Pipeline anti-freezing system based on seawater freezing time margin and application method thereof

Abstract

The invention provides a seawater freezing time margin-based pipeline anti-freezing system and an application method thereof, and the pipeline anti-freezing system and the application method are particularly suitable for seawater transportation and distribution pipeline anti-freezing protection in an extremely cold environment. The system comprises a pipeline testing section and a water taking pipe section which are connected through a flange, an electric tracing band is spirally wound on the outer wall of the water taking pipe section, temperature sensors are arranged at intervals, and a seawater parameter sensor is arranged in the testing section. The environment monitoring system comprises an anemograph and a temperature and humidity sensor. Each sensor is connected with a computer through a signal acquisition card to acquire pipeline temperature, seawater parameters and environmental data in real time. The system takes the seawater freezing time margin as a control index, the electric heat tracing power is dynamically adjusted, and graded intelligent temperature control is achieved. The method solves the problem of polar region pipeline freezing, has the characteristics of quick response, accurate control and the like, and is suitable for pipeline anti-freezing protection in extreme environments such as the south pole and the north pole.

Description

Technical Field

[0001] This invention relates to the field of electric heat tracing technology, and in particular to a pipeline antifreeze system based on seawater freezing time margin and its application method. Background Technology

[0002] Antarctica possesses significant scientific development value. Its unique geographical location, extreme climate conditions, and abundant scientific resources have attracted more than 30 countries worldwide to establish research stations there. Since its first Antarctic expedition in 1984, China has built four research stations: Great Wall Station, Zhongshan Station, Kunlun Station, and Taishan Station.

[0003] In most parts of Antarctica, extreme storms are frequent and accompanied by blowing snow year-round, resulting in scarce and contaminated snowmelt, making it impossible to obtain fresh water through snowmelt. Therefore, seawater desalination has been widely adopted as a water intake solution. However, the extreme low temperatures in polar regions can easily cause water bodies to freeze or even rupture water intake pipelines, affecting the stable supply of water resources and posing a significant challenge to the stable operation of water intake systems.

[0004] Electric heat tracing technology is a commonly used method for preventing pipeline freezing. It uses an electric heating device to heat the pipeline and maintain the temperature of the fluid inside the pipeline within a safe range. For example, Du Wenfang has authorized a utility model invention, "An electric heat tracing device for pipeline freezing (CN 221857778U)," to prevent liquids in distribution pipelines from freezing in low-temperature environments and to solve the problems of low stability and poor heating effect of traditional electric heat tracing devices.

[0005] However, traditional pipeline electric heat tracing control strategies typically employ fixed temperature control thresholds. Specifically, the electric heat tracing device is activated when the pipeline temperature falls below a certain set value and deactivated when the pipeline temperature exceeds a certain set value. For example, Li Yilian et al. disclosed an invention patent entitled "An Electric Heat Tracing Temperature Control Strategy (CN 117170432A)," which controls the heating power of the electric heat tracing by setting upper and lower threshold values ​​for the pipeline surface temperature. The shortcomings of this traditional control strategy are:

[0006] (1) The response is delayed: Traditional strategies require waiting for the actual temperature change to reach the threshold before the heating action can be triggered, which results in a delay. In extreme low-temperature environments in polar regions, a sudden drop in temperature may cause the pipeline to freeze before the control is triggered.

[0007] (2) Frequent start-stop and low energy efficiency: When the upper and lower temperature limits are set to a small value, the electric heat tracing system will frequently start and stop in scenarios with small temperature fluctuations, resulting in low energy efficiency. Furthermore, frequent start-stop can easily cause the electric heat tracing system to malfunction due to power fluctuations or short circuits (such as leakage or overheating).

[0008] (3) Insufficient adaptability to complex environmental factors: The freezing process of seawater is not linearly related to water temperature. Using temperature as the only control variable and ignoring other key factors affecting freezing (such as ambient temperature, wind speed, permafrost temperature, fluid flow rate, etc.) will lead to a one-sided control strategy and insufficient adaptability to complex environmental factors. Summary of the Invention

[0009] The purpose of this invention is to provide a pipeline antifreeze system and its application method based on seawater freezing time margin for the antifreeze protection of polar seawater transmission and distribution pipelines. The system employs a machine learning LSTM recurrent neural network to train a seawater freezing prediction model. It uses the seawater freezing time margin (i.e., the time required for the coldest point of the pipeline to fully freeze, abbreviated as TTF) instead of seawater temperature as the indicator for electric heat tracing power control, forming a graded heating control strategy for pipeline electric heat tracing based on the seawater freezing time margin. This improves the control logic of the electric heat tracing system, thereby solving the problem of electric heat tracing temperature control strategy.

