Nitrogen plug early warning and automatic disposal control method and related device

Abstract

The application is suitable for the technical field of industrial process control, and particularly relates to a nitrogen plug early warning and automatic disposal control method and related device. The method comprises the following steps: acquiring multi-parameter real-time data of each measuring point of a cold box; determining calculation data corresponding to the multi-parameter real-time data; determining nitrogen plug risk data based on the determined calculation data; acquiring operation data of the cold box and calculating current operation condition data; calculating an emergency disposal threshold value and an enhanced monitoring threshold value based on the operation condition data; and performing hierarchical processing based on a comparison result between the nitrogen plug risk data and the emergency disposal threshold value and the enhanced monitoring threshold value. The method provided by the application can solve the problem that the existing technology mainly adopts a single-parameter fixed threshold value alarm method for nitrogen plug detection and control, thereby leading to low accuracy of nitrogen plug early warning.

Description

Technical Field

[0001] This application belongs to the field of industrial process control technology, and in particular relates to a nitrogen plugging early warning and automatic handling control method and related device. Background Technology

[0002] In LNG (liquefied natural gas) liquefaction and LNG receiving terminal regasification processes, the cold box is the heat exchange equipment for achieving cryogenic liquefaction of natural gas. Taking the common propane pre-cooled mixed refrigerant liquefaction process as an example, the interior of the cold box consists of hundreds of layers of aluminum plates and fins brazed together to form multiple isolated heat exchange channels. The feedstock natural gas undergoes counter-current heat exchange with the mixed refrigerant within these channels, gradually decreasing in temperature from ambient to approximately -162°C to achieve liquefaction. In actual operation, the feedstock natural gas inevitably contains trace amounts of nitrogen. Since nitrogen has a boiling point of -196°C, far below the cold box operating temperature (approximately -162°C), nitrogen cannot liquefy in the cryogenic section of the cold box. Instead, it gradually accumulates in gaseous form at the cold end of the heat exchange channels, forming a gas phase blockage. This blockage occupies the effective flow cross-section of the channels, hindering the normal flow of liquid hydrocarbons and causing a gradual deterioration in local heat transfer efficiency. This phenomenon is known as nitrogen blockage in the LNG industry.

[0003] Current technologies for detecting and controlling nitrogen plugging primarily employ a single-parameter fixed threshold alarm method. This involves setting a fixed alarm threshold for a key parameter, triggering an alarm and executing a standardized response procedure when the parameter exceeds the threshold. However, the normal range of parameter variation differs under different operating conditions. The fluctuation amplitude of the same parameter, which is normal under high-load operation, may exhibit abnormal signal characteristics under low-load operation. In this situation, existing detection mechanisms frequently misinterpret normal fluctuations as abnormalities during high-load operation, triggering unnecessary actions. During low-load or stable operation, they fail to adequately identify true early nitrogen plugging signals, allowing the nitrogen plug to progress to a severe stage before being detected, thus leading to low accuracy in nitrogen plugging early warning systems. Summary of the Invention

[0004] This application provides a nitrogen plugging early warning and automatic handling control method and related device, which can solve the problem that the existing technology mainly uses a single-parameter fixed threshold alarm method for nitrogen plugging detection and control, resulting in low accuracy of nitrogen plugging early warning.

[0005] In a first aspect, embodiments of this application provide a nitrogen plugging early warning and automatic handling control method, including: Acquire real-time data of multiple parameters at various measurement points in the cold box; The calculation data corresponding to the multi-parameter real-time data is determined; wherein, the calculation data corresponding to the multi-parameter real-time data includes thermal deviation data, abnormal pattern data, sensor reliability data, and process safety level data; the thermal deviation data is used to reflect the degree of difference between the multi-parameter real-time data and the statistical characteristics of historical time windows; the abnormal pattern data is used to reflect the rarity of the abnormal pattern presented by the current data combination in historical periods; the sensor reliability data is used to reflect the data reliability weight of the current sensor based on the consistency of measurements by adjacent sensors within the same heat exchange section; the process safety level data is used to reflect the nitrogen plugging hazard level corresponding to the multi-parameter real-time data. Nitrogen plug risk data are determined based on the calculated data; Acquire the operating data of the cold box and calculate the current operating condition data; Calculate emergency response thresholds and enhanced monitoring thresholds based on the aforementioned operational data; The data is classified based on the comparison between the nitrogen plugging risk data and the emergency response threshold and the enhanced monitoring threshold.

[0006] The technical solutions described in this application embodiment have at least the following technical effects: The nitrogen plugging early warning and automatic handling control method provided in this application determines nitrogen plugging risk data based on calculated data from real-time multi-parameter data, including thermal deviation data, abnormal mode data, sensor reliability data, and process safety level data. This allows for a comprehensive quantitative assessment of the nitrogen plugging risk level for each set of real-time multi-parameter data from multiple dimensions, improving the accuracy of early nitrogen plugging signal identification. By using sensor reliability data, the influence of sensor data with large consistency deviations on the nitrogen plugging risk data determination process is mitigated, reducing the possibility of misjudgments due to sensor failure or drift. Furthermore, by acquiring cold box operating data and calculating current operating condition data, emergency handling thresholds and enhanced monitoring thresholds are calculated based on the operating condition data. This correlation between the calculation process of these two thresholds and the operating condition data allows the thresholds to adapt to parameter fluctuations under different operating conditions, reducing the possibility of false alarms or missed alarms due to changes in operating conditions. By classifying and processing the results of comparisons between nitrogen surge risk data and emergency response thresholds and enhanced monitoring thresholds, different levels of response actions can be executed according to the position of the nitrogen surge risk data relative to the two thresholds. This enables differentiated automatic response that matches the degree of nitrogen surge risk, which helps to reduce false alarm rates while ensuring the timeliness of nitrogen surge response, thereby improving the accuracy of nitrogen surge early warning.

[0007] Secondly, embodiments of this application provide a nitrogen plugging warning and automatic handling control system, applied to a nitrogen plugging warning device, for implementing the nitrogen plugging warning and automatic handling control method described in any one of the first aspects above. The nitrogen plugging warning and automatic handling control system includes: The acquisition unit is used to acquire real-time data of multiple parameters at various measurement points in the cold box. An analysis unit is used to determine the calculated data corresponding to the multi-parameter real-time data; wherein, the calculated data corresponding to the multi-parameter real-time data includes thermal deviation data, anomaly pattern data, sensor reliability data, and process safety level data; the thermal deviation data is used to reflect the degree of difference between the multi-parameter real-time data and the statistical characteristics of historical time windows; the anomaly pattern data is used to reflect the rarity of the anomaly pattern presented by the current data combination in historical periods; the sensor reliability data is used to reflect the data reliability weight of the current sensor based on the consistency of measurements by adjacent sensors within the same heat exchange section; and the process safety level data is used to reflect the nitrogen plugging hazard level corresponding to the multi-parameter real-time data. A calculation unit is used to determine nitrogen plug risk data based on the determined calculation data; The data acquisition unit is used to acquire the operating data of the cold box and calculate the current operating condition data; The generation unit is used to calculate emergency response thresholds and enhanced monitoring thresholds based on the operating condition data. The processing unit is used to perform graded processing based on the comparison results between the nitrogen plug risk data and the emergency response threshold and the enhanced monitoring threshold.

[0008] Thirdly, embodiments of this application provide a nitrogen plugging early warning device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the nitrogen plugging early warning and automatic handling control method described in any of the first aspects above.

[0009] It is understood that the beneficial effects of the second and third aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1This is a schematic flowchart of a nitrogen plugging early warning and automatic handling control method provided in an embodiment of this application; Figure 2 This is a schematic diagram of a nitrogen plug provided in an embodiment of this application; Figure 3 This is a schematic diagram of residual trend analysis in a nitrogen plugging early warning and automatic handling control method provided in an embodiment of this application; Figure 4 This is a schematic diagram of setting a pending execution flag in a nitrogen plugging early warning and automatic handling control method provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of the nitrogen plugging early warning and automatic handling control system provided in the embodiments of this application; Figure 6 This is a schematic diagram of the nitrogen plugging warning device provided in the embodiments of this application. Detailed Implementation

[0012] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0013] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0014] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0015] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if the described condition or event is detected" may be interpreted, depending on the context, as "once determined," "in response to determination," "once the described condition or event is detected," or "in response to the detection of the described condition or event."

