicate the switching of the two adsorbers between the purification stage and the regeneration stage within the remaining regeneration time; a rearrangement device, which communicates with the purification decision model and the regeneration operating condition adjustment model according to the The regeneration condition adjustment model calculates the execution time of each to-be-executed condition of the adsorber in the regeneration phase and the target operating parameters of each to-be-executed condition in an emergency switching state based on the current three-dimensional temperature chromatography matrix, three-dimensional pressure chromatography matrix, and current operating conditions of the adsorber in the regeneration phase transmitted by the regeneration condition adjustment model, as well as the remaining regeneration time transmitted by the purification decision model. An emergency switching instruction carrying the execution time and the target operating parameters is generated and sent to the main controller. After receiving the emergency switching instruction, the main controller sends the execution time and the target operating parameters in the emergency switching instruction to the execution mechanism, so that the execution mechanism performs corresponding operations on the purification system. The main controller also sends an emergency switching notification carrying the execution time and the target operating parameters to the regeneration condition adjustment model. The regeneration condition adjustment model calculates the current three-dimensional temperature chromatography matrix, three-dimensional pressure chromatography matrix, current operating conditions, the execution time, and the target operating parameters of the adsorber in the regeneration phase to perform a condition adjustment judgment in the emergency switching state.

[0011] Preferably, for the adsorber in the purification stage, the control confirmation device obtains the sensor data of the purified air collected by the outlet temperature sensor and the outlet pressure sensor, and determines whether the sensor data meets the design indicators; if the design indicators are not met, an alarm signal is sent to the main controller; after receiving the alarm signal, the main controller sends a high-temperature regeneration instruction to the actuator, so that the actuator changes the purification system to a high-temperature regeneration mode; the main controller also sends a high-temperature regeneration notification to the regeneration condition adjustment model, so that the regeneration condition adjustment model executes the condition adjustment judgment under the high-temperature regeneration mode.

[0012] Preferably, the feature engineering device cuts the three-dimensional temperature field into multiple parallel sections, arranges the temperature value of each parallel section into a temperature vector, and combines the temperature vectors of multiple parallel sections into the three-dimensional temperature tomography matrix; the feature engineering device cuts the three-dimensional pressure field into multiple parallel sections, arranges the pressure value of each parallel section into a pressure vector, and combines the pressure vectors of multiple parallel sections into the three-dimensional pressure tomography matrix.

[0013] Preferably, the operating conditions of the adsorber in the regeneration stage are: pressure relief condition, heating condition, cold blowing condition, and pressurizing condition, or, preparation for pressure relief condition, pressure relief condition, preparation for heating condition, heating condition, cold blowing condition, preparation for pressurizing condition, pressurizing condition, preparation for switching condition, and switching condition.

[0014] Preferably, the necessary condition for adjusting the pressure relief condition is that the air pressure at the adsorber outlet is less than the first pressure threshold; the necessary condition for adjusting the heating condition is that the air temperature at the adsorber outlet is greater than the first temperature threshold; the necessary condition for adjusting the cold blowing condition is that the absolute value of the air temperature difference between the adsorber outlet and inlet is less than the preset second temperature threshold; the necessary condition for adjusting the charging condition is that the air pressure at the adsorber outlet is greater than the second pressure threshold.

[0015] According to the technical solution of the present invention, one embodiment of the above invention has the following advantages or beneficial effects: it is able to accurately perceive the working environment of the purification system and accurately calculate the timing of operating condition adjustment through sensor technology, automatic control technology and artificial intelligence technology for multi-point temperature and pressure monitoring, avoiding the defects of relying on manual experience to adjust the operating conditions, and helping to improve the quality and efficiency of air separation.

[0016] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are provided for a better understanding of the present invention and are not intended to limit the present invention.