[0010] To achieve the above objectives, the present invention adopts the following technical solution:

[0011] A pipeline antifreeze system based on seawater freezing time margin, suitable for intelligent temperature control of pipeline components, including electric heat tracing components, detection components, signal acquisition card, environmental monitoring system and computer;

[0012] The pipeline assembly includes at least one pipeline test section and multiple water intake pipe sections, which are connected by flanges.

[0013] The electric heat tracing assembly includes an electric heat tracing tape spirally wound and laid on the outer wall of the water intake pipe section;

[0014] The detection component includes a pipe outer wall temperature sensor that is uniformly and tightly attached to the outer wall of the water intake pipe section, and each of the pipe outer wall temperature sensors is spaced apart in the spiral gap formed by the electric heating tape on the water intake pipe section;

[0015] And seawater temperature sensors, electromagnetic flow meters, and seawater salinity sensors are centrally installed in the pipeline test section;

[0016] The environmental monitoring system includes an ultrasonic anemometer and an ambient temperature sensor.

[0017] The detection component and the environmental monitoring system are respectively connected to the signal acquisition card via cables, and the signal acquisition card is connected to a computer for communication.

[0018] Preferably, the pipe assembly is wrapped with a pipe insulation layer.

[0019] Preferably, the environmental monitoring component is installed inside a radiation-proof ventilation hood, and the surface of the radiation-proof ventilation hood is provided with an electrically heated defrosting device.

[0020] This invention also provides an application method for a pipeline antifreeze system based on seawater freezing time margin, comprising the following steps:

[0021] Step S1: Laying of electric heating tape and installation of sensors:

[0022] At least one pipeline test section and multiple water intake pipe sections are connected by flanges; and an electric heating tape is spirally wound and laid on the outer wall of the water intake pipe section in the pipeline assembly. Multiple pipe outer wall temperature sensors are installed at intervals in the spiral gaps formed on the water intake pipe section. Seawater temperature sensor, electromagnetic flow meter and seawater salinity sensor are centrally installed in the pipeline test section.

[0023] The pipe outer wall temperature sensor, seawater temperature sensor, electromagnetic flow meter, seawater salt concentration sensor, ultrasonic anemometer, and ambient temperature sensor are connected to the signal acquisition card via cables, and the collected data is transmitted to the computer via communication connection for real-time data acquisition.

[0024] Step S2: Data Acquisition: Obtain at least six parameters of the pipeline components through the real-time data acquired in Step S1. The six parameters include the pipeline outer wall temperature, seawater flow velocity, seawater temperature, seawater salinity, surface wind speed, and ambient temperature data. Outliers and missing values ​​are removed, interpolation is performed, spatial-temporal grid mapping is performed, temperature change rate and comprehensive heat transfer coefficient are calculated, and finally normalization is performed.

[0025] Step S3: Construction of a seawater freezing prediction model based on machine learning LSTM: Using normalized eight-dimensional time series data, including pipe outer wall temperature, seawater flow velocity, seawater temperature, seawater salinity, surface wind speed, ambient temperature, temperature change rate, and comprehensive heat transfer coefficient as input, and combining the lower limit of static water freezing time and the constraint of flowing water freezing time, an LSTM neural network is constructed and trained to obtain a seawater freezing time margin (TTF) prediction model.

[0026] Step S4: Risk classification of seawater freezing time: Based on the predicted real-time TTF value obtained from the seawater freezing time margin TTF prediction model, four risk levels are divided into "safe", "early warning", "danger" and "emergency", and the operation mode of the electric heat tracing system is adjusted according to different levels.