[0016] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0017] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0018] In related technologies, existing technologies are aimed at nitrogen plugging (such as...) Figure 2 The detection and control of nitrogen plugging mainly adopts the single-parameter fixed threshold alarm method. However, the dynamic changes in the operating conditions of the cold box cause the normal fluctuation range of the process parameters to change continuously with the operating status. The fluctuation amplitude of the same parameter is either a normal phenomenon or an abnormal signal under different operating conditions, resulting in the problem of low accuracy of nitrogen plugging warning.

[0019] Furthermore, existing technologies do not consider the impact of sensor malfunction or drift on detection results. In the cryogenic operating environment of LNG cold boxes, sensors face harsh conditions such as vibration, ice blockage, and electromagnetic interference, and the readings of individual sensors may malfunction or drift. When abnormal readings are mistaken for nitrogen blockage signals, it can lead to false alarms; when the readings of faulty sensors mask the true abnormal signals, it can lead to missed detections.

[0020] To address the aforementioned issues, embodiments of this application provide a nitrogen plugging early warning and automatic handling control method and related apparatus.

[0021] This method determines nitrogen plug risk data by identifying calculated data from real-time multi-parameter data, including thermal deviation data, abnormal mode data, sensor reliability data, and process safety level data. Based on this calculated data, it enables a comprehensive quantitative assessment of the nitrogen plug risk level for each set of real-time multi-parameter data from multiple dimensions, improving the accuracy of early nitrogen plug signal identification. By using sensor reliability data, the influence of sensor data with large consistency deviations on the nitrogen plug risk data determination process is mitigated, reducing the possibility of misjudgments due to sensor failure or drift. Furthermore, by acquiring cold box operating data and calculating current operating condition data, emergency response thresholds and enhanced monitoring thresholds are calculated based on this operating condition data. This correlation between the calculation process of these two thresholds and the operating condition data allows the thresholds to adapt to parameter fluctuations under different operating conditions, reducing the possibility of false alarms or missed alarms due to changes in operating conditions. By classifying and processing the results of comparisons between nitrogen surge risk data and emergency response thresholds and enhanced monitoring thresholds, different levels of response actions can be executed according to the position of the nitrogen surge risk data relative to the two thresholds. This enables differentiated automatic response that matches the degree of nitrogen surge risk, which helps to reduce false alarm rates while ensuring the timeliness of nitrogen surge response, thereby improving the accuracy of nitrogen surge early warning.

[0022] The nitrogen plugging warning and automatic handling control method provided in this application embodiment can be applied to a nitrogen plugging warning device. In this case, the nitrogen plugging warning device is the executing entity of the nitrogen plugging warning and automatic handling control method provided in this application embodiment. This application embodiment does not impose any restrictions on the specific type of nitrogen plugging warning device.

[0023] For example, a nitrogen plugging warning device may include a data acquisition device (e.g., temperature sensors, pressure sensors, etc. installed on LNG pipelines for real-time acquisition of temperature and pressure), a data processing and judgment device (e.g., a programmable logic controller (PLC) or a distributed control system (DCS) dedicated computing module for receiving data acquired by the data acquisition device and performing real-time calculations), and a warning and protection execution device (e.g., a control room audible and visual alarm, a field alarm bell, and an interlocking shut-off valve or pump speed regulation circuit for control or warning based on the results calculated by the data processing and judgment device).

[0024] To better understand the nitrogen plugging warning and automatic handling control method provided in the embodiments of this application, the specific implementation process of the nitrogen plugging warning and automatic handling control method provided in the embodiments of this application will be described by way of example below.

[0025] Figure 1 This illustration shows a schematic flowchart of a nitrogen plugging warning and automatic handling control method provided in an embodiment of this application. The nitrogen plugging warning and automatic handling control method includes: S100 acquires real-time data of multiple parameters at various measurement points in the cold box.

[0026] It can be understood that the measurement points in the cold box can be the measurement locations corresponding to various sensors distributed at key positions in each heat exchange section of the cold box. Multi-parameter real-time data can be the measured values ​​of a series of process parameters periodically collected by these sensors and transmitted to the data processing and judgment device during the operation of the cold box. Multi-parameter real-time data includes at least: (a) Approach temperature of each heat exchange section: This can be collected by multiple sets of thermocouples distributed at the hot end, intermediate section, and cold end of the cold box. In a counter-current heat exchange configuration of a plate-fin heat exchanger, the approach temperature can be the difference between the outlet temperature of the higher-temperature fluid and the inlet temperature of the lower-temperature fluid within each heat exchange section, used to measure the heat exchange efficiency of the current heat exchange section. When nitrogen accumulates in this heat exchange section, the effective heat exchange area decreases, heat exchange is insufficient, and the approach temperature will gradually increase.

[0027] (b) Differential pressure in each heat exchange channel: This can be collected by differential pressure transmitters installed at the inlet and outlet of each channel to reflect the fluid flow resistance within the channel. The differential pressure is relatively stable during normal operation. When air resistance reduces the effective cross-section of the channel, the resistance of the fluid through the channel increases, and the differential pressure rises accordingly.

[0028] (c) Outlet nitrogen mole fraction: This can be collected by a gas chromatograph installed on the cold box outlet pipeline to reflect the trend of nitrogen concentration change at the cold box outlet. Under normal operation, the outlet nitrogen mole fraction is basically consistent with the feed nitrogen mole fraction. When nitrogen accumulates in the cold box, the outlet nitrogen mole fraction will fluctuate or decrease (because nitrogen is trapped inside the cold box).

[0029] (d) Fluid flow rate in each channel: This can be collected by mass flow meters installed in each channel. During normal operation, the flow rate in each channel is stable. When air resistance increases, the flow rate in the obstructed channel will decrease accordingly.

[0030] The acquisition cycle for the above four types of data can be configured according to the specific data type and the data refresh frequency of the cold box DCS system. The nitrogen plugging early warning device reads the latest measurement value of each data from the DCS system at a fixed polling cycle (e.g., 10 seconds), and combines the measurement values ​​of each parameter obtained at the same polling time into a set of multi-parameter real-time data. Each set of data also includes a corresponding timestamp and identification information of each measurement point.

[0031] S200, determine the calculated data corresponding to the multi-parameter real-time data. This calculated data includes thermal deviation data, anomaly pattern data, sensor reliability data, and process safety level data. Thermal deviation data reflects the degree of difference between the multi-parameter real-time data and the statistical characteristics of historical time windows. Anomaly pattern data reflects the rarity of the anomaly pattern presented by the current data combination within historical periods. Sensor reliability data reflects the data reliability weight of the current sensor based on the consistency of measurements from adjacent sensors within the same heat exchange section. Process safety level data reflects the nitrogen plug hazard level corresponding to the multi-parameter real-time data.

[0032] After receiving each set of multi-parameter real-time data, the current data can be evaluated from multiple dimensions to quantify the nitrogen plugging risk level reflected by that set of data. The determined indicators may include: thermal deviation data, abnormal mode data, sensor reliability data, and process safety level data. The methods for determining each indicator are explained in detail below.

[0033] Thermal deviation data is used to characterize the degree of deviation of multi-parameter real-time data based on the statistical characteristics of historical time windows, and to quantify the magnitude of the deviation between the current data point and the normal operating mode. It can be understood that during normal operation, the parameter values ​​of each measurement point in the cold box will fluctuate within a certain range. This fluctuation is caused by various normal operating disturbances (such as load fine-tuning, changes in ambient temperature, etc.). Thermal deviation data uses statistical methods to measure the degree to which the current data deviates from the normal fluctuation range; the greater the deviation, the more likely the current data reflects an abnormal event.

[0034] The determination of thermal deviation data can be based on the mean and standard deviation within a sliding time window. Specifically, for a measurement value x(t) at time t, let the historical data sequence within the sliding time window W (e.g., data from the past 2 hours, the length of which can be configured according to the operating characteristics of the cold box) be {x(t-N+1), x(t-N+2), ..., x(t)}, where N is the number of data points within the window. Calculate the mean μ and standard deviation σ of the historical data within this window: mean μ = (1 ÷ N) × Σ(x(ti)), where i ranges from 0 to N-1; standard deviation σ = ... , where i ranges from 0 to N-1. The thermal deviation Z(t) of the current measured value x(t) is calculated as follows: Z(t) = |x(t) - μ| ÷ (σ + L), where L is a very small positive number (such as 0.001) to avoid division by zero errors when the standard deviation approaches zero (i.e., when the historical data has almost no fluctuations).

[0035] For multi-parameter real-time data containing multiple measurement points, the thermal deviation is calculated for the approach temperature of each heat exchange section. Then, the maximum value of the thermal deviation of all heat exchange sections is taken as the overall thermal deviation data of the data set. The purpose of taking the maximum value is that, in the initial stage of nitrogen plugging, anomalies usually first appear in a certain local area (such as the cold end), and using the maximum value can sensitively capture the local anomaly signal.

