A gas-liquid low-temperature two-phase flow state monitoring system

Abstract

The application discloses a kind of gas-liquid low temperature two-phase flow state monitoring systems, it is related to low temperature fluid monitoring technical field, for solving gas-liquid low temperature two-phase flow state monitoring precision low, early warning lag problem;The application is fused by real-time monitoring two-phase flow state data and processing by capacitor and ultrasonic sensor array;Image acquisition is carried out by camera, and fluid feature analysis and fluid feature library matching judge flow pattern category;By analyzing the state characteristics and flow parameters of gas-liquid two-phase flow in low temperature pipeline, comprehensive fluid state evaluation is provided;Based on monitoring, identification and analysis results, intelligent control and fault prevention of low temperature pipeline are realized, comprehensive monitoring and analysis of gas-liquid two-phase flow in low temperature pipeline are realized, and the identification precision of two-phase flow state and abnormal early warning are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of cryogenic fluid monitoring technology, specifically to a gas-liquid cryogenic two-phase flow state monitoring system. Background Technology

[0002] With the rapid development of industries such as aerospace propulsion, liquefied natural gas, and cryogenic refrigeration, gas-liquid cryogenic two-phase flow is widely present in various industrial systems. Real-time monitoring and precise control of its flow state are directly related to the safety, stability, and energy efficiency of the system operation, and are the core technical link to ensure the reliable operation of cryogenic industrial processes.

[0003] With the development of sensing technology and data processing methods, the field of gas-liquid two-phase flow monitoring has evolved from single-parameter detection to multi-parameter collaborative monitoring. By integrating sensors such as pressure, temperature, and flow velocity, basic perception of the state of low-temperature fluids has been initially achieved, providing some data support for system operation status assessment. However, existing technologies still have many shortcomings:

[0004] Traditional monitoring schemes often rely on single-type sensors, fail to establish a quantitative correlation between core parameters and flow state of gas-liquid cryogenic two-phase flow, and do not integrate the spatiotemporal characteristics of multi-source sensor data in their state analysis strategies. They also lack dynamic calibration mechanisms for sensor drift under cryogenic conditions and collaborative analysis methods for multi-parameter coupling. Furthermore, flow state prediction does not incorporate the nonlinear variation law of fluid characteristics under cryogenic conditions. These factors lead to problems such as low accuracy in two-phase flow state identification and delayed anomaly warnings.

[0005] To address the aforementioned shortcomings, a technical solution is provided. Summary of the Invention

[0006] The purpose of this invention is to solve the problems of low accuracy in flow state identification and delayed abnormal early warning, and to propose a gas-liquid low-temperature two-phase flow state monitoring system.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] A gas-liquid cryogenic two-phase flow state monitoring system includes:

[0009] Phase monitoring module: By deploying capacitance sensors to collect capacitance signal changes, it obtains the current gas phase state, liquid phase state, and gas-liquid two-phase mixing state of the fluid, and at the same time collects the pressure, temperature and flow velocity of the gas-liquid two-phase flow in the pipeline;

[0010] Acoustic signals are collected and analyzed in real time by deploying ultrasonic sensors;

[0011] Data from the capacitor and ultrasonic sensor is fused based on a data conflict resolution mechanism.

[0012] Flow pattern recognition module: It acquires images through a group of deployed cameras, analyzes the fluid morphology characteristics and matches them with a pre-built flow pattern feature library, and determines the flow pattern category by combining image features, including stratified flow, wavy flow, slug flow, annular flow and diffuse flow;

[0013] Two-phase flow state analysis module: Based on phase state monitoring and flow pattern identification data, it analyzes the state characteristics and flow parameters of gas-liquid two-phase flow, including a flow parameter calculation unit, a phase stability assessment unit, and a flow pattern transformation prediction unit;

[0014] The intelligent control and fault prevention module, based on monitoring, identification, and analysis results, enables intelligent control and fault prevention of cryogenic pipelines, including control units and fault diagnosis and maintenance units.

[0015] Furthermore, the process of acquiring and analyzing the acoustic signals includes:

[0016] Set acoustic markers, establish the mapping relationship between acoustic time and fluid phase state, generate a phase state virtual network, and divide the standard acoustic response region;

[0017] The acoustic signal features are extracted and pattern fitted. The liquid phase value is output when the liquid phase feature intensity exceeds 40% of the total signal.