[0027] Step S5: Based on the data collected in Step S1 and the data processed in Step S2, eight-dimensional normalized time series data is formed. The real-time TTF value is calculated based on the seawater freezing prediction model trained in Step S3 using machine learning LSTM. The real-time TTF value is judged according to the risk level in Step S4, and the operation mode of the electric heat tracing system is dynamically adjusted according to the risk level.

[0028] Preferably, in step S1, the data acquisition frequency of the signal acquisition card 10 is once every 10 seconds.

[0029] Preferably, step S2 further includes data preprocessing, which includes: when the surface wind speed exceeds 40 m / s, using an anti-interference filtering algorithm to remove outliers in the data and using interpolation to fill in missing data; mapping sensor data collected at different locations in the pipeline assembly to a unified spatial grid and time grid through pipeline distance coordinates and acquisition time scale; calculating the rate of temperature change over time and the pipeline's overall heat transfer coefficient; and performing Min-Max normalization on the six test parameters and two calculation parameters.

[0030] In the site selection step S3, the specific steps include using eight-dimensional real-time normalized time series data (including pipe outer wall temperature, pipe seawater flow velocity, pipe seawater temperature, seawater salinity, surface wind speed, ambient temperature, temperature change rate, and comprehensive heat transfer coefficient) as input parameters; constructing an LSTM recurrent neural network with the minimum freezing time of static water as the lower limit of seawater freezing time margin and the constraint that the TTF of flowing water with a velocity greater than 1.8 m / s tends to infinity; dividing the measured data into training and testing sets according to a certain ratio, and training a seawater freezing time margin prediction model by combining mean square error and constraint penalty terms.

[0031] Preferably, in step S4, the predicted TTF value is based on equilibrium thermodynamic characteristics, system response capability, safety redundancy, and freezing data of the water intake pipeline at Zhongshan Station in Antarctica, and the risk of seawater freezing in the water intake pipeline is divided into four levels: "safe", "early warning", "dangerous" and "emergency".

[0032] Preferably, the operating mode of the electric heat tracing system in step S5 specifically includes:

[0033] When TTF > 240 minutes, it is considered "safe" and the electric heat tracing system will not work;

[0034] When 60 minutes ≤ TTF ≤ 240 minutes, it is judged as "early warning", and only the high-risk section of the pipeline is heated with low power;

[0035] When 30 minutes ≤ TTF < 60 minutes, it is judged as "dangerous" and the entire pipe section is heated at 50% power; at the same time, the high-risk section is heated at full power.

[0036] When TTF < 30 minutes, it is determined to be "emergency", the electric heat tracing system operates at full power and hydraulic pulse flushing is activated to prevent ice crystal blockage.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1. Using the freezing time margin (TTF) as a direct control indicator to replace the traditional temperature threshold control can appropriately widen the heating response range, avoid frequent start-ups and shutdowns of the electric heat tracing system, and improve energy efficiency.

[0039] 2. By utilizing the LSTM algorithm from machine learning, the risk of freezing can be predicted in advance, replacing the post-event response in traditional control strategies, and realizing the strategy shift from "passive insulation" to "active antifreeze".

[0040] 3. Taking into account multiple environmental factors affecting seawater freezing and pipeline heat loss, the environmental adaptability of the electric heat tracing system has been enhanced.

[0041] 4. By using the TTF freezing risk classification response mechanism, the sectional and variable power control of pipeline electric heat tracing can be realized, thereby reducing the risk of freezing accidents and the energy consumption of electric heat tracing.

[0042] In summary, this invention proposes a pipeline antifreeze system based on seawater freezing time margin and its application method. Leveraging the powerful performance of machine learning in prediction and control, an LSTM neural network seawater freezing prediction model is established, upgrading the traditional "temperature threshold control" to freezing risk time prediction control. This enables intelligent control of the electric heating tape, achieving the goals of energy saving, improved response speed, and enhanced environmental adaptability. Attached Figure Description

[0043] Figure 1 A flowchart illustrating the application method of a pipeline antifreeze system based on seawater freezing time margin, provided for an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of the laying of electric heating tape and arrangement of multi-parameter sensors for a pipeline antifreeze system based on seawater freezing time margin provided for an embodiment of the present invention.