[0036] Anomaly pattern data is used to characterize the rarity of the anomaly pattern presented by the current multi-parameter data combination within a historical period, and to measure the rarity of the current anomaly pattern. It is understood that various types of anomaly patterns may occur during cold box operation, and different patterns have different frequencies of occurrence historically. Frequently occurring patterns are more likely to be variations of normal perturbations, while rare patterns are more likely to indicate a true nitrogen plugging event. By matching the current data combination with historical patterns, the occurrence of the current anomaly can be quantified.

[0037] The method for determining abnormal pattern data can be as follows: construct a feature vector from multi-parameter data within the current time window (consistent with the time window of the thermal deviation data). Specifically, three types of parameters are selected as key parameters: approach temperature, channel differential pressure, and fluid flow rate. For each type of parameter, four statistics are calculated within the current window: mean, standard deviation, linear regression slope, and maximum value. The cosine similarity of this feature vector with the standard feature vectors of each pattern in the feature pattern library is then calculated. The formula for calculating cosine similarity is: (A · B) ÷ (||A|| × ||B||), where A is the current feature vector, B is the standard feature vector of a certain pattern in the pattern library, A · B is the inner product of the two vectors, and ||A|| and ||B|| are the Euclidean norms of the two vectors, respectively. Before calculating cosine similarity, the components in the 12-dimensional feature vector need to be normalized to eliminate the influence of different parameter dimensions.

[0038] The feature pattern library can include multiple patterns. The standard feature vectors for each pattern can be pre-obtained based on historical nitrogen plugging event data and normal operation data by taking the cluster centers of feature vectors of similar events. Furthermore, after each confirmed nitrogen plugging event, the feature pattern library adds the event's feature vector as a new sample to the corresponding pattern's cluster sample set and recalculates the cluster centers to continuously update and optimize the pattern library. For example, it can include a gradual pattern: corresponding to the slow development of nitrogen plugging, where the approach temperature of multiple heat exchange sections increases synchronously and slowly, and the differential pressure rises slowly; and a sudden pattern: corresponding to the abnormal process of rapid nitrogen plugging, where the approach temperature of one or a few heat exchange sections increases sharply in a short period of time, indicating rapid accumulation of nitrogen in the channel or the merging and expansion of existing small air resistances.

[0039] After calculating the cosine similarity between the current feature vector and the standard feature vectors of each pattern, the pattern with the highest similarity that exceeds a preset matching threshold (for example, the preset matching threshold can be selected based on historical data verification, such as 0.85) is selected as the matching result. If the highest similarity does not exceed the matching threshold, it is determined that no known pattern has been matched, and the scarcity value is set to a default neutral value (such as 0.5, indicating that it is neither scarce nor common).

[0040] If a pattern is successfully matched, a scarcity value is calculated based on the frequency of that pattern's occurrence within a historical statistical period (such as the past year). Specifically, scarcity = 1 - (number of times the pattern occurred ÷ total number of times all anomalous patterns occurred within the historical statistical period). The closer the scarcity value is to 1, the rarer the pattern and the higher the risk; the closer the scarcity value is to 0, the more common the pattern and the more likely it is to be just a normal disturbance.

[0041] Sensor reliability data reflects the differences in reliability among different sensors in a cold box system under harsh operating conditions such as low-temperature vibration, ice blockage, and signal line interference. If a sensor generates an abnormal reading due to its own failure or drift, this reading should not be considered a valid signal of nitrogen plugging risk. Therefore, sensor reliability data can be used to assess the reliability of each sensor's data, and sensor data with low reliability can be weighted less in the scoring calculation.

[0042] Sensor reliability data can be evaluated based on the measurement consistency of adjacent sensors within the same heat exchange section. Specifically, for the i-th sensor in a heat exchange section (such as the cold end), its current reading xi is taken, along with the readings xj of 2-3 similar sensors physically adjacent to sensor i within the same heat exchange section (j∈N(i), where N(i) is the set of adjacent sensors of sensor i). The relative deviation between sensor i and its adjacent sensors is calculated as: relative deviation i = |xi - mean(xj)| ÷ mean(xj), where mean(xj) is the average value of the readings of adjacent sensors. When the relative deviation i exceeds a preset threshold (which can be set according to industry standards or actual needs, for example, 10%), it is determined that the sensor has a large consistency deviation with its adjacent sensors, and the weight coefficient of its data in the scoring calculation is reduced from the standard weight coefficient to a weakened weight coefficient (for example, the weight value can be reduced). In a specific embodiment, the standard weight coefficient is 0.75, and the weakened weight coefficient is 0.3, that is, the weight of the sensor with a large consistency deviation in the scoring calculation is reduced to 40% of the standard value. When the relative deviation i does not exceed the preset threshold, the sensor consistency is determined to be normal, and the standard weighting coefficient remains unchanged.

[0043] It should be noted that the temperature difference between different heat exchange sections (hot end, middle section, cold end) can reach tens to hundreds of degrees Celsius, which is normal operating condition. Therefore, the consistency comparison is performed between similar sensors within the same heat exchange section.

[0044] Process safety level data is used to characterize the nitrogen plugging hazard level corresponding to the data, and a four-level label can be used. This can be understood as being determined by directly comparing key process parameters with preset safety thresholds.

[0045] The specific definition of the Level 4 identifier is as follows: Level 0, Normal Operating Range: All key process parameters are within the historical normal fluctuation range. Level 1, Early Warning: A single or a few parameters slightly deviate from the normal range. For example, only the cold end approach temperature exceeds the historical normal range but does not exceed 80% of the maximum allowable design value. Level 2, Significant Warning: Multiple parameters significantly deviate from the normal range but have not yet approached the safety threshold. For example, the approach temperatures of multiple heat exchange sections exceed the historical normal range, or the channel differential pressure reaches 90% of the maximum allowable design value. Level 3 (Preset Highest Hazard Level): Triggered by any of the following conditions: The cold box outlet temperature continuously approaches the LNG product non-compliance threshold (e.g., continuously above -155°C for more than 10 minutes); the differential pressure of any heat exchange channel approaches the equipment structural safety limit (e.g., reaches 95% of the maximum allowable design value); the approach temperatures of multiple heat exchange sections (≥3) simultaneously exceed the maximum allowable design value.

[0046] S300 determines nitrogen plug risk data based on the determined calculation data.

[0047] It can be understood that nitrogen embrittlement risk data is a comprehensive quantitative value reflecting the degree of nitrogen embrittlement risk from each set of multi-parameter real-time data. This comprehensive quantitative value is calculated by normalizing each data point, multiplying it by its corresponding configurable weight parameter, and then summing the results.

[0048] Specifically, the thermal deviation data Z, abnormal pattern data S, and sensor confidence data C are normalized to the [0,1] interval. The normalization method uses saturated linear mapping: for thermal deviation data, when the Z value exceeds the preset upper limit Zmax (e.g., 5), it is mapped to 1; when it is below the preset lower limit Zmin (e.g., 0), it is mapped to 0; and intermediate values ​​are mapped linearly. For abnormal pattern data and sensor confidence data, since their original values ​​are already within the [0,1] interval, no additional normalization is required.

[0049] After normalization, the nitrogen plugging risk data R is calculated using the following formula: R = w1 × Z + w2 × S + w3 × C + w4 × P, where w1, w2, and w3 are the configurable weight parameters corresponding to thermal deviation data, abnormal mode data, and sensor reliability data, respectively; P is the numerical representation of the process safety level data (level 0 corresponds to 0, level 1 corresponds to 0.33, level 2 corresponds to 0.67, and level 3 corresponds to 1.0).

[0050] The weights can be optimized and adjusted based on the specific design characteristics and operational experience of the cold box. Each weight parameter satisfies the preset constraint: w1+w2+w3+w4=1, and the value range of each weight parameter is [0,1]. In a specific embodiment, the preset weight values ​​can be: w1=0.35, w2=0.25, w3=0.15, w4=0.25. Thermal deviation data has the highest weight because it is the most direct data to measure the deviation of the current data from the normal range; process safety level data has the second highest weight because it is directly related to equipment safety and product quality; abnormal pattern data is used to distinguish between known common disturbances and rare anomalies; sensor reliability data has the lowest weight because it mainly plays an auxiliary role in weakening the impact. For data that is not used (e.g., when the system configuration only uses part of the data), the corresponding weight parameter is zero.