[0018] Using the baseline acoustic features under standard conditions as a template, the similarity of the current features is compared, and if it is lower than a preset threshold, the change value is output.

[0019] A filtering algorithm is used to calculate the phase change rate and generate a sequence. The stable phase or phase transition is determined by the active window of the pipe section, and the phase probability value is output. The decision value is obtained by combining the weighted formula.

[0020] When the decision value is greater than the preset threshold and liquid phase features are continuously detected, acoustic features are successfully matched, and the phase state remains stable for more than the preset time, it is determined to be a liquid phase state.

[0021] When the decision value is less than the preset threshold and no effective liquid phase characteristics are detected continuously, it is determined to be a gas phase state.

[0022] When the decision value is between two preset thresholds and the liquid phase characteristic is below 40%, it is determined to be a gas-liquid mixture state.

[0023] Furthermore, the specific operation process of the flow parameter calculation unit is as follows:

[0024] Extract pressure value and temperature The mixed density was calculated. ; through formula The cavitation rate is obtained, where, These represent the density of the pure liquid phase and the density of the pure gas phase, respectively.

[0025] The liquid phase characteristic time delay YC and the gas phase characteristic time delay XC are obtained by analyzing the time delay of two adjacent capacitive sensor signals; the gas phase flow velocity is obtained based on the flow velocity data. Liquid phase flow rate YS; normalized and then substituted into the formula Obtain the gas-liquid velocity ratio;

[0026] The cavitation rate and gas-liquid velocity ratio are combined using a weighted fusion formula to construct a flow state index.

[0027] Furthermore, the specific operation steps of the phase stability evaluation unit include:

[0028] The spectral features of the acoustic signal are extracted, filtered, and the DC component is extracted to obtain the instantaneous situation wave. Periodic characteristic spectrum obtained based on Fourier transform ;

[0029] Extract the largest angular frequency from the periodic characteristic spectrum and convert it into the dominant frequency. Simultaneously, the maximum absolute value is extracted from the instantaneous situation wave and marked as the pulsation amplitude. Normalization process followed by input into the formula The stability coefficient is obtained, where, This represents the average pressure. Indicates the natural frequency of the pipe;

[0030] when When the two-phase flow is in a stable state, it is determined that the two-phase flow is in a stable state.

[0031] when When the two-phase flow is in a critically unstable state, it is determined that the two-phase flow is in a critically unstable state.

[0032] when At that time, the two-phase flow state was determined to be severely unstable.

[0033] Furthermore, the specific operation steps of the flow pattern transformation prediction unit include:

[0034] Record the historical flow pattern change sequence and corresponding operating parameters of the monitoring points, construct a historical flow pattern data database and perform time series analysis, extract flow pattern transformation characteristic parameters, including flow pattern duration, transformation frequency and transformation rate, and obtain the characteristic parameter value TZ through a weighted formula;

[0035] The probability value of flow pattern transition is obtained by combining the cavitation ratio, gas-liquid velocity ratio, and stability coefficient of the flow parameter calculation unit. ;

[0036] When the transition probability value exceeds a preset threshold and the stability coefficient shows a decreasing trend over multiple consecutive sampling periods, a flow pattern transition warning is triggered, using the formula... Obtain the expected transformation time ,in, For the current time, The characteristic time constant; The baseline probability threshold; The characteristic decay factor represents the decay characteristic of the transition probability over time. This represents an operator; it is used to quantize the ratio of the transition probability value to the baseline probability threshold and the characteristic decay factor into a unitless constant.

[0037] Based on the predicted flow pattern shift and the expected shift time, a flow pattern shift risk assessment report is generated and mapped to three risk levels: low, medium, and high.

[0038] Furthermore, the specific process for determining the risk level includes:

[0039] when and If the risk level is determined to be low at a given minute, the standard monitoring frequency will be maintained; flow pattern parameters will be recorded but no warning will be triggered.

[0040] when and At the specified time, the risk level was determined to be medium, the monitoring frequency was increased, and a yellow alert was triggered; the gas-liquid ratio was adjusted to a safe range; and technical personnel were notified to be on standby.