[0045] The serial numbers in the diagram are as follows:

[0046] 1-1. Water intake pipe section; 1-2. Pipeline test section; 3. Pipeline outer wall temperature sensor; 4. Seawater temperature sensor; 5. Electromagnetic flow meter; 6. Seawater salinity sensor; 7. Ultrasonic anemometer; 8. Ambient temperature sensor; 9. Radiation protection ventilation cover; 10. Signal acquisition card; 11. Computer; 12. Pipeline insulation layer. Detailed Implementation

[0047] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0048] like Figure 2As shown, this embodiment provides a pipeline antifreeze system based on seawater freezing time margin, which is suitable for intelligent temperature control of pipeline components. It is characterized by including an electric heat tracing component, a detection component, a signal acquisition card 10, an environmental monitoring system, and a computer 11.

[0049] In the pipeline assembly, any one pipe segment is designated as pipeline test segment 1-1, and the remaining pipe segments are designated as water intake pipe segments 1-2. Alternatively, a segment in water intake pipe segment 1-2 can be replaced as pipeline test segment 1-1. All pipe segments are connected by flanges.

[0050] The electric heat tracing assembly includes an electric heat tracing tape 2 spirally wound and laid on the outer wall of the water intake pipe section 1-2, used to achieve intelligent temperature control of the water intake pipe section 1-2. Furthermore, in this embodiment, the pipe assembly is wrapped with a pipe insulation layer 12, used to cooperate with the electric heat tracing assembly for heat preservation.

[0051] The detection component includes pipe outer wall temperature sensors 3 that are uniformly and tightly attached to the outer wall of the water intake pipe section 1-2. Each pipe outer wall temperature sensor 3 is spaced apart in the spiral gap formed by the electric heating tape 2 on the water intake pipe section 1-2, and is used to detect the pipe wall temperature of the water intake pipe section 1-2 in real time.

[0052] In addition, a seawater temperature sensor 4, an electromagnetic flow meter 5, and a seawater salinity sensor 6 are centrally installed in the pipeline test section 1-1.

[0053] Table 1 provides the sensor models available for each data acquisition, including but not limited to the sensor types shown.

[0054] Table 1 Summary of Sensors

[0055] Data to be tested Sensor type sampling frequency Pipe outer wall temperature, °C Surface mount PT100 platinum resistance temperature sensor 10 seconds / time Seawater flow velocity inside the pipe, m / s Electromagnetic flowmeter 10 seconds / time Seawater temperature inside the pipe, °C PT100 Platinum Resistance Temperature Sensor 10 seconds / time Seawater salinity, PSU Flow-through conductivity sensor 10 seconds / time Surface wind speed, m / s Ultrasonic anemometer 10 seconds / time Ambient temperature, °C PT100 Platinum Resistance Temperature Sensor 10 seconds / time

[0056] The environmental monitoring system includes an ultrasonic anemometer 7 and an ambient temperature sensor 8.

[0057] The detection component and the environmental monitoring system are connected to the signal acquisition card 10 via cables. The signal acquisition card 10 is connected to the computer 11 for real-time data aggregation, display and storage.

[0058] According to claim 1, a pipeline antifreeze system based on seawater freezing time margin is characterized in that the environmental monitoring component is installed inside the radiation protection ventilation hood 9, and the surface of the radiation protection ventilation hood 9 is provided with an electric heating defrosting device.

[0059] like Figure 1 As shown, this embodiment also provides an application method for a pipeline antifreeze system based on seawater freezing time margin, including the following steps:

[0060] Step S1: Laying of electric heating tape and installation of sensors:

[0061] A test section 1-1 and multiple intake pipe sections 1-2 are connected by flanges, with the test section 1-1 located in the middle of the multiple intake pipe sections 1-2; and an electric heating tape 2 is spirally wound and laid on the outer wall of the intake pipe section 1-2 in the pipe assembly. Multiple pipe outer wall temperature sensors 3 are installed at intervals in the spiral gaps formed on the intake pipe section 1-2. A seawater temperature sensor 4, an electromagnetic flow meter 5, and a seawater salinity sensor 6 are centrally installed in the test section 1-1.

[0062] The pipe outer wall temperature sensor 3, seawater temperature sensor 4, electromagnetic flow meter 5, seawater salt concentration sensor 6, ultrasonic anemometer 7, and ambient temperature sensor 8 are connected to signal acquisition card 10 via cables. The data acquisition frequency of signal acquisition card 10 is once every 10 seconds, and the acquired data is transmitted to computer 11 via communication connection for real-time data acquisition.