[0051] The optimization and adjustment method for weight parameters can be as follows: Based on historical operating data of the cold box and corresponding nitrogen plugging event records, regression analysis is used to optimize the weights with the occurrence of nitrogen plugging events as the label, so as to maximize the predictive ability of operating condition data for nitrogen plugging risk. Specific optimization processes can employ grid search or gradient descent methods.

[0052] In one possible implementation, the optimization method for configurable weight parameters w1, w2, w3, and w4 includes: (1) collecting historical operating datasets of the target cold box, including real-time multi-parameter data of the time series and corresponding labels indicating whether a nitrogen blockage event has occurred (which can be confirmed through post-event analysis); (2) using whether the nitrogen blockage risk data R exceeds the preset warning line as the prediction target, dividing the historical data into training sets; (3) using a grid search method, traversing different weight combinations under the constraints of w1+w2+w3+w4=1, wi∈[0,1]; (4) for each weight set, calculating the risk score R of all data points in the training set, and evaluating its precision, recall, or F1 score in predicting nitrogen blockage events; (5) selecting the combination that optimizes the evaluation indicators as the final weight parameters. The optimization method for calculating weights a1, a2, a3, and a4 for operating condition data is similar, but the evaluation target can be adjusted to maximize the correlation between operating condition data o and historical operating condition classification labels.

[0053] It should be noted that this application uses four-dimensional calculation data (thermal deviation, abnormal modes, sensor reliability, and process safety level) for risk assessment, rather than using only thermal deviation data. This is because, in actual LNG cold box operation, single-dimensional data may be subject to multiple interfering factors. For example, an increase in approach temperature may be caused by various factors such as load adjustments, ambient temperature changes, or sensor drift, and thermal deviation data alone cannot distinguish these factors. This application improves the accuracy of nitrogen plugging signal identification in environments with multiple interfering factors by introducing abnormal mode data (distinguishing between known and unknown abnormal modes) and sensor reliability data (eliminating false signals caused by sensor malfunctions). Furthermore, this application uses dynamic thresholds instead of fixed thresholds, which can adapt to the differences in the normal fluctuation range of parameters under different operating conditions of the cold box, further reducing the risk of false alarms and missed alarms.

[0054] S400 acquires the operating data of the cold box and calculates the current operating condition data.

[0055] The normal fluctuation range of various process parameters in a cold box varies significantly under different operating conditions (such as different load levels, different usage stages, and different feeding conditions). For example, at 100% load operation, the normal fluctuation range of the approaching temperature may reach 1-2°C, while at 50% load operation, the normal fluctuation range may only be 0.3-0.5°C. If a fixed threshold is used to judge anomalies, normal fluctuations are easily misjudged as anomalies at high loads (false alarms), while true anomaly signals may be missed at low loads (false negatives). Therefore, it is necessary to perceive the current operating status of the cold box through operating condition data and dynamically adjust the entry thresholds of each level of response accordingly.

[0056] The operating condition data of the cold box may include: (1) the ratio of the current production load rate to the designed throughput (i.e., relative load data): the current production load rate can be obtained in real time through the production management log, and divided by the designed throughput of the cold box to obtain the relative load data. This relative load data is used to reflect the intensity of the heat load borne by the cold box. The higher the ratio, the more the cold box is operating under high load, the greater the fluctuation of normal parameters, and the higher the fault tolerance requirement for abnormal judgment.

[0057] (2) Cold box approach temperature deviation coefficient (heat transfer attenuation data): The cold box approach temperature deviation coefficient is calculated by comparing the relative deviation between the current actual approach temperature and the design approach temperature. The coefficient is calculated as: |current actual approach temperature - design approach temperature| ÷ design approach temperature. This heat transfer attenuation data reflects the degree of attenuation of heat transfer performance relative to the design state.

[0058] (3) Ratio of time since last temperature regeneration to designed regeneration cycle (regeneration cycle data): The regeneration cycle data is obtained by dividing the time since the last temperature regeneration obtained through the maintenance management system by the designed regeneration cycle. This regeneration cycle data is used to reflect the cumulative usage since the last full recovery. The larger the ratio, the longer the cold box has been since the last regeneration, and the higher the risk of nitrogen accumulation.

[0059] (4) Feed nitrogen mole fraction change rate (feed fluctuation data): The relative change rate of nitrogen content over the past 24 hours is calculated using data measured at the feed end of the gas chromatograph. Specifically, feed fluctuation data = |xN2(t) - xN2(t-24h)| ÷ xN2(t-24h), where xN2(t) is the current feed nitrogen mole fraction, and xN2(t-24h) is the feed nitrogen mole fraction 24 hours ago. This feed fluctuation data reflects the degree of fluctuation in nitrogen content in the feed gas. When the unit is not equipped with a feed chromatograph, the feed methane mole fraction change rate can be used as an alternative indirect data. For example, if the feed nitrogen mole fraction was 1.5% 24 hours ago and is currently 1.8%, then the feed fluctuation data = |(1.8-1.5) ÷ 1.5| = 0.2.

[0060] The above four data points are weighted and summed according to preset weights to obtain the current operating condition data. The specific calculation formula is: Current operating condition data = Relative load data × a1 + Heat exchange attenuation data × a2 + Regeneration cycle data × a3 + Feed fluctuation data × a4. Among them, a1, a2, a3, and a4 can be optimized and adjusted according to the specific design characteristics and operating experience of the cold box, a1 + a2 + a3 + a4 = 1, and each value ranges from [0,1]. For example, a1 = 0.30, a2 = 0.25, a3 = 0.30, a4 = 0.15.

[0061] It should be noted that although both the heat transfer attenuation data in S400 and the thermal deviation data in S200 involve approaching temperatures, their functions and comparison benchmarks differ: the thermal deviation data uses the historical average within a sliding time window as a benchmark to measure the short-term deviation of current data from recent normal operating conditions, serving as a direct input factor for risk assessment; the heat transfer attenuation data uses the cold box design approaching temperature as a benchmark to measure the long-term attenuation of current heat transfer performance relative to the design state, serving as an input factor for dynamic threshold adjustment in operational condition perception. The two data points respectively reflect information from two different dimensions: short-term anomalies and long-term attenuation.

[0062] In one possible implementation, the operating data includes the ratio of the current production load rate to the designed throughput, the cold box approach temperature deviation coefficient, the ratio of the time since the last temperature-state regeneration to the designed regeneration cycle, and the change rate of the feed nitrogen mole fraction. S400, acquire the cold box operating data and calculate the current operating condition data, including: S410: Obtain the ratio of the current production load rate to the designed throughput and calculate the relative load data.

[0063] It is understandable that the load relative data is used to reflect the intensity of the heat load borne by the cold box.

[0064] Specifically, the current production load rate can be obtained in real time through the production management log, and then divided by the cold box's designed capacity to obtain the relative load data.

[0065] S420, obtains the cold box approach temperature deviation coefficient, and calculates heat transfer attenuation data.

[0066] It is understandable that heat transfer attenuation data is used to reflect the degree of degradation of heat transfer performance relative to the design state.

[0067] Specifically, heat transfer attenuation data can be calculated by comparing the relative deviation between the current actual approach temperature and the design approach temperature (i.e., the cold box approach temperature deviation coefficient). The current actual approach temperature can be the difference between the outlet temperature of the higher-temperature fluid and the inlet temperature of the lower-temperature fluid within the heat exchange section during actual operation of the cold box. The design approach temperature can be the approach temperature value that each heat exchange section should reach under design conditions, as specified in the cold box's design specifications. It is the target value of heat transfer efficiency determined through theoretical calculations or performance simulations based on the cold box's structural design parameters (such as fin area, number of channels, fluid flow distribution, etc.) and design operating conditions (such as design throughput, design feed composition, etc.).

[0068] S430: Obtain the ratio of the time since the last temperature regeneration to the designed regeneration cycle, and calculate the regeneration cycle data.

[0069] It can be understood that the last warm-state regeneration time can be the cumulative operating time of the cold box from the completion of the last warm-state regeneration operation to the current moment. Warm-state regeneration can be a restorative operation that raises the cold box temperature from the deep cryogenic operating temperature (e.g., -162°C) to room temperature, causing the gaseous nitrogen accumulated at the cold end of the heat exchange channel to evaporate and be discharged. The design regeneration cycle can be the longest recommended operating time between two consecutive warm-state regenerations specified in the cold box design specifications. It is an operating period with controllable nitrogen accumulation risk determined based on the cold box design parameters and typical operating conditions. The regeneration cycle data is used to reflect the cumulative usage since the last full recovery.

[0070] Specifically, the regeneration cycle data can be calculated by obtaining the time since the last temperature regeneration through the maintenance management system, dividing it by the designed regeneration cycle.