[0018] Figure 1 It is a schematic diagram of the main components of the air separation plant;

[0019] Figure 2 It is a structural schematic diagram of the intelligent working condition adjustment system of an air separation plant based on multi-point temperature and pressure monitoring of the present invention;

[0020] Figure 3 This is a first schematic diagram of the working condition flow of the present invention;

[0021] Figure 4 This is a second schematic diagram of the working condition flow of the present invention. DETAILED DESCRIPTION

[0022] The following description of exemplary embodiments of the present invention is made in conjunction with the accompanying drawings, in which various details of the embodiments of the present invention are included to facilitate understanding. These details should be considered as merely exemplary. Therefore, it should be appreciated by those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0023] Figure 1 This is a schematic diagram of the main components of the air separation plant, such as Figure 1 As shown, the air separation equipment may include air filters, air compression systems, pre-cooling systems, purification systems, heat exchange systems, turbine expanders, distillation systems and other parts. The functions of these parts are known technologies and will not be introduced here.

[0024] Figure 2 This is a structural diagram of the intelligent operating condition adjustment system of the air separation plant based on multi-point temperature and pressure monitoring of the present invention, see Figure 2 .

[0025] In specific applications, the purification system in an air separation plant consists of two adsorbers and an electric heater. Each adsorber is in either a purification phase or a regeneration phase, switching between the two. The purification phase adsorbs adsorbents such as water, carbon dioxide, and hydrocarbons from the air being treated. The regeneration phase restores the adsorption function of adsorbents. The adsorbers typically have a double-layer bed structure, with activated alumina on the bottom and molecular sieve on the top. The intelligent operating condition adjustment system for the air separation plant includes the following components.

[0026] A plurality of internal temperature sensors are arranged inside the two adsorbers, and may be thermocouples, infrared thermal imagers, etc. A plurality of internal pressure sensors are fixed inside the two adsorbers, and may be air pressure sensors.

[0027] The regeneration data collection device is used to collect the internal temperature data detected in real time by the internal temperature sensor and the internal pressure data detected in real time by the internal pressure sensor in the adsorber during the regeneration phase and perform pre-processing such as data cleaning and standardization.

[0028] The reconstruction device is configured to reconstruct the adsorber's three-dimensional temperature field based on the preprocessed internal temperature data, and to reconstruct the adsorber's three-dimensional pressure field based on the preprocessed internal pressure data. In practical applications, the reconstruction device can reconstruct the three-dimensional temperature and pressure fields based on known data differences and a reconstruction algorithm. More preferably, the reconstruction device can achieve this reconstruction based on a mature machine learning model, resulting in more accurate three-dimensional temperature and pressure field models.

[0029] The feature engineering device is used to extract the three-dimensional temperature tomography matrix of the adsorber from the three-dimensional temperature field, and to extract the three-dimensional pressure tomography matrix of the adsorber from the three-dimensional pressure field. In an embodiment of the present invention, the feature engineering device cuts the three-dimensional temperature field into multiple parallel sections, arranges the temperature value of each parallel section into a temperature vector, and combines the temperature vectors of the multiple parallel sections into the three-dimensional temperature tomography matrix; the feature engineering device cuts the three-dimensional pressure field into multiple parallel sections, arranges the pressure value of each parallel section into a pressure vector, and combines the pressure vectors of the multiple parallel sections into the three-dimensional pressure tomography matrix.

[0030] The regeneration working condition adjustment model is a pre-trained machine learning model, which can adopt a traditional classification model or a pre-trained large language model. The regeneration working condition adjustment model is used to calculate the input three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix and the pre-recorded current working condition, and output a maintenance current working condition instruction or an adjustment working condition instruction. The maintenance current working condition instruction indicates that no working condition adjustment is required, and the adjustment working condition instruction indicates that an adjustment to the next working condition needs to be performed. The main controller is used to send out a working condition adjustment signal after receiving the adjustment working condition instruction, and the actuator can adjust the working condition of the adsorber in the regeneration stage according to the received working condition adjustment signal. The above working condition adjustment method of the regeneration working condition adjustment model is the working condition adjustment judgment logic under the default state.