[0063] Step S2: Data Acquisition and Preprocessing: Six parameters of the pipeline components are obtained from the real-time data acquired in Step S1. These six parameters include the pipeline outer wall temperature, seawater flow velocity, seawater temperature, seawater salinity, surface wind speed, and ambient temperature. Outliers and missing values ​​are removed, interpolation is performed, spatial-temporal grid mapping is applied, and the temperature change rate and comprehensive heat transfer coefficient are calculated. Finally, the data are normalized. These parameters affect the nonlinear process of seawater freezing or the heat loss of the pipeline.

[0064] Data preprocessing includes: when the surface wind speed exceeds 40 m / s, an anti-interference filtering algorithm is used to remove outliers from the data, and interpolation is used to fill in missing data; sensor data collected at different locations in the pipeline assembly are mapped to a unified spatial and temporal grid using pipeline distance coordinates and acquisition time scales; the rate of temperature change over time and the overall heat transfer coefficient of the pipeline are calculated; and Min-Max normalization is performed on six test parameters and two calculation parameters.

[0065] Step S3: Construction of a seawater freezing prediction model based on machine learning LSTM: Using eight-dimensional real-time normalized time series data, including pipe outer wall temperature, seawater flow velocity inside the pipe, seawater temperature inside the pipe, seawater salinity, surface wind speed, ambient temperature, temperature change rate, and comprehensive heat transfer coefficient, as input parameters; using the minimum static water freezing time as the lower limit of the seawater freezing time margin (TTF), and the constraint that the TTF of flowing water with a flow velocity greater than 1.8 m / s tends to infinity, an LSTM recurrent neural network is constructed; the measured data is divided into training and testing sets according to a certain ratio, and the seawater freezing time margin (TTF) prediction model is trained by combining the mean squared error (MSE) and the constraint penalty term (applying additional loss to the predicted value that violates the static water lower limit).

[0066] By using the freezing time margin (TTF) as a direct control indicator to replace the traditional temperature threshold control, the heating response range can be appropriately widened, effectively avoiding frequent start-ups and shutdowns of the electric heat tracing system and improving energy efficiency.

[0067] By utilizing the LSTM algorithm from machine learning, freezing risks can be predicted in advance, replacing the reactive response in traditional control strategies and realizing a strategy shift from "passive insulation" to "active antifreeze".

[0068] Step S4: Seawater Freezing Time Risk Classification: Based on the predicted real-time TTF value obtained from the seawater freezing time margin (TTF) prediction model, the predicted TTF value is used to classify the seawater freezing risk of the water intake pipeline into four levels: "Safe," "Warning," "Danger," and "Emergency," taking into account equilibrium thermodynamic characteristics, system response capability, safety redundancy, and freezing data of the water intake pipeline at Zhongshan Station in Antarctica (all of the above data are existing data and will not be elaborated here). The TTF value range for each level is shown in Table 2.

[0069] Table 2 TTF Risk Classification

[0070] TTF range Risk level Control Action TTF > 240min Safety Turn off electric heat tracing 60min < TTF ≤ 240min Warning High-risk pipe sections require 30% power heating 30min < TTF ≤ 60min Danger 50% power for the entire section + 100% power for the elbows TTF ≤ 30min urgent Full power + hydraulic pulse

[0071] Step S5: Based on the data collected in Step S1 and the data processed in Step S2, eight-dimensional normalized time series data is formed. The real-time TTF value is calculated based on the seawater freezing prediction model of machine learning LSTM trained in Step S3. The real-time TTF value is judged according to the risk level in Step S4, and the operation mode of the electric heat tracing system is dynamically adjusted according to the risk level.

[0072] The above-mentioned electric heat tracing system operation modes specifically include:

[0073] When TTF > 240 minutes, it is considered "safe". This means that under the current operating conditions, the time for the coldest point of the pipeline to freeze completely is greater than 240 minutes. The time required for the pipeline to freeze and become blocked is relatively long, and the electric heat tracing system will not work.