[0071] S440, obtain the rate of change of nitrogen mole fraction in the feed and calculate the feed fluctuation data.

[0072] It can be understood that the feed nitrogen mole fraction can be the molar proportion of nitrogen in the natural gas feedstock of the cold box. The feed nitrogen mole fraction change rate can be the relative change of the feed nitrogen mole fraction relative to the initial value over a period of time (e.g., the last 24 hours). The calculation formula is: subtract the nitrogen mole fraction 24 hours ago from the nitrogen mole fraction at the current moment, divide by the nitrogen mole fraction 24 hours ago, and take the absolute value.

[0073] Specifically, the relative change rate of nitrogen content over the past 24 hours can be calculated using data measured at the feed end of a gas chromatograph.

[0074] S450 calculates operating condition data by weighting relative load data, heat exchange attenuation data, regeneration cycle data, and feed fluctuation data according to preset weights.

[0075] Specifically, the current operating condition data can be obtained by weighting and summing the relative load data, heat exchange attenuation data, regeneration cycle data, and feed fluctuation data according to preset weights. The specific calculation formula is: Current operating condition data = Relative load data × a1 + Heat exchange attenuation data × a2 + Regeneration cycle data × a3 + Feed fluctuation data × a4. Among them, a1, a2, a3, and a4 can be optimized and adjusted according to the specific design characteristics and operating experience of the cold box. For example, a1 = 0.30, a2 = 0.25, a3 = 0.30, and a4 = 0.15.

[0076] S500 calculates emergency response thresholds and enhanced monitoring thresholds based on operational condition data.

[0077] It is understandable that the emergency response threshold and the enhanced monitoring threshold serve as the dividing lines for categorizing nitrogen plug risk data into different response zones. Both thresholds dynamically change with the operating condition data. The design logic is as follows: when the operating condition data is high (i.e., the cold box is under high load and high stress), the fluctuation range of normal parameters is large, requiring the threshold to be increased to widen the safe distance between the normal amplitude and the threshold, avoiding misjudging normal fluctuations as abnormalities; when the operating condition data is low (i.e., the cold box is under low load and stable state), the normal fluctuation range is small, and the threshold should be decreased to improve the detection sensitivity of weak abnormal signals.

[0078] The emergency response threshold Ta can be calculated using a growth function: Ta = min(Tb × exp(β × o), Rmax), where Tb is the basic emergency response threshold (the basic emergency response threshold Tb can be determined by: based on the historical operating data of the cold box, statistically analyzing the 95th percentile value of the nitrogen plug risk score during normal operation, taking this value as the initial value of Tb, and fine-tuning it according to the actual operating effect, for example, it can be set to 0.4), β is the sensitivity parameter (which can be set according to the design manual or the value input by the operator, for example, it can be set to 2), o is the operating condition data, Rmax is the preset upper limit of the nitrogen plug risk data (for example, it can be set according to actual needs, such as 1), and min(·) is the minimum value calculation.

[0079] The meaning of this function is as follows: Under low load conditions (o close to 0), exp(β×o) is close to 1, the threshold is close to the baseline level Tb, and the system operates with standard sensitivity; when the load increases, exp(β×o) increases exponentially, the threshold rises rapidly, the admission conditions are tightened, and normal amplitude under high load is prevented from being misjudged as a nitrogen blockage event. The minimum value is calculated to ensure that the threshold does not grow indefinitely, with an upper limit of Rmax.

[0080] The enhanced monitoring threshold Tc can be calculated using a polynomial function: Tc = Td × (1 + o^n), where Td is the base threshold for enhanced monitoring (which can be set according to the design manual or a value input by the operator, such as 0.25), n is the polynomial degree (it can be a quadratic or cubic polynomial, and the appropriate polynomial degree can be selected according to the actual cold box characteristics and threshold sensitivity requirements), and o is the operating condition data. Both thresholds are determined by their respective base thresholds (Tb and Td), which together determine the initial level and response rate.

[0081] In one possible implementation, the method for determining the basic threshold Tb for emergency response can be: statistically analyzing all calculated nitrogen blockage risk score data obtained during the normal operation of the cold box over the past year (with no confirmed nitrogen blockage events), and taking its 95th percentile value as the initial value of Tb. The method for determining the sensitivity parameter β and the polynomial degree n can be: on a historical data simulation platform, optimizing β and n with the objective function of minimizing the weighted sum of the false alarm rate and the false negative rate. The method for constructing the feature pattern library can include: (1) collecting historically confirmed nitrogen blockage event cases and corresponding multi-parameter data windows; (2) extracting a 12-dimensional feature vector for each case; (3) using the K-means clustering algorithm (preset K=2, corresponding to progressive and abrupt types) to cluster all case feature vectors; (4) storing the final cluster centers as standard feature vectors of each pattern in the library. The update method is as follows: once a new nitrogen slugging event is finally confirmed (e.g., confirmed by checking after temperature regeneration), its feature vector is added to the cluster sample set of the corresponding pattern, and K-means clustering is run again to update the cluster centers.

[0082] After calculating the two thresholds, threshold consistency correction is performed: when the calculated emergency response threshold is less than the enhanced monitoring threshold (this may occur under certain extreme operating condition parameter combinations), the emergency response threshold is adjusted to the enhanced monitoring threshold, i.e., Ta = max(Ta, Tc). For example, when the operating condition data o = 0.5, the polynomial degree n = 2, and the enhanced monitoring base threshold Td = 0.25, the enhanced monitoring threshold Tc = 0.25 × (1 + 0.5²) = 0.3125; the emergency response threshold Ta = min(0.4 × exp(2 × 0.5), 1.0) = min(0.4 × e, 1.0) ≈ min(1.087, 1.0) = 1.0. After correction, Ta = 1.0 ≥ Tc = 0.3125, which meets the consistency requirement.

[0083] In one possible implementation, S500 calculates emergency response thresholds and enhanced monitoring thresholds based on operational condition data, including: Using operational condition data as independent variables, emergency response thresholds and enhanced monitoring thresholds are calculated separately using functions, ensuring that the larger the operational condition data, the larger at least one of the emergency response thresholds and enhanced monitoring thresholds becomes. Specifically, the emergency response threshold is calculated using a growth function, causing it to increase rapidly with increasing operational condition data, and a minimum value operation is used to limit the emergency response threshold from exceeding a preset upper limit. The enhanced monitoring threshold is calculated using a polynomial growth function, causing it to increase smoothly with increasing operational condition data. The initial levels and response rates of both the emergency response threshold and the enhanced monitoring threshold are determined by their respective base thresholds.

[0084] Specifically, the emergency response threshold Ta can be calculated using a growth function: Ta = min(Tb × exp(β × o), Rmax). The enhanced monitoring threshold Tc can be calculated using a polynomial function: Tc = Td × (1 + o^n).

[0085] S600 classifies data based on comparisons between nitrogen plug risk data and emergency response and enhanced monitoring thresholds.

[0086] Specifically, each set of data can be routed to different treatment paths for tiered processing based on the comparison results between nitrogen plugging risk data and emergency response thresholds, enhanced monitoring thresholds, and process safety levels.

[0087] With this configuration, this application achieves improved accuracy in nitrogen plugging early warning through the synergistic effect of the following technical features: First, by fusing thermal deviation data (Z), abnormal pattern data (S), sensor reliability data (C), and process safety level data (P) for calculation, it can comprehensively assess risk from four orthogonal dimensions: data deviation, pattern rarity, sensor reliability, and direct safety status. Compared to single-parameter detection, it can effectively distinguish between single-parameter fluctuations caused by load adjustments, environmental changes, or sensor drift and true multi-parameter synergistic deterioration precursors of nitrogen plugging, thereby improving the accuracy of early signal identification. Second, both the emergency response threshold and the enhanced monitoring threshold are dynamically correlated with the operating condition data o. When the operating condition o increases (e.g., high load), Ta and Tc are simultaneously increased, mathematically expanding the safe zone of the risk score R and avoiding misjudging small fluctuations as abnormalities during normal high-load fluctuation periods (reducing false alarms); when the operating condition o decreases, the threshold is lowered, enhancing the system's sensitivity to weak abnormal signals under low load (reducing false alarms). Finally, the tiered processing logic (Level 3: direct action; Level 2: action + waiting period; Level 1: trend analysis) ensures rapid response to high-level risks and prudent assessment of medium- and low-level risks, avoiding abuse or delay in handling actions. This multi-dimensional, dynamic, and tiered design works together to solve the false alarm and missed alarm problems caused by the single-parameter fixed threshold method in existing technologies, which cannot adapt to changes in operating conditions.