[0041] when and If the risk level is determined to be high at a given minute, a red alert and audible / visual alarm will be triggered; the pipeline vibration frequency will be monitored in real time. With pressure fluctuations The maximum load value is then calculated.

[0042] If the maximum load value is lower than the preset threshold, the load value will be gradually reduced until it stabilizes; otherwise, the machine will be forcibly shut down and maintenance will be notified.

[0043] Furthermore, the specific operation process of the intelligent control and fault prevention module includes:

[0044] Upon receiving a risk assessment report, if the risk level is medium or above, calculate the optimal control parameters to obtain control adjustment values, generate execution instructions, and send them to the actuators in stages; construct a parameter adaptive adjustment mechanism, set an initial value for the adjustment step size, increase the adjustment step size when the adjusted parameter response is insufficient, and decrease the adjustment step size when an over-response occurs;

[0045] A fault feature library is built based on historical data, and abnormal states are associated with flow pattern transformation features. When flow pattern transformation features are detected, potential fault types are matched. The similarity between the current state and historical fault features is calculated to generate a fault probability distribution. When the fault probability exceeds a preset threshold, a fault warning report is generated. The maintenance cycle is calculated based on the phase stability assessment results. The maintenance cycle is shortened when the stability coefficient continues to decrease and extended when it is stable at a high value.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] This invention monitors the gas-liquid two-phase flow state in a cryogenic pipeline in real time using a multimodal sensor array. Based on a data conflict resolution mechanism, it fuses data from capacitive and ultrasonic sensors, effectively solving the problem of phase misjudgment caused by interference of a single sensor under complex cryogenic conditions, and significantly improving the accuracy of phase state identification of gas-liquid two-phase flow.

[0048] This invention, based on phase monitoring and flow pattern identification data, achieves comprehensive analysis of the two-phase flow state through flow parameter calculation, phase stability assessment, and flow pattern transition prediction. It dynamically triggers different control strategies according to risk level: low risk maintains the regular monitoring frequency, medium risk increases the monitoring frequency and adjusts operating parameters, and high risk initiates emergency measures. Through parameter adaptive adjustment and fault diagnosis and maintenance mechanisms, it achieves optimized control throughout the entire process, significantly improving the safety and stability of cryogenic industrial systems and effectively enhancing the early warning of abnormal states. Attached Figure Description

[0049] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0050] Figure 1 This is the overall system block diagram of the present invention. Detailed Implementation

[0051] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] It should be understood that the terms “comprising” and “including” used in this disclosure and claims indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0053] It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this disclosure. As used in this disclosure and claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this disclosure and claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations.

[0054] like Figure 1 As shown, the present invention is a gas-liquid low-temperature two-phase flow state monitoring system, including a phase state monitoring module, a flow pattern identification module, a two-phase flow state analysis module, and an intelligent control and fault prevention module.

[0055] The phase state monitoring module uses a multi-modal sensor array to monitor the gas-liquid two-phase flow state data in the cryogenic pipeline in real time; the specific monitoring process includes:

[0056] The multimodal sensor array includes a group of capacitive sensors and an ultrasonic sensor array. The capacitive sensors are uniformly installed along the pipeline axis and generate dynamic monitoring signals by continuously collecting changes in capacitance signals. This allows for real-time acquisition of the current fluid phase state, including gas phase, liquid phase, and gas-liquid two-phase mixed state. Simultaneously, core data of the gas-liquid two-phase flow in the pipeline under low-temperature conditions are collected, including pressure, temperature, and flow velocity.

[0057] When fluid flows in a pipe, fluids in different phases have different dielectric constants;

[0058] When the capacitance sensor detects a change in capacitance signal exceeding a preset threshold, it determines the state to be liquid.

[0059] When the capacitance sensor detects a change in capacitance signal value below a preset threshold, it is determined to be in a gas phase state.

[0060] When the change in capacitance signal fluctuates between two preset thresholds, it is determined to be a gas-liquid two-phase mixture state.

[0061] Ultrasonic sensors are used as an auxiliary means to monitor the phase state of fluids. High-frequency ultrasonic sensors are deployed at various key nodes in cryogenic pipelines, with each sensor covering the complete acoustic profile of the pipeline cross-section. These sensors are used to collect acoustic signals and perform real-time analysis. The specific analysis process is as follows:

[0062] By pre-setting acoustic markers on the inner wall of the pipe, the ultrasonic sensor automatically establishes a mapping relationship between acoustic time and fluid phase state through a calibration algorithm, and generates a virtual network of fluid phase state. Each measurement section is divided into a standard acoustic response area and bound with a unique number; the distribution of different phase states is accurately located through acoustic time mapping.