[0031] Through the above steps, it is possible to utilize multi-point temperature and pressure sensing as well as automatic control technology and artificial intelligence technology to realize automatic adjustment of the working condition of the purification system, thereby improving the accuracy of the working condition adjustment. The temperature sensor and pressure sensor arranged inside the adsorber can capture more accurate real-time temperature and pressure data, and have better monitoring performance than the outlet sensing method in the prior art. In the above process, a reconstruction device is used to reconstruct the three-dimensional temperature field and three-dimensional pressure field models based on the monitored temperature and pressure data. The mature three-dimensional reconstruction machine learning model can be used to enrich the temperature data and pressure data, which helps to improve the working condition classification performance of the regeneration working condition adjustment model. Then, a feature engineering device is used to extract a large amount of temperature and pressure distribution data from the three-dimensional model to form a three-dimensional temperature tomography matrix and a three-dimensional pressure tomography matrix for the input model, thereby ensuring the accuracy of the model calculation.

[0032] See also Figure 3 and Figure 4 In one embodiment, the adsorber operating sequence during the regeneration phase can be in two scenarios. First, the operating sequence is: pressure relief, heating, cold blow, and pressurizing. Second, based on the first scenario, a preparation and switching condition are added. The operating sequence is: pressure relief, pressure relief, heating, cold blow, pressurizing, pressurizing, switching, and switching.

[0033] As a preferred solution, the intelligent operating condition adjustment system of the air separation equipment also includes: an inlet temperature sensor arranged at the inlet of the adsorber, and an outlet temperature sensor and an outlet pressure sensor arranged at the outlet of the adsorber, a sensor data processing device communicating with the inlet temperature sensor, the outlet temperature sensor and the outlet pressure sensor, and a control confirmation device communicating with the sensor data processing device.

[0034] After receiving the operating condition adjustment instruction, the main controller sends a working condition adjustment confirmation request to the control confirmation device, the working condition adjustment confirmation request carrying the current working condition identifier. Upon receiving the working condition adjustment confirmation request, the control confirmation device determines whether the sensor data transmitted by the inlet temperature sensor, the outlet temperature sensor, and the outlet pressure sensor meet pre-configured necessary adjustment conditions for the current working condition (i.e., the minimum conditions required to perform working condition adjustment), and returns the determination result to the main controller. The main controller receives the determination result and, if the determination result indicates that the necessary adjustment conditions for the current working condition are met, indicating that the calculation results of the regeneration working condition adjustment model are correct, issues a working condition adjustment signal. If the determination result indicates that the necessary adjustment conditions for the current working condition are not met, indicating that the calculation results of the regeneration working condition adjustment model may be erroneous, issues a manual intervention signal to the management unit. In this way, the main controller can utilize the control confirmation device and its connected external sensors to perform necessity verification on the adsorber inlet and outlet temperatures and pressures, thereby avoiding control decision errors and improving control system reliability.

[0035] Exemplarily, the necessary condition for adjusting the pressure relief condition is that the air pressure at the adsorber outlet is less than the first pressure threshold; the necessary condition for adjusting the heating condition is that the air temperature at the adsorber outlet is greater than the first temperature threshold; the necessary condition for adjusting the cold blowing condition is that the absolute value of the air temperature difference between the adsorber outlet and inlet is less than the preset second temperature threshold; the necessary condition for adjusting the charging condition is that the air pressure at the adsorber outlet is greater than the second pressure threshold.

[0036] Preferably, after the actuator adjusts the operating condition of the adsorber in the regeneration stage according to the operating condition adjustment signal, the purification system operates according to pre-configured default operating parameters of the adjusted operating condition; the default operating parameters may include: the heating temperature of the electric heater and the blowing flow rate of the contaminated nitrogen gas from the cold box.

[0037] In one embodiment, after adjusting the operating condition of the adsorber in the regeneration phase according to the operating condition adjustment signal, the actuator returns an adjustment completion notification to the main controller, and the main controller returns an adjustment success notification to the regeneration operating condition adjustment model. The regeneration operating condition adjustment model locally updates and records the current operating condition for subsequent calculations.

[0038] In an embodiment of the present invention, the intelligent operating condition adjustment system of the air separation equipment also includes: an operating parameter regression model that communicates with the regeneration operating condition adjustment model, and the operating parameter regression model is a pre-trained machine learning regression model, which can be a traditional regression model or a pre-trained large language model.