[0088] In one possible implementation, S600 performs tiered processing based on comparisons between nitrogen plug risk data and emergency response thresholds and enhanced monitoring thresholds, including: S610: If the process safety level is the highest preset hazard level, the nitrogen plugging risk data will not be compared with the threshold, and the emergency control action will be executed directly.

[0089] For example, when the process safety level is the preset highest hazard level (level 3), there is no need to compare the nitrogen plugging risk data with the threshold, and the emergency control action (i.e., the first-level adjustment action) is executed directly.

[0090] Specifically, emergency control actions may include: (x) adjusting the superheat setting or opening of the cold-end expansion valve of the cold box to reduce the refrigerant output at the cold end, thereby appropriately raising the minimum temperature of the cold-end heat exchange channel and reducing the driving force for further nitrogen accumulation in this area. Specifically, a superheat setting adjustment command can be sent to the cold-end expansion valve, with the adjustment range being an increase of 2-5°C (e.g., from 8°C to 12°C), or a corresponding opening adjustment command can be sent to the cold-end expansion valve (e.g., increasing the opening by 3%-5%).

[0091] (y) Modify the precooling temperature setpoint of the raw material gas to increase the precooling temperature and reduce the heat load at the cold end of the cold box. Specifically, a new temperature setpoint command can be sent to the temperature controller of the raw material gas precooler, and the adjustment range can be to increase the precooling temperature by 3-8°C. After the precooling temperature is increased, the temperature of the raw material gas entering the cold end of the cold box will increase accordingly, reducing the cooling demand at the cold end and helping to alleviate the tendency of nitrogen accumulation.

[0092] (z) By adjusting the opening of the refrigerant bypass valves at the intermediate and cold ends, a portion of the refrigerant flow can be bypassed from the cold end to the intermediate section. This reduces the refrigeration depth at the cold end and increases the refrigeration capacity at the intermediate section, thereby alleviating the nitrogen accumulation trend at the cold end and maintaining the heat exchange efficiency of the intermediate section. Specifically, an opening command can be sent to the intermediate-cold end refrigerant bypass valve to increase the bypass valve opening by 10%-20% (e.g., from 15% to 30%). The above three adjustment actions are sent to the corresponding actuators in a preset priority order: the default priority action (x) is to adjust the superheat setting of the cold end expansion valve; after execution, the first observation period begins (e.g., 10-30 minutes, which can be configured according to the thermal inertia of the cold box). If the nitrogen plugging risk data does not fall below the emergency response threshold, then action (y) is executed to modify the raw material gas precooling temperature setting; if the situation is still not alleviated after another observation, then action (z) is executed to adjust the refrigerant bypass valve opening. Once the nitrogen plug risk data falls back to a safe range after any action is performed, subsequent actions will not be executed to avoid excessive reduction of the cooling capacity at the cold end, which could lead to drastic fluctuations in the thermal balance of the cold box. In another implementation, the three adjustment actions can also be sent to the corresponding actuators simultaneously, but the adjustment range of each action will be reduced accordingly (e.g., each will be 50%-70% of the default range).

[0093] S620 If the process safety level is not the highest hazard level and the nitrogen plug risk data is not lower than the emergency response threshold, an active adjustment action is executed and a waiting period is entered. During the waiting period, the current level of control action is not repeated, but the channel to upgrade the response level is retained.

[0094] For example, when the process safety level does not reach the highest level (Level 3) and the nitrogen plugging risk data is not lower than the emergency response threshold, an active adjustment action (i.e., Level 2 adjustment action) can be performed.

[0095] Specifically, the active adjustment action can be: adjusting the superheat setting value of the cold end expansion valve (e.g., increasing it by 2°C, a smaller adjustment range than the emergency control action) and the precooling temperature setting value (e.g., increasing it by 3°C). The adjustment range of the active adjustment action is smaller than the adjustment range of the emergency control action.

[0096] After performing the above proactive adjustment actions, the system enters an action confirmation waiting period. During this waiting period, the following steps are executed: (1) Temporarily suspend the execution of the current level of control actions to avoid system oscillation. It is understood that the cold box is a system with a large thermal inertia (the thermal time constant is usually tens of minutes to several hours), and it takes a certain amount of time for the effect of the adjustment action to be reflected in the process parameters. If the same adjustment action is repeated before the effect is manifested, it may lead to over-adjustment and system oscillation. Therefore, the system suspends the execution of the same level of control actions during the waiting period.

[0097] (2) A channel for escalating the response level is retained. Specifically, the system continuously monitors process safety level data and nitrogen plug risk data during the waiting period. If the process safety level is upgraded to a higher level (e.g., from level 1 to level 2 or level 3) during the waiting period, or if the nitrogen plug risk data continues to rise and exceeds the threshold of the next higher level (i.e., the higher sensitivity boundary of the emergency response threshold), the system immediately lifts the waiting period restriction and executes a higher-level response. For example, if the process safety level is upgraded to level 3 during the waiting period, the system immediately executes emergency control actions.

[0098] The waiting period can be dynamically calculated based on the cold box's heat capacity and historical response characteristics. In one specific embodiment, the waiting period is calculated as follows: based on the ratio of the effective heat capacity of the cold box heat exchanger (which can be estimated by multiplying the total mass of the cold box's aluminum fins by the specific heat capacity of aluminum) to the typical heat exchange power under the current operating conditions (which can be estimated by multiplying the current production load rate by the design unit load heat exchange power).

[0099] After the waiting period, the nitrogen suffocation risk data can be reassessed. If the nitrogen suffocation risk data has fallen below the enhanced monitoring threshold, the adjustment action is deemed effective, and normal monitoring is restored; if the nitrogen suffocation risk data remains high, the response level is further upgraded.

[0100] S630: If the nitrogen plug risk data is below the emergency response threshold but not below the enhanced monitoring threshold, then aggregation trend analysis is performed. If the nitrogen plug risk data is below the enhanced monitoring threshold, no processing is performed. The aggregation trend analysis is based on the statistical characteristics of nitrogen plug risk data over at least several consecutive periods.

[0101] For example, when nitrogen plug risk data is below the emergency response threshold but not below the enhanced monitoring threshold, aggregated trend analysis can be performed. It is understood that while this level of data does not warrant immediate intervention, it indicates certain abnormal signals. Further statistical analysis can be used to make a definitive judgment. That is, for data at low to medium risk levels, a single data point is insufficient for a definitive assessment; multiple data points need to be accumulated for overall statistical analysis to improve the ability to identify progressive nitrogen plugs while avoiding false alarms due to accidental fluctuations in a single data point.

[0102] Specifically, for Level 3 (nitrogen plug risk data is below the emergency response threshold but not below the enhanced monitoring threshold), aggregated statistical characteristics are calculated for multiple data points at this level, including the mean change rate, standard deviation change rate, and cumulative anomaly count for each key parameter (approach temperature, channel differential pressure) within the batch.

[0103] When calculating the mean change rate, considering the frequent load adjustments of LNG units, changes in production load can cause normal fluctuations in parameters such as temperature and differential pressure. These changes may mask or disguise as nitrogen plugging signals. Therefore, the following methods can be used to eliminate the interference of load fluctuations on parameter trends: Using changes in production load as one of the independent variables in a multiple linear regression model, and the parameter values ​​as the dependent variable, regression analysis is performed. Figure 3 The specific regression model is: y(t) = β0 + β1 × t + β2 × L(t) + ε(t), where y(t) is the parameter value at time t (e.g., approaching temperature), t is the time variable, L(t) is the production load rate at time t, β0 is the intercept, β1 is the time coefficient (used to reflect the trend of parameter change over time), β2 is the load coefficient (reflecting the impact of load change on the parameter), and ε(t) is the residual.

[0104] The regression coefficient β2 reflects the impact of load changes on the parameters, while the persistent positive trend in the time coefficient β1 and the residual ε(t) indicates a parameter increase that cannot be explained by load changes, i.e., a possible signal of nitrogen plugging development.

[0105] When the number of data entries in a batch is less than the preset minimum sample size (e.g., 20 entries), load compensation regression will not be performed temporarily; only the original trend slope will be calculated.

[0106] The development trend of nitrogen plugging is assessed based on residual trends (i.e., the trend of parameter changes over time). A deteriorating trend is defined as when all of the following conditions are met: Condition 1: For at least two consecutive complete processing cycles, the slope of the linear regression of the residual trend is greater than the preset minimum rate of change threshold (the minimum rate of change threshold can be determined based on data under normal conditions, such as 0.05). Condition 2: The cumulative abnormal count (the percentage of data entries whose thermal deviation exceeds the preset sub-threshold within the processing cycle) exceeds the preset percentage threshold (which can be determined based on the percentage of data entries with normal thermal deviation within the processing cycle, such as 30%).