[0063] Based on an improved deep learning network, feature extraction and pattern fitting are performed on the detected acoustic signals. When the liquid phase feature intensity accounts for more than 40% of the total signal intensity, the liquid phase value YX is output.

[0064] Under standard conditions, the reference acoustic features inside the pipeline are pre-collected as a reference template, and the similarity between the current acoustic features and the reference template is compared in real time. When the similarity of the acoustic features is lower than a preset threshold, it is determined that the fluid state inside the pipeline has changed significantly, and the change value BH is output.

[0065] An improved Kalman filter is used to track fluid state changes. The fluid phase change rate is calculated at a sampling rate of 100Hz to generate a state change sequence. An active window for the pipe cross section is set. When the phase change rate is detected to drop from the initial value to near zero within 20 consecutive sampling points, it is determined to be a stable phase. When the phase changes from a stable state and reaches another stable state within 15 consecutive sampling points, it is determined to be a phase transition, and the phase probability value GL is output.

[0066] After normalizing the liquid phase value YX, the change value BH, and the phase probability value GL, they were substituted into the formula. The decision value is obtained through calculation, where, These are the influence weighting factors for liquid phase value, change value, and phase probability value, respectively;

[0067] When the decision value is greater than the preset threshold and simultaneously meets the following conditions: liquid phase characteristics are detected for 5 consecutive sampling cycles, acoustic characteristics are successfully matched, and the phase state remains stable in the pipeline for more than 2 seconds, it is determined to be a liquid phase state.

[0068] When the decision value is less than the preset threshold and no effective liquid phase features are detected for 30 consecutive sampling cycles or the phase tracking shows that the liquid phase has completely disappeared, it is determined to be a gas phase state.

[0069] When the decision value fluctuates between two preset thresholds and the liquid phase characteristic is below 40%, it is determined to be a gas-liquid mixture state.

[0070] Data from capacitive and ultrasonic sensors are fused based on a data conflict resolution mechanism.

[0071] When the judgment results monitored by the capacitive sensor and the ultrasonic sensor are consistent, the current fluid phase state is directly determined and the phase state is updated.

[0072] When the judgment results monitored by the capacitive sensor and the ultrasonic sensor are inconsistent, a second judgment is performed, and the time window second verification mechanism is activated.

[0073] Three data acquisitions are triggered within 15 seconds at 5-second intervals to simultaneously acquire the capacitance signal change sequence of the capacitance sensor and the acoustic signal of the ultrasonic sensor. The timing of the appearance of phase characteristics in the acoustic signal is compared with the capacitance change curve using a dynamic time warping algorithm. If the time difference between the sudden change in capacitance signal and the phase change in acoustic signal is less than 2 seconds in at least two acquisitions, it is determined to be a liquid phase state. If all three acquisitions show that the capacitance change is consistently below the preset threshold and there are no liquid phase characteristics in the acoustic signal, it is determined to be a gas phase state. If two acquisitions show that the capacitance change is consistently below the preset threshold, but there are still weak liquid phase characteristics in the acoustic signal, it is determined to be a gas-liquid mixed state.

[0074] The flow pattern recognition module deploys a camera array within the transparent observation window of the cryogenic pipeline to identify and record the fluid flow pattern. The specific process is as follows:

[0075] As fluid flows through the transparent observation window, a deployed camera system captures omnidirectional images of the fluid flow. These images are then transmitted to a flow pattern recognition server for processing. The server analyzes the fluid morphology characteristics and matches them against a pre-built flow pattern feature library, while also combining image features to determine the flow pattern category. Flow pattern categories include stratified flow, wavy flow, slug flow, annular flow, and diffuse flow. Simultaneously, image enhancement technology is used to extract flow pattern boundary features, and the flow pattern category is determined based on the boundary morphology. The flow pattern category, boundary features, and phase distribution obtained at the same time are integrated into the current fluid flow pattern information.