[0039] The regeneration operating condition adjustment model calculates the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix as well as the current operating condition. If the calculation result matches a pre-configured abnormal operating mode for the current operating condition (e.g., a low-temperature mode occurring during a heating condition), a working parameter calculation request is sent to the working parameter regression model. The working parameter calculation request carries the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix, the current operating condition, and the matched abnormal operating mode.

[0040] The operating parameter regression model regressively calculates target operating parameters for responding to the abnormal operating mode based on the received operating parameter calculation request, and sends an operating parameter update instruction carrying the target operating parameters to the main controller; after receiving the operating parameter update instruction, the main controller sends a parameter adjustment signal carrying the target operating parameters to the actuator; the target operating parameters include: the heating temperature of the electric heater and the blowing flow rate of the contaminated nitrogen gas;

[0041] The actuator operates the purification system according to the target operating parameters in the parameter adjustment signal, and upon completion of the operation, returns a change completion notification to the main controller. Upon receiving the change completion notification, the main controller returns a change success notification to the operating parameter regression model. Upon receiving the change success notification, the operating parameter regression model issues an operating parameter update notification containing the target operating parameters to the regeneration operating condition adjustment model. The regeneration operating condition adjustment model calculates the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix, the current operating conditions, and the target operating parameters to execute the operating condition adjustment judgment logic after the operating parameters are updated. Through the above steps, the regeneration operating condition adjustment model can detect abnormal patterns, the operating parameter regression model can perform the update calculation of the operating parameters, and the main controller can drive the update of the operating parameters, thereby achieving dynamic update of the operating parameters during the operating condition maintenance process, which is conducive to optimizing the operation process and operation quality through the change of operating parameters. The above operating condition adjustment method of the regeneration operating condition adjustment model is to execute the operating condition adjustment judgment logic after the operating parameter update, which is different from the operating condition adjustment judgment logic in the default state.

[0042] In an optional technical solution, the intelligent operating condition adjustment system of the air separation equipment of the present invention further includes: multiple humidity sensors fixed inside the adsorber, multiple adsorbate detection devices installed at different positions near the adsorber outlet, and the following parts.

[0043] The purification data collection device is used to collect and preprocess the internal temperature data detected in real time by the internal temperature sensor of the adsorber during the purification phase, the internal humidity data detected in real time by the humidity sensor, and the adsorbate front position data collected by the adsorbate detection device. The front refers to the front face where the adsorbate arrives (i.e., the face closest to the adsorber outlet). The feature extraction device is used to combine the preprocessed internal temperature data, internal humidity data, and adsorbate front position data into purification features. The purification decision model is a pretrained machine learning model that calculates the input purification features and outputs a default operation signal or an emergency switch signal. The default operation signal indicates that no emergency operation is required, while the emergency switch signal indicates that the adsorber is in an emergency switch state and needs to switch between the purification phase and the regeneration phase within a short period of time. The emergency switch signal carries the remaining regeneration time of the adsorber in the regeneration phase and is used to instruct the two adsorbers to switch between the purification phase and the regeneration phase within this remaining regeneration time.

[0044] The rearrangement device communicates with the purification decision model and the regeneration condition adjustment model. Based on the current three-dimensional temperature chromatography matrix, three-dimensional pressure chromatography matrix, and current operating conditions of the adsorber during the regeneration phase transmitted by the regeneration condition adjustment model, as well as the remaining regeneration time transmitted by the purification decision model, it calculates the execution time of each pending operating condition of the adsorber during the regeneration phase and the target operating parameters of each pending operating condition under the emergency switching state. The device then generates an emergency switching instruction carrying the execution time and target operating parameters and sends it to the main controller. The rearrangement device can be implemented using a traditional algorithm or a machine learning model.