[0074] When 60 minutes ≤ TTF ≤ 240 minutes, it is judged as "warning". At this time, the high-risk section of the pipeline, such as elbows and joints, will freeze after 60 minutes, while the freezing time of other pipe sections is longer. Only the high-risk section of the pipeline is heated with low power.

[0075] When 30 minutes ≤ TTF < 60 minutes, it is judged as "dangerous". At this time, high-risk sections such as bends in the pipeline will freeze in about 30 minutes, and the remaining sections will freeze within 60 minutes. The risk of freezing and blockage of the pipeline is high. The entire pipeline section is heated at 50% power; at the same time, high-risk sections are heated at full power.

[0076] When TTF < 30 minutes, it is judged as "emergency". At this time, the freezing time of the entire water intake pipe is less than 30 minutes, the risk of freezing and blockage is extremely high, the electric heat tracing system runs at full power and the hydraulic pulse flushing is activated to prevent ice crystal blockage.

[0077] By using the Freezing Time Margin (TTF) freezing risk classification response mechanism, the sectional and variable power control of pipeline electric heat tracing can be achieved, thereby reducing the risk of freezing accidents while reducing the energy consumption of electric heat tracing.

[0078] In summary, this embodiment leverages the powerful performance of machine learning in prediction and control to establish an LSTM neural network-based seawater freezing prediction model. This upgrades the traditional "temperature threshold control" to freezing risk time prediction control, thereby achieving intelligent temperature control of the electric heating cable and realizing the goals of saving energy, improving response speed, and enhancing environmental adaptability.

[0079] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0080] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0081] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A freeze protection system for a pipeline based on the time margin to freezing of seawater, suitable for intelligent temperature control of a pipeline assembly, characterized in that, The electric heat tracing assembly, the detection assembly, the signal acquisition card (10), the environment monitoring system and the computer (11); The pipeline assembly comprises at least one pipeline test section (1-1) and a plurality of water taking pipe sections (1-2), and each pipe section is connected through a flange; The electric heat tracing assembly comprises an electric heat tracing belt (2) spirally wound and laid on the outer wall surface of the water taking pipe section (1-2); The detection assembly comprises pipeline outer wall surface temperature sensors (3) uniformly and closely arranged on the outer wall surface of the water taking pipe section (1-2), and each pipeline outer wall surface temperature sensor (3) is arranged in the spiral gap formed by the electric heat tracing belt (2) on the water taking pipe section (1-2); and a seawater temperature sensor (4), an electromagnetic flowmeter (5) and a seawater salinity sensor (6) are centrally installed in the pipeline test section (1-1); The environment monitoring system comprises an ultrasonic wind speed meter (7) and an environmental temperature sensor (8); The detection assembly and the environment monitoring system are respectively connected to the signal acquisition card (10) through cables, and the signal acquisition card (10) is in communication connection with the computer (11).

2. A freeze protection system for a pipeline based on a time margin to freezing of seawater according to claim 1, wherein, The pipeline assembly is wrapped with a pipeline heat preservation layer (12).

3. A freeze protection system for a pipe based on a time margin to freezing of seawater according to claim 1, wherein, The environment monitoring assembly is installed in a radiation-proof ventilation cover (9), and the surface of the radiation-proof ventilation cover (9) is provided with an electric heating defrosting device.