[0076] The two-phase flow state analysis module, based on the output data of the phase monitoring module and the flow pattern recognition module, analyzes the state characteristics and flow parameters of the gas-liquid two-phase flow in the cryogenic pipeline in real time, providing a comprehensive fluid state assessment, including a flow parameter calculation unit, a phase stability assessment unit, and a flow pattern transformation prediction unit; the specific analysis process includes:

[0077] The flow parameter calculation unit calculates the cavitation rate and gas-liquid velocity ratio in real time based on the fused data from capacitive and ultrasonic sensors to form flow state indices; among which:

[0078] Cavitation rate calculation: Based on the collected core data, pressure values ​​are extracted. and temperature The mixing density was calculated. ; through formula The cavitation rate is calculated, where, These represent the density of the pure liquid phase and the density of the pure gas phase, respectively.

[0079] Gas-liquid velocity ratio: The characteristic time delay YC of the liquid phase and the characteristic time delay XC of the gas phase are obtained by analyzing the time delay of two adjacent capacitive sensor signals; the gas phase velocity is obtained based on the collected flow velocity data. Liquid phase flow rate YS; normalized and then substituted into the formula The gas-liquid velocity ratio was calculated.

[0080] A weighted fusion formula is used to fuse cavitation rate and gas-liquid velocity ratio to construct a flow state index; the larger the flow state index, the easier it is for the flow state to change.

[0081] Phase stability assessment unit: Extracts the spectral characteristics of the acoustic signal, performs bandpass filtering on the spectral characteristics, and extracts the DC component to obtain the instantaneous state wave. ; through formula The periodic characteristic spectrum is obtained by calculation. ; Indicates the frequency of target analysis; Indicates the sampling frequency; Indicates the number of signal sampling points; Represents the first digit in the discretized time-domain signal. The serial number of each sampling point; Indicates the first A discrete frequency point; Represents the imaginary unit; , indicating the sampling time interval;

[0082] Extract the largest angular frequency from the periodic feature spectrum And convert it to the main frequency. Simultaneously, the maximum absolute value is extracted from the instantaneous situation wave and marked as the pulsation amplitude. Normalization process followed by input into the formula The stability coefficient is calculated, where, This represents the average pressure. Indicates the natural frequency of the pipe;

[0083] when If the two-phase flow is determined to be in a stable state and will not affect the safe operation of the pipeline, then routine measures are taken, including equipment inspections according to the routine inspection cycle and continuous monitoring of the trend of stability coefficient changes.

[0084] when When the two-phase flow is in a critically unstable state, it is determined that the two-phase flow is in a critical state of instability. When the two-phase flow reaches the critical state that causes pipeline vibration, critical measures are taken, including shortening the equipment inspection cycle, strengthening the detection of pipeline vibration and abnormal noise, using vibration monitoring instruments to measure the frequency and amplitude of pipeline vibration and analyze the vibration source, checking whether the operating parameters of the equipment are normal, and adjusting the operating conditions of the equipment or performing maintenance and repair if necessary.

[0085] when If the two-phase flow state is determined to be severely unstable, posing a serious threat to the safety of the pipeline structure and causing pipeline vibration fatigue and sensor failure, then emergency measures should be taken, including reducing the pipeline operating load or suspending pipeline operation; a comprehensive inspection of the pipeline should be carried out to investigate the cause of the instability.

[0086] Flow pattern change prediction unit: Based on the collected multidimensional data and flow pattern identification results, it predicts the flow pattern change of gas-liquid two-phase flow in a cryogenic pipeline. The specific process is as follows:

[0087] Record the flow pattern change sequence and its corresponding operating parameters for each monitoring point over the past 24 hours to construct a historical flow pattern data database; perform time-series analysis on the historical data to extract flow pattern transition characteristic parameters, including flow pattern duration, transition frequency, and transition rate, and calculate the characteristic parameter value TZ using a weighted formula;

[0088] Combining the flow parameter calculation unit's cavitation rate, gas-liquid velocity ratio, and stability coefficient, the formula is used to calculate the cavitation rate, gas-liquid velocity ratio, and stability coefficient. The probability value of manifold transition is obtained by calculation. ;in, This represents the probability of transitioning from manifold i to manifold j. These represent the weighting factors affecting the characteristic parameter values ​​of cavitation rate, gas-liquid velocity ratio, and stability coefficient, respectively.