[0045] After receiving the emergency switching instruction, the main controller sends the execution duration and the target working parameters in the emergency switching instruction to the execution mechanism, so that the execution structure performs corresponding operations on the purification system; the main controller also sends an emergency switching notification carrying the execution duration and the target working parameters to the regeneration working condition adjustment model. The regeneration working condition adjustment model calculates the current three-dimensional temperature chromatography matrix, three-dimensional pressure chromatography matrix, current working condition, execution duration and target working parameters of the adsorber in the regeneration stage to execute the working condition adjustment judgment logic under the emergency switching state. The working condition adjustment judgment logic under the emergency switching state is different from the working condition adjustment judgment logic under the default state.

[0046] Through the above steps, it is possible to use sensors inside the adsorber in the purification stage to monitor emergency situations such as high temperature (which will cause a serious decrease in adsorption capacity and lead to the presence of impurities in the output gas), high humidity (which will cause irreversible damage to the molecular sieve), and adsorbate penetration (i.e., the adsorbate approaches the adsorber outlet). The rearrangement device calculates the execution time of each operating condition for response (generally shortening the duration of each operating condition) and the working parameters, which are finally driven and implemented by the main controller, thereby completing adsorber regeneration and stage switching in a short time, achieving effective response to the above emergencies. Accordingly, the regeneration operating condition adjustment model will execute the operating condition adjustment judgment logic under the emergency switching state, and generally complete the operating condition adjustment according to the execution time of each operating condition calculated by the rearrangement device.

[0047] Furthermore, in an embodiment of the present invention, for the adsorber in the purification stage, the control confirmation device acquires sensor data of the purified air collected by the outlet temperature sensor and the outlet pressure sensor, and determines whether the sensor data meets design specifications. If not, an alarm signal is issued to the main controller. Upon receiving the alarm signal, the main controller issues a high-temperature regeneration instruction to the actuator, causing the actuator to switch the purification system to a high-temperature regeneration mode. The main controller also issues a high-temperature regeneration notification to the regeneration condition adjustment model, causing it to perform condition adjustment determinations under the high-temperature regeneration mode. The high-temperature regeneration mode has higher regeneration intensity and efficiency than the conventional regeneration mode. In the above process, if the outlet sensor detects that the purified gas does not meet design specifications (e.g., the presence of free water or substandard temperature), indicating that the conventional regeneration mode cannot meet operational requirements, the high-temperature regeneration mode is activated to improve purification and regeneration quality. At this time, the regeneration operating condition adjustment model can execute the operating condition adjustment judgment logic under the high-temperature regeneration mode. In actual applications, in addition to executing the operating condition adjustment judgment logic under the default state, the regeneration operating condition adjustment model can also execute the operating condition adjustment judgment logic after the working parameters are updated, the operating condition adjustment judgment logic under the emergency switching state, and the operating condition adjustment judgment logic under the high-temperature regeneration mode.

[0048] In the technical solution of the present invention, accurate perception of the working environment of the purification system and precise calculation of the timing of operating condition adjustment can be achieved through sensor technology, automatic control technology and artificial intelligence technology for multi-point temperature and pressure monitoring, avoiding the defects of relying on manual experience to adjust the operating conditions, and helping to improve the quality and efficiency of air separation.

[0049] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. An intelligent operating condition adjustment system for an air separation unit based on multi-point temperature and pressure monitoring, the air separation unit comprising: A purification system comprising two adsorbers and an electric heater, wherein each adsorber is in a purification stage or a regeneration stage; wherein the intelligent operating condition adjustment system for the air separation plant comprises: Multiple internal temperature sensors are arranged inside the two adsorbers; Multiple internal pressure sensors, fixed inside the two adsorbers; a regeneration data collection device for collecting internal temperature data detected in real time by an internal temperature sensor and internal pressure data detected in real time by an internal pressure sensor in the adsorber during the regeneration phase and performing preprocessing; a reconstruction device for reconstructing the three-dimensional temperature field of the adsorber according to the pre-processed internal temperature data, and reconstructing the three-dimensional pressure field of the adsorber according to the pre-processed internal pressure data; a feature engineering device for extracting a three-dimensional temperature tomography matrix of the adsorber from the three-dimensional temperature field and extracting a three-dimensional pressure tomography matrix of the adsorber from the three-dimensional pressure field; The regeneration operating condition adjustment model is a pre-trained machine learning model that calculates the input three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix and the pre-recorded current operating condition, and outputs an instruction to maintain the current operating condition or an instruction to adjust the operating condition; The main controller sends out an operating condition adjustment signal after receiving the operating condition adjustment instruction; The actuator adjusts the operating condition of the adsorber during the regeneration phase according to the received operating condition adjustment signal.