4. A method of using the seawater-based freeze time margin-based pipe freeze protection system according to any one of claims 1 to 3, characterized in that, The method comprises the following steps: Step S1: electric heat tracing belt laying and sensor installation: At least one pipeline test section (1-1) and a plurality of water taking pipe sections (1-2) are connected through a flange; and an electric heat tracing belt (2) is spirally wound and laid on the outer wall surface of the water taking pipe section (1-2) in the pipeline assembly, and a plurality of pipeline outer wall surface temperature sensors (3) are arranged in the spiral gap formed by the electric heat tracing belt (2) on the water taking pipe section (1-2); and a seawater temperature sensor (4), an electromagnetic flowmeter (5) and a seawater salinity sensor (6) are centrally installed in the pipeline test section (1-1); The pipeline outer wall surface temperature sensor (3), the seawater temperature sensor (4), the electromagnetic flowmeter (5), the seawater salinity sensor (6), the ultrasonic wind speed meter (7) and the environmental temperature sensor (8) are connected to the signal acquisition card (10) through cables, and the collected data is transmitted to the computer (11) in communication connection for real-time data collection; Step S2: data collection: at least six parameters of the pipeline assembly are obtained from the data collected in real time in step S1, the six parameters include pipeline outer wall surface temperature, seawater flow rate, seawater temperature, seawater salinity, ground surface wind speed and environmental temperature data, and abnormal values, missing values, space-time grid mapping, temperature change rate and comprehensive heat exchange coefficient are eliminated, and finally normalized processing is performed; Step S3: seawater freezing prediction model construction based on machine learning LSTM: eight-dimensional time series data after normalization processing, including pipeline outer wall surface temperature, seawater flow rate, seawater temperature, seawater salinity, ground surface wind speed, environmental temperature, temperature change rate and comprehensive heat exchange coefficient, are taken as input, an LSTM neural network is constructed in combination with static water freezing time lower limit and flowing water freezing time constraint, and a seawater freezing time margin (TTF) prediction model is obtained through training; Step S4: Sea water freezing time risk classification: dividing the predicted TTF real-time value obtained based on the sea water freezing time margin (TTF) prediction model into four risk levels of "safe", "warning", "dangerous", and "emergency", and adjusting the electric heat tracing system operation mode based on different levels; Step S5: forming eight-dimensional normalized time series data based on the data collected in step S1 and the data processed in step S2, and calculating the TTF real-time value based on the sea water freezing prediction model of the trained machine learning LSTM in step S3; the TTF real-time value is judged according to the risk level in step S4, and the electric heat tracing system operation mode is dynamically adjusted according to the risk level.

5. A method of applying a pipe freeze protection system based on the time margin to freezing of seawater according to claim 4, characterized in that, In the step S1, the data acquisition frequency of the signal acquisition card (10) is 10s once.

6. The method of claim 4, wherein the method further comprises: The step S2 further includes data preprocessing, which includes: when the local surface wind speed exceeds 40m / s, using an anti-interference filtering algorithm to remove abnormal values in the data, and completing the missing data by interpolation method; mapping the sensor data collected at different positions in the pipeline assembly to a unified spatial grid and time grid through pipeline distance coordinates and collection time scale; calculating the temperature change rate and the comprehensive heat exchange coefficient of the pipeline; and performing Min-Max normalization processing on the six test parameters and two calculation parameters.

7. The method of claim 4, wherein the method further comprises: In the step S3, specifically, the eight-dimensional real-time normalized time series data of the pipeline outer wall temperature, the pipeline inner sea water flow rate, the pipeline inner sea water temperature, the sea water salt concentration, the surface wind speed and the environment temperature, and the temperature change rate and the comprehensive heat exchange coefficient are taken as input parameters; the static water minimum freezing time is taken as the lower limit of the sea water freezing time margin, and the TTF of the flowing water with a flow rate greater than 1.8m / s tends to infinity as a constraint condition, to construct an LSTM recurrent neural network; the measured data is divided into a training set and a test set according to a certain proportion, and a sea water freezing time margin prediction model is trained by combining the mean square error and the constraint penalty term.

8. The method of claim 4, wherein the method is used for a pipe freeze protection system based on the time margin of seawater freezing. In the step S4, the predicted TTF value is divided into four levels of "safe", "warning", "dangerous", and "emergency" based on the balance thermodynamic characteristics, system response capability, safety redundancy, and the freezing data of the water intake pipeline in the Antarctic Zhongshan Station.

9. The method of claim 4, wherein the method is used for a pipe freeze protection system based on the time margin of seawater freezing. The electric heat tracing system operation mode in the step S5 specifically includes: When TTF>240 minutes, it is determined as "safe", and the electric heat tracing system does not work; When 60 minutes≤TTF≤240 minutes, it is determined as "warning", and only the high-risk section of the pipeline is heated at low power; When 30 minutes≤TTF<60 minutes, it is determined as "dangerous", and the whole pipeline section is heated at 50% power; at the same time, the high-risk section is heated at full power; When TTF<30 minutes, it is determined as "emergency", and the electric heat tracing system operates at full power, and the hydraulic pulse flushing is started to prevent ice crystal blockage.

Images