[0089] When the transition probability value When the stability coefficient exceeds a preset threshold and shows a decreasing trend for three consecutive sampling periods, a flow pattern transition warning is triggered, using the formula... The expected transition time was calculated. ,in, For the current time, The characteristic time constant; The baseline probability threshold; The characteristic decay factor represents the decay characteristic of the transition probability over time. This represents an operator; it is used to quantize the ratio of the transition probability value to the baseline probability threshold and the characteristic decay factor into a unitless constant.

[0090] Based on the predicted flow pattern shift and the expected shift time, a flow pattern shift risk assessment report is generated and mapped to a three-level risk level: low, medium, and high.

[0091] when and If the risk level is determined to be low at a given minute, the standard 30-minute monitoring frequency is maintained; flow pattern parameters are recorded but no alerts are triggered; and 24-hour trend reports are generated regularly for maintenance personnel to refer to.

[0092] when and If the risk level is determined to be medium after a certain number of minutes, the monitoring frequency is increased to once every 5 minutes, triggering a yellow alert; the sensor sampling rate is increased to 200Hz to improve transient detection capability; the inlet temperature or flow rate is adjusted; the load is reduced by 15% to 25%; the gas-liquid ratio is adjusted to a safe range; and technical personnel are notified to be on standby and prepare possible intervention measures.

[0093] when and If the risk level is determined to be high at a given minute, a red alert and audible / visual alarm will be triggered; the pipeline vibration frequency will be monitored in real time. With pressure fluctuations ; through formula The maximum load value is obtained through calculation; where, Indicates the baseline load; Indicates the safe threshold for vibration frequency; Indicates the fluctuation of the benchmark pressure; Indicates the safety threshold for pressure fluctuations; These are the influencing factors of vibration frequency and pressure fluctuation, respectively;

[0094] If the maximum load value is lower than the preset threshold, the load value will be gradually reduced at a frequency of 10% until it stabilizes. If the maximum load value is higher than the preset threshold, the machine will be forcibly shut down and technicians will be notified to carry out maintenance work.

[0095] By comparing the deviation between the actual transition time and the expected transition time, the calculation parameters of the flow pattern transition probability matrix are adjusted in reverse or the characteristic decay factor of the expected transition time is optimized. ;

[0096] The intelligent control and fault prevention module, based on monitoring, identification, and analysis results, enables intelligent control and fault prevention of cryogenic pipelines. It includes a control unit and a fault diagnosis and maintenance unit, the specific implementation process of which is as follows:

[0097] Based on the output of the two-phase flow state analysis module, the control unit performs parameter adjustments and flow state optimization. The specific process includes:

[0098] The system receives the risk assessment report generated by the flow pattern change prediction unit. When the risk level reaches medium or above, it automatically calculates the optimal control parameters, obtains the control adjustment value based on the optimal control parameters, generates execution instructions, and sends them to the actuators in stages.

[0099] For medium risk, the actuator adjusts parameters, including flow rate reduction and temperature adjustment; for higher risk, the actuator reduces load; for high risk, the actuator executes deep load reduction or emergency shutdown commands; an adaptive parameter adjustment mechanism is constructed, setting the adjustment step size to an initial value. When the adjusted parameter response is insufficient, the adjustment step size is increased; when the adjusted parameter exhibits an over-response, the adjustment step size is decreased.

[0100] The fault diagnosis and maintenance unit constructs a fault feature database based on historical data from phase monitoring and flow pattern identification to achieve early fault identification and prediction. The specific process includes:

[0101] Record abnormal states and corresponding flow pattern transformation characteristics that occur during operation, and establish a database that associates faults with flow patterns;

[0102] When flow pattern transformation characteristics are detected, potential fault types are matched; the similarity between the current state and historical fault characteristics is calculated to generate a fault probability distribution; when the fault probability exceeds a preset threshold, a fault warning report is generated.

[0103] Based on the phase stability assessment results, the maintenance cycle is calculated; when the stability coefficient continues to decrease, the maintenance cycle is shortened; when the stability coefficient stabilizes at a high value, the maintenance cycle is extended.

