Intelligent monitoring method for freezing pipe leakage in freezing method construction

Abstract

The invention discloses an intelligent monitoring method for freezing pipe leakage in freezing method construction, and the method comprises the steps: collecting a negative pressure wave signal generated near a freezing pipe leakage point and the position and size information of the freezing pipe leakage point corresponding to the negative pressure wave signal through an indoor test; performing de-noising processing and feature extraction on the acquired negative pressure wave signal, and constructing and training an LSTM model; during on-site freezing method construction, pressure sensors are installed at the two ends of a freezing pipe; a negative pressure wave signal of the freezing pipe is monitored through a pressure sensor, denoising processing and feature extraction are conducted on the monitoring signal, the trained LSTM model is imported in real time, freezing pipe state information is output, and real-time monitoring of leakage of the freezing pipe is achieved; and according to a monitoring result, by integrating multi-factor analysis and consideration, proper construction treatment measures are taken for the fracture of the freezing pipe. The method can help a project to monitor the state of the freezing pipe in real time, quickly discover the position and size of a leakage point, improve the tunneling operation efficiency and guarantee the tunnel construction safety.

Description

Technical Field

[0001] The invention relates to the technical field of tunnel construction and discloses an intelligent monitoring method for freezing pipe leakage in freezing method construction. Background Art

[0002] In recent years, with the continuous development of urbanization and the increasing demand for underground space, urban underground projects have made great progress. In underground engineering construction, the freezing method is widely used because of its good water sealing, environmental friendliness and carrying capacity. The refrigerant medium in the freezing construction process is mostly low-temperature CaCl2 solution, which takes heat out of the stratum by flowing in the freezing pipe. During the freezing construction process, especially when the freezing pipe passes through multi-layer composite strata, freezing pipe fracture accidents often occur. The fracture of the freezing pipe will cause the low-temperature brine circulating inside it to leak. On the one hand, it will damage the frozen wall and affect the safety of construction. In severe cases, it will also cause water in the soil layer to enter the construction working surface, causing the freezing construction to fail. On the other hand, the leakage of brine will cause soluble salts to enter the soil layer, which will reduce the initial freezing temperature of the soil body, prolong the subsequent freezing time, and even cause frozen soil to be unable to form, resulting in the failure of frozen soil construction, causing irreversible impact and loss. How to effectively monitor the leakage of freezing pipes in real time is a technical problem that has long plagued the freezing construction. Summary of the invention

[0003] In order to solve the above-mentioned technical deficiencies in the prior art for real-time monitoring of freezing pipe leakage during freezing construction, the present invention provides an intelligent monitoring method for freezing pipe leakage during freezing construction.

[0004] In order to achieve the above technical purpose, the technical solution adopted by the present invention is:

[0005] An intelligent monitoring method for freezing pipe leakage during freezing construction includes the following steps:

[0006] S1: Through indoor tests, collect the negative pressure wave signal generated near the freezing pipe leakage point and its corresponding freezing pipe leakage point location and size information;

[0007] S2: De-noise and extract features of the collected negative pressure wave signal, build and train an LSTM model, and the trained LSTM model