[0107] If the situation is deemed to have worsened, the response level will be upgraded from batch monitoring to proactive adjustment action, and the aforementioned second-level proactive adjustment action will be executed.

[0108] It should be noted that the load compensation regression method is used instead of directly calculating the trend of the risk score R because: in LNG cold box operation, short-term fluctuations in production load will cause significant changes in the risk score R. These changes are normal adjustments rather than a manifestation of nitrogen plugging deterioration. If a trend analysis is performed directly on R, load fluctuations will seriously interfere with or even mask the trend of the true nitrogen plugging signal. This application uses the load change L(t) as the independent variable of the regression model, which can separate the fluctuation part explained by load change from the original parameter y(t), thereby extracting the residual trend ε(t) that cannot be explained by load and is more likely related to nitrogen plugging. Its time coefficient β1 can more purely reflect the gradual development trend of nitrogen plugging and improve the accuracy of trend judgment in the low-to-medium risk range.

[0109] In one possible implementation, step S630, if the nitrogen plugging risk data is below the emergency response threshold and not below the enhanced monitoring threshold, then aggregation trend analysis processing is performed, including: If the nitrogen plug risk data is below the emergency response threshold but not below the enhanced monitoring threshold, the following actions are taken: When calculating the mean change rate, the change in production load is used as the independent variable in the multiple linear regression model to eliminate the interference of load fluctuations on the parameter trend. The nitrogen plug development trend is assessed based on the residual trend after eliminating the interference. When the slope of the residual trend exceeds the preset minimum rate of change threshold and the cumulative count of anomalies exceeds the preset proportion threshold for at least two consecutive processing cycles, the nitrogen plug is determined to be deteriorating, and the response level is upgraded to proactive adjustment.

[0110] Specifically, the change in production load can be used as one of the independent variables in a multiple linear regression model, with the parameter value as the dependent variable for regression analysis. The specific regression model is: y(t) = β0 + β1 × t + β2 × L(t) + ε(t), where y(t) is the parameter value at time t (e.g., approaching temperature), t is the time variable, L(t) is the production load rate at time t, β0 is the intercept, β1 is the time coefficient (used to reflect the parameter's trend over time), β2 is the load coefficient (reflecting the impact of load changes on the parameter), and ε(t) is the residual, representing the residual fluctuation of the parameter after removing the time trend and load influence.

[0111] When the slope of the residual trend exceeds the preset minimum rate of change threshold and the cumulative count of anomalies exceeds the preset proportion threshold for at least two consecutive processing cycles, it is determined that the nitrogen plugging is showing a deteriorating trend, and the response level is upgraded to an active adjustment action.

[0112] In one possible implementation, S300 determines nitrogen plugging risk data based on the determined calculation data, including: After normalizing the thermal deviation data, abnormal mode data, and sensor reliability data, each data point is multiplied by its corresponding configurable weight parameter and summed. This summation is then added to the product of the process safety level data and its corresponding configurable weight parameter to obtain the nitrogen plug risk data. Specifically, the sensor reliability data is used to mitigate the impact of sensor data with large consistency deviations on nitrogen plug risk assessment: for target sensors whose measurement consistency deviation with adjacent sensors within the same heat exchange section exceeds a preset threshold, the target sensor's weight coefficient in the scoring calculation is reduced. For sensors with normal consistency, the standard weight coefficient remains unchanged.

[0113] In one possible implementation, emergency control actions include: Automatically perform at least one of the following adjustment actions: adjust the superheat setting or opening of the cold end expansion valve of the cold box to reduce the cold end cooling output and cause the cold end temperature to rise; modify the raw material gas precooling temperature setting; and bypass part of the refrigerant flow from the cold end to the intermediate section by adjusting the opening of the refrigerant bypass valves of the intermediate section and the cold end.

[0114] The system generates a suggested load reduction plan and displays it on the confirmation screen. After receiving confirmation from the operator, the load adjustment is executed.

[0115] In one possible implementation, the current level of control action is not repeated during the waiting period, while a channel for escalating the response level is retained, including: During the waiting period, the current level of control actions will not be repeated. When the process safety level data is upgraded to the target level, or the nitrogen plugging risk data continues to rise and exceeds the threshold of the next higher level, the waiting period restriction will be lifted immediately and a higher level response will be executed. The duration of the waiting period is dynamically calculated based on the cold box heat capacity and historical response characteristics.

[0116] In one possible implementation, please refer to Figure 4 The nitrogen plugging early warning and automatic handling control methods also include: If the control action fails to execute on the first attempt, the action will be retried a preset number of times, and a pending execution flag will be set.

[0117] After the system returns to normal, check if there are any pending execution flags. If there are, re-execute the control actions that were not successfully executed before.

[0118] Specifically, when a control action command sent to the actuator fails to execute (e.g., DCS communication timeout, actuator feedback error, etc.), it retryes a preset number of times (e.g., 3 times), with a 10-second interval between each retry. If the retry still fails, all control outputs are maintained at their last valid value before the failure and marked as pending execution. After the system recovers, it automatically checks the pending execution mark and re-executes the previously unsuccessful control actions.

[0119] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0120] Corresponding to the nitrogen plug warning and automatic handling control method described in the above embodiments, this application also provides a nitrogen plug warning and automatic handling control system, in which each unit can implement each step of the nitrogen plug warning and automatic handling control method. Figure 5 A structural block diagram of the nitrogen plugging early warning and automatic handling control system provided in an embodiment of this application is shown. For ease of explanation, only the parts related to the embodiment of this application are shown.

[0121] Reference Figure 5 The nitrogen plugging early warning and automatic handling control system includes: The acquisition unit is used to acquire real-time data of multiple parameters at various measurement points in the cold box.

[0122] The analysis unit is used to determine the calculated data corresponding to the multi-parameter real-time data. This calculated data includes thermal deviation data, anomaly pattern data, sensor reliability data, and process safety level data. Thermal deviation data reflects the degree of difference between the multi-parameter real-time data and the statistical characteristics of historical time windows. Anomaly pattern data reflects the rarity of the anomaly pattern presented by the current data combination within historical periods. Sensor reliability data reflects the data reliability weight of the current sensor based on the consistency of measurements from adjacent sensors within the same heat exchange section. Process safety level data reflects the nitrogen plug hazard level corresponding to the multi-parameter real-time data.

[0123] The calculation unit is used to determine nitrogen plugging risk data based on the determined calculation data.

[0124] The data acquisition unit is used to acquire the operating data of the cold box and calculate the current operating condition data.

[0125] The generation unit is used to calculate emergency response thresholds and enhanced monitoring thresholds based on operational condition data.

[0126] The processing unit is used to perform tiered processing based on the comparison results between nitrogen plug risk data and emergency response thresholds and enhanced monitoring thresholds.

[0127] It should be noted that the information interaction and execution process between the above systems / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0128] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units is merely an example. In practical applications, the above functions can be assigned to different functional units as needed, that is, the internal structure of the system can be divided into different functional units to complete all or part of the functions described above. The functional units in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0129] This application also provides a nitrogen plugging early warning device. Figure 6 This is a schematic diagram of the structure of a nitrogen plugging warning device provided in one embodiment of this application. Figure 6 As shown, the nitrogen plugging warning device 6 of this embodiment includes: at least one processor 60 ( Figure 6 Only one is shown in the image), at least one memory 61 ( Figure 6 (Only one is shown in the image) and a computer program 62 stored in the at least one memory 61 and executable on the at least one processor 60. When the processor 60 executes the computer program 62, it causes the nitrogen plug warning device 6 to implement the steps in any of the above-described nitrogen plug warning and automatic handling control method embodiments, or causes the nitrogen plug warning device 6 to implement the functions of each unit in the above-described system embodiments.

[0130] For example, the computer program 62 may be divided into one or more units, which are stored in the memory 61 and executed by the processor 60 to complete this application. The one or more units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 62 in the nitrogen plugging warning device 6.