[0104] Based on historical flow pattern transition data, identify operating conditions prone to flow pattern transitions and generate operating condition optimization suggestions to avoid unstable operating conditions; regularly generate pipeline health assessment reports, including wear predictions for key components, vibration fatigue accumulation, and sealing performance change trends; and combine the assessment results to prioritize maintenance levels and specific maintenance items.

[0105] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A gas-liquid cryogenic two-phase flow state monitoring system, characterized in that, include: Phase monitoring module: By deploying capacitance sensors, it continuously collects capacitance signal changes, generates dynamic monitoring signals, and obtains the current gas phase state, liquid phase state, and gas-liquid two-phase mixing state of the fluid in real time. At the same time, it collects the pressure, temperature, and flow velocity of the gas-liquid two-phase flow in the pipeline under low temperature conditions. Ultrasonic sensors are deployed by pre-set acoustic markers, with each sensor covering a complete acoustic profile of the pipe section for acquiring acoustic signals and analyzing them in real time. Data from capacitive and ultrasonic sensors are fused based on a data conflict resolution mechanism. Flow pattern recognition module: It acquires images through a deployed camera array, analyzes fluid morphology characteristics and matches them with a pre-built flow pattern feature library; at the same time, it combines image features to determine the flow pattern category, including stratified flow, wavy flow, slug flow, annular flow and diffuse flow; it uses image enhancement technology to extract flow pattern boundary features and determine the flow pattern category. Two-phase flow state analysis module: Based on the output data of the phase state monitoring module and the flow pattern recognition module, it analyzes the state characteristics and flow parameters of the gas-liquid two-phase flow in the cryogenic pipeline, including a flow parameter calculation unit, a phase stability assessment unit and a flow pattern transformation prediction unit; The specific operation steps of the phase stability evaluation unit include: Extract the spectral features of the acoustic signal, perform bandpass filtering on the spectral features, and extract the DC component to obtain the instantaneous situation wave. Periodic characteristic spectrum obtained based on discrete Fourier transform ; Extract the largest angular frequency from the periodic feature spectrum And convert it to the main frequency. Simultaneously, the maximum absolute value is extracted from the instantaneous situation wave and marked as the pulsation amplitude. Normalization process followed by input into the formula The stability coefficient is obtained, where, This represents the average pressure. Indicates the natural frequency of the pipe; when At this time, it is determined that the two-phase flow is in a stable state and will not affect the safe operation of the pipeline; when When the two-phase flow is in a critically unstable state, it is determined that the two-phase flow state has reached the critical state that will cause pipe vibration. when When the two-phase flow state is determined to be severely unstable, it poses a serious threat to the safety of the pipeline structure, leading to pipeline vibration fatigue and sensor failure. Intelligent control and fault prevention module: Based on monitoring, identification and analysis results, it realizes intelligent control and fault prevention of cryogenic pipelines, including control unit and fault diagnosis and maintenance unit.

2. The gas-liquid low-temperature two-phase flow state monitoring system according to claim 1, characterized in that, The process of acquiring and analyzing the acoustic signals includes: By pre-setting acoustic markers, the sensor establishes a mapping relationship between acoustic time and fluid phase state using a calibration algorithm, generates a phase state virtual network, and divides the measurement section into standard acoustic response regions with unique numbers. Based on an improved deep learning network, acoustic signal features are extracted and pattern fitting is performed. When the liquid phase feature intensity exceeds 40% of the total signal, the liquid phase value is output. Using the baseline acoustic features under standard conditions as a template, the similarity of the current features is compared in real time, and the change value is output if it is lower than the preset threshold. An improved Kalman filter is used to calculate the phase change rate and generate a sequence at a sampling rate of 100Hz. The stable phase state or phase transition is determined by the active window of the pipe section, and the phase probability value is output. The decision value is obtained by a weighted formula. When the decision value is greater than the preset high threshold and liquid phase characteristics are detected for 5 consecutive sampling cycles, acoustic characteristics are successfully matched, and the phase state remains stable in the pipeline for more than 2 seconds, it is determined to be a liquid phase state. When the decision value is less than the preset low threshold and no effective liquid phase characteristics are detected for 30 consecutive sampling cycles, it is determined to be a gas phase state. When the decision value is between the preset low threshold and high threshold and the liquid phase characteristic is less than 40%, it is determined to be a gas-liquid mixed state.