2. The intelligent operating condition adjustment system for air separation equipment according to claim 1, characterized in that: Also includes: an inlet temperature sensor disposed at the adsorber inlet, an outlet temperature sensor and an outlet pressure sensor disposed at the adsorber outlet, a sensor data processing device communicating with the inlet temperature sensor, the outlet temperature sensor and the outlet pressure sensor, and a control confirmation device communicating with the sensor data processing device; After receiving the working condition adjustment instruction, the main controller sends a working condition adjustment confirmation request to the control confirmation device, wherein the working condition adjustment confirmation request carries a current working condition identifier; The control confirmation device receives the operating condition adjustment confirmation request, determines whether the sensor data transmitted by the inlet temperature sensor, the outlet temperature sensor, and the outlet pressure sensor meet pre-configured necessary conditions for adjustment of the current operating condition, and returns the determination result to the main controller; The main controller receives the judgment result; if the judgment result indicates that the necessary adjustment conditions of the current working condition are met, it sends a working condition adjustment signal; If the judgment result indicates that the necessary adjustment conditions for the current working condition are not met, a manual intervention signal is sent to the management unit.

3. The intelligent operating condition adjustment system for air separation equipment according to claim 2, characterized in that: After the actuator adjusts the operating condition of the adsorber in the regeneration phase according to the operating condition adjustment signal, the purification system operates according to pre-configured default operating parameters of the adjusted operating condition; The default operating parameters include: the heating temperature of the electric heater and the blowing flow rate of the contaminated nitrogen gas from the cold box.

4. The intelligent operating condition adjustment system for an air separation plant according to claim 3, characterized in that: After adjusting the operating condition of the adsorber in the regeneration phase according to the operating condition adjustment signal, the actuator returns an adjustment completion notification to the main controller, and the main controller returns an adjustment success notification to the regeneration operating condition adjustment model, and the regeneration operating condition adjustment model locally updates and records the current operating condition.

5. The intelligent operating condition adjustment system for an air separation plant according to claim 4, characterized in that: Also includes: an operating parameter regression model in communication with the regeneration operating condition adjustment model, the operating parameter regression model being a pre-trained machine learning regression model; The regeneration operating condition adjustment model calculates the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix as well as the current operating condition. If the calculation result matches the abnormal operating mode pre-configured for the current operating condition, a working parameter calculation request is sent to the working parameter regression model; the working parameter calculation request carries the current three-dimensional temperature tomography matrix and the three-dimensional pressure tomography matrix, the current operating condition, and the matched abnormal operating mode. The operating parameter regression model outputs target operating parameters for responding to the abnormal operating mode based on the received operating parameter calculation request, and sends an operating parameter update instruction carrying the target operating parameters to the main controller; after receiving the operating parameter update instruction, the main controller sends a parameter adjustment signal carrying the target operating parameters to the actuator; the target operating parameters include: the heating temperature of the electric heater and the blowing flow rate of the contaminated nitrogen gas; The actuator operates the purification system according to the target operating parameters in the parameter adjustment signal, and returns a change completion notification to the main controller after the operation is completed; after receiving the change completion notification, the main controller returns a change success notification to the operating parameter regression model; after receiving the change success notification, the operating parameter regression model sends an operating parameter update notification carrying the target operating parameters to the regeneration operating condition adjustment model; the regeneration operating condition adjustment model calculates the current three-dimensional temperature chromatography matrix and three-dimensional pressure chromatography matrix, the current operating conditions and the target operating parameters to perform operating condition adjustment judgment after the working parameters are updated.