[0131] The nitrogen plugging early warning device 6 may include a data acquisition device (e.g., temperature sensors, pressure sensors, etc. installed on the LNG pipeline for real-time acquisition of temperature and pressure), a data processing and judgment device (e.g., a programmable logic controller (PLC) or a dedicated computing module of a distributed control system (DCS) for receiving data acquired by the data acquisition device and performing real-time calculations), and an early warning and protection execution device (e.g., a control room audible and visual alarm, a field alarm bell, and an interlocking shut-off valve or pump speed regulation circuit for control or early warning based on the calculation results of the data processing and judgment device). The nitrogen plugging early warning device 6 may include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art will understand that... Figure 6 This is merely an example of the nitrogen plug warning device 6 and does not constitute a limitation on the nitrogen plug warning device 6. It may include more or fewer components than shown, or combine certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0132] The processor 60 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0133] In some embodiments, the memory 61 can be an internal storage unit of the nitrogen choke warning device 6, such as a hard drive or memory of the nitrogen choke warning device 6. In other embodiments, the memory 61 can be an external storage device of the nitrogen choke warning device 6, such as a plug-in hard drive, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the nitrogen choke warning device 6. Furthermore, the memory 61 can include both internal and external storage units of the nitrogen choke warning device 6. The memory 61 is used to store operating systems, applications, bootloaders, data, and other programs, such as the program code of computer programs. The memory 61 can also be used to temporarily store data that has been output or will be output.

[0134] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0135] This application provides a computer program product that, when run on a nitrogen plugging warning device, enables the nitrogen plugging warning device to implement the steps in any of the above-described method embodiments.

[0136] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to the nitrogen rush warning device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, such as a USB flash drive, a portable hard drive, a magnetic disk, or an optical disk.

[0137] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0138] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0139] In the embodiments provided in this application, it should be understood that the disclosed nitrogen plug warning and automatic handling control system / nitrogen plug warning device and nitrogen plug warning and automatic handling control method can be implemented in other ways. For example, the nitrogen plug warning and automatic handling control system / nitrogen plug warning device embodiments described above are merely illustrative. For instance, the division of units is merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0140] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0141] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A nitrogen plugging early warning and automatic handling control method, characterized in that, The method includes: Acquire real-time data of multiple parameters at various measurement points in the cold box; The calculation data corresponding to the multi-parameter real-time data is determined; wherein, the calculation data corresponding to the multi-parameter real-time data includes thermal deviation data, abnormal pattern data, sensor reliability data, and process safety level data; the thermal deviation data is used to reflect the degree of difference between the multi-parameter real-time data and the statistical characteristics of historical time windows; the abnormal pattern data is used to reflect the rarity of the abnormal pattern presented by the current data combination in historical periods; the sensor reliability data is used to reflect the data reliability weight of the current sensor based on the consistency of measurements by adjacent sensors within the same heat exchange section; the process safety level data is used to reflect the nitrogen plugging hazard level corresponding to the multi-parameter real-time data. Nitrogen plug risk data are determined based on the calculated data; Acquire the operating data of the cold box and calculate the current operating condition data; Calculate emergency response thresholds and enhanced monitoring thresholds based on the aforementioned operational data; The data is classified based on the comparison between the nitrogen plugging risk data and the emergency response threshold and the enhanced monitoring threshold.

2. The nitrogen plugging early warning and automatic handling control method as described in claim 1, characterized in that, The operational data includes the ratio of the current production load rate to the designed throughput, the cold box approach temperature deviation coefficient, the ratio of the time since the last temperature-state regeneration to the designed regeneration cycle, and the rate of change of the feed nitrogen mole fraction. The process of acquiring the operating data of the cold box and calculating the current operating condition data includes: Obtain the ratio of the current production load rate to the designed processing capacity, and calculate the relative load data; Obtain the temperature deviation coefficient of the cold box and calculate the heat transfer attenuation data; Obtain the ratio of the time since the last temperature regeneration to the designed regeneration cycle, and calculate the regeneration cycle data; Obtain the rate of change of nitrogen mole fraction in the feed and calculate the feed fluctuation data; The operating condition data are obtained by weighting the relative load data, the heat exchange attenuation data, the regeneration cycle data, and the feed fluctuation data according to preset weights.

3. The nitrogen plugging early warning and automatic handling control method as described in claim 1, characterized in that, The calculation of emergency response thresholds and enhanced monitoring thresholds based on the operational condition data includes: Using the operating condition data as independent variables, emergency response thresholds and enhanced monitoring thresholds are calculated using functions, such that the larger the operating condition data, the larger at least one of the emergency response thresholds and the enhanced monitoring thresholds. Specifically, the emergency response threshold is calculated using a growth function, causing it to increase rapidly as the operating condition data increases, and a minimum value operation is used to limit the emergency response threshold from exceeding a preset upper limit. The enhanced monitoring threshold is calculated using a polynomial growth function, causing it to increase smoothly as the operating condition data increases. Both the emergency response threshold and the enhanced monitoring threshold have their initial levels and response rates determined by their respective base thresholds.

4. The nitrogen plugging early warning and automatic handling control method as described in claim 1, characterized in that, The tiered processing based on the comparison results between the nitrogen plug risk data and the emergency response threshold and the enhanced monitoring threshold includes: If the process safety level is the highest preset hazard level, the nitrogen plugging risk data will not be compared with the threshold, and emergency control actions will be executed directly. If the process safety level is not the highest hazard level and the nitrogen plugging risk data is not lower than the emergency response threshold, an active adjustment action is executed and a waiting period is entered. During the waiting period, the control action at the current level is not repeated, but the channel to upgrade the response level is retained. If the nitrogen slugging risk data is lower than the emergency response threshold and not lower than the enhanced monitoring threshold, then aggregation trend analysis is performed; if the nitrogen slugging risk data is lower than the enhanced monitoring threshold, no processing is performed; wherein, the aggregation trend analysis is based on the statistical characteristics of the nitrogen slugging risk data over at least several consecutive periods.

5. The nitrogen plugging early warning and automatic handling control method as described in claim 4, characterized in that, If the nitrogen plug risk data is lower than the emergency response threshold and not lower than the enhanced monitoring threshold, then aggregated trend analysis processing is performed, including: If the nitrogen plug risk data is lower than the emergency response threshold and not lower than the enhanced monitoring threshold, the following processing is performed: When calculating the mean change rate, the change in production load is used as the independent variable of the multiple linear regression model to eliminate the interference of load fluctuations on the parameter trend. The nitrogen plug development trend is assessed based on the residual trend after eliminating the interference. When the slope of the residual trend exceeds the preset minimum change rate threshold and the cumulative count of anomalies exceeds the preset proportion threshold for at least two consecutive processing cycles, it is determined that the nitrogen plug is showing a deteriorating trend, and the response level is upgraded to an active adjustment action.

6. The nitrogen plugging early warning and automatic handling control method as described in claim 1, characterized in that, The determination of nitrogen plug risk data based on the determined calculation data includes: After normalizing the thermal deviation data, the abnormal mode data, and the sensor reliability data, they are multiplied by their respective configurable weight parameters and summed. This sum is then added to the product of the process safety level data and its corresponding configurable weight parameter to obtain the nitrogen plug risk data. The sensor reliability data is used to mitigate the impact of sensor data with large consistency deviations on nitrogen plug risk assessment: for target sensors whose measurement consistency deviation with adjacent sensors in the same heat exchange section exceeds a preset threshold, the weight coefficient of the target sensor in the scoring calculation is reduced; for sensors with normal consistency, the standard weight coefficient remains unchanged.

7. The nitrogen plugging early warning and automatic handling control method as described in claim 4, characterized in that, The emergency control actions include: Automatically perform at least one of the following adjustment actions: adjust the superheat setting or opening of the cold end expansion valve of the cold box to reduce the cold end cooling output and cause the cold end temperature to rise; modify the raw material gas precooling temperature setting; and bypass part of the refrigerant flow from the cold end to the intermediate section by adjusting the opening of the refrigerant bypass valves of the intermediate section and the cold end. The system generates a suggested load reduction plan and displays it on the confirmation screen. After receiving confirmation from the operator, the load adjustment is executed.

8. The nitrogen plugging early warning and automatic handling control method as described in claim 4, characterized in that, During the waiting period, the current level of control action will not be repeated, but the channel to escalate the response level will be retained, including: During the waiting period, the current level of control action will not be repeated. When the process safety level data is upgraded to the target level, or the nitrogen plugging risk data continues to rise and exceeds the threshold of the previous level, the waiting period restriction will be lifted immediately and a higher level response will be executed. The duration of the waiting period is dynamically calculated based on the cold box heat capacity and historical response characteristics.

9. The nitrogen plugging early warning and automatic handling control method as described in claim 4, characterized in that, The method further includes: If the control action fails to execute on the first attempt, the action will be retried a preset number of times, and a pending execution flag will be set. After the system returns to normal, check if there is a pending execution flag. If there is a pending execution flag, re-execute the control action that was not successfully executed before.

10. A nitrogen plugging early warning device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as claimed in any one of claims 1 to 9.

Images