3. The gas-liquid low-temperature two-phase flow state monitoring system according to claim 1, characterized in that, The specific operation process of the flow parameter calculation unit is as follows: Based on the fusion data from capacitive and ultrasonic sensors, the cavitation rate and gas-liquid velocity ratio are calculated to form flow state indices; where: Based on the collected core data, extract the pressure value. and temperature The mixed density was calculated. ; through formula The cavitation rate is obtained, where, These represent the density of the pure liquid phase and the density of the pure gas phase, respectively. The liquid phase characteristic time delay YC and the gas phase characteristic time delay XC are obtained by analyzing the time delay of two adjacent capacitive sensor signals; the gas phase flow velocity is obtained based on the collected flow velocity data. Liquid phase flow rate YS; normalized and then substituted into the formula Obtain the gas-liquid velocity ratio; The cavitation rate and gas-liquid velocity ratio are fused using a weighted fusion formula to construct a flow state index; the larger the flow state index, the easier it is for the flow state to change.

4. The gas-liquid low-temperature two-phase flow state monitoring system according to claim 3, characterized in that, The specific operation steps of the flow pattern transformation prediction unit include: Record the flow pattern change sequence and its corresponding operating parameters for each monitoring point over the past 24 hours to construct a historical flow pattern data library; perform time-series analysis on the historical data to extract flow pattern transition characteristic parameters, including flow pattern duration, transition frequency, and transition rate, and obtain the characteristic parameter value TZ through a weighted formula; The probability value of flow pattern transition is obtained by combining the cavitation ratio, gas-liquid velocity ratio, and stability coefficient of the flow parameter calculation unit. ; When the transition probability value exceeds a preset threshold and the stability coefficient shows a decreasing trend over three consecutive sampling periods, a flow pattern transition warning is triggered, using the formula... Obtain the expected transformation time ,in, For the current time, The characteristic time constant; The baseline probability threshold; The characteristic decay factor represents the decay characteristic of the transition probability over time. This represents an operator; it is used to quantize the ratio of the transition probability value to the baseline probability threshold and the characteristic decay factor into a unitless constant. Based on the predicted flow pattern shift and the expected shift time, a flow pattern shift risk assessment report is generated and mapped to a three-level risk level, including low, medium, and high.

5. The gas-liquid low-temperature two-phase flow state monitoring system according to claim 4, characterized in that, The specific process for determining the risk level includes: when and If the risk level is determined to be low at a given minute, the standard 30-minute monitoring frequency will be maintained; flow pattern parameters will be recorded but no alerts will be triggered. when and If the risk level is determined to be medium at a certain time, the monitoring frequency is increased to once every 5 minutes, triggering a yellow alert; the gas-liquid ratio is adjusted to a safe range; and technical personnel are notified to be on standby. when and If the risk level is determined to be high at a given minute, a red alert and audible / visual alarm will be triggered; the pipeline vibration frequency will be monitored in real time. With pressure fluctuations And calculate the maximum load value; If the maximum load value is lower than the preset threshold, the load value will be gradually reduced at a frequency of 10% until it stabilizes. Conversely, if it is higher than the preset threshold, the machine will be forced to shut down and technicians will be notified to carry out maintenance work.

6. The gas-liquid low-temperature two-phase flow state monitoring system according to claim 1, characterized in that, The specific operation process of the intelligent control and fault prevention module includes: The system receives the risk assessment report generated by the flow pattern change prediction unit. When the risk level reaches medium or above, it calculates the optimal control parameters and obtains the control adjustment value, generates execution instructions and sends them to the actuators in stages. It also constructs a parameter adaptive adjustment mechanism, sets the adjustment step size as the initial value, increases the adjustment step size when the adjusted parameter response is insufficient, and decreases the adjustment step size when an over-response occurs. A fault feature library is constructed based on historical data from phase state monitoring and manifold recognition. An association database is established by recording abnormal states during operation and manifold transformation features. When manifold transformation features are detected, potential fault types are matched. The similarity between the current state and historical fault features is calculated to generate a fault probability distribution. When the fault probability exceeds a preset threshold, a fault warning report is generated. The maintenance cycle is calculated based on the phase stability assessment results. When the stability coefficient continues to decrease, the maintenance cycle is shortened. When it stabilizes at a high value, the maintenance cycle is extended.

Images