6. The intelligent operating condition adjustment system for an air separation plant according to claim 5, characterized in that: Also includes: a plurality of humidity sensors fixed inside the adsorber; a plurality of adsorbate detection devices installed at different locations near the adsorber outlet; a purification data collection device for collecting the internal temperature data detected in real time by the internal temperature sensor in the adsorber during the purification stage, the internal humidity data detected in real time by the humidity sensor, and the adsorbate front position data collected by the adsorbate detection device, and performing preprocessing; a feature extraction device for splicing the pre-processed internal temperature data, the internal humidity data, and the adsorbate front position data into purified features; The purification decision model is a pre-trained machine learning model that calculates the input purification features and outputs a default working signal or an emergency switching signal; the emergency switching signal carries the remaining regeneration time of the adsorber in the regeneration phase, which is used to indicate whether the two adsorbers should switch between the purification phase and the regeneration phase within the remaining regeneration time; a rearrangement device, communicating with the purification decision model and the regeneration operating condition adjustment model, and calculating, based on the current three-dimensional temperature chromatography matrix, three-dimensional pressure chromatography matrix, and current operating condition of the adsorber in the regeneration phase transmitted by the regeneration operating condition adjustment model, and the remaining regeneration time transmitted by the purification decision model, an execution time of each pending operating condition of the adsorber in the regeneration phase and a target operating parameter of each pending operating condition in an emergency switching state; generating an emergency switching instruction carrying the execution duration and the target working parameters and sending it to the main controller; After receiving the emergency switching instruction, the main controller sends the execution duration and the target working parameters in the emergency switching instruction to the actuator, so that the actuator performs corresponding operations on the purification system; the main controller also sends an emergency switching notification carrying the execution duration and the target working parameters to the regeneration condition adjustment model. The regeneration condition adjustment model calculates the current three-dimensional temperature chromatography matrix, three-dimensional pressure chromatography matrix, current working conditions, execution duration and target working parameters of the adsorber in the regeneration stage to perform condition adjustment judgment under the emergency switching state.

7. The intelligent operating condition adjustment system for an air separation plant according to claim 6, characterized in that: For the adsorber in the purification stage, the control confirmation device obtains sensor data of the purified air collected by the outlet temperature sensor and the outlet pressure sensor, and determines whether the sensor data meets the design indicators; If the design indicators are not met, an alarm signal is sent to the main controller; After receiving the alarm signal, the main controller sends a high-temperature regeneration instruction to the actuator, so that the actuator changes the purification system to the high-temperature regeneration mode; the main controller also sends a high-temperature regeneration notification to the regeneration condition adjustment model, so that the regeneration condition adjustment model executes the condition adjustment judgment under the high-temperature regeneration mode.

8. The intelligent operating condition adjustment system for an air separation plant according to claim 7, characterized in that: The feature engineering device cuts the three-dimensional temperature field into a plurality of parallel sections, arranges the temperature value of each parallel section into a temperature vector, and combines the temperature vectors of the plurality of parallel sections into the three-dimensional temperature tomography matrix; The feature engineering device cuts the three-dimensional pressure field into multiple parallel sections, arranges the pressure value of each parallel section into a pressure vector, and combines the pressure vectors of multiple parallel sections into the three-dimensional pressure tomography matrix.

9. The intelligent operating condition adjustment system for an air separation plant according to claim 2, characterized in that: The operating conditions of the adsorber during the regeneration stage are: pressure relief condition, heating condition, cold blowing condition, and pressurizing condition, or, preparation for pressure relief condition, pressure relief condition, preparation for heating condition, heating condition, cold blowing condition, preparation for pressurizing condition, pressurizing condition, preparation for switching condition, and switching condition.

10. The intelligent operating condition adjustment system for an air separation plant according to claim 9, characterized in that: The necessary condition for adjusting the pressure relief condition is that the air pressure at the adsorber outlet is less than the first pressure threshold; the necessary condition for adjusting the heating condition is that the air temperature at the adsorber outlet is greater than the first temperature threshold; the necessary condition for adjusting the cold blowing condition is that the absolute value of the air temperature difference between the adsorber outlet and inlet is less than the preset second temperature threshold; the necessary condition for adjusting the charging condition is that the air pressure at the adsorber outlet is greater than the second pressure threshold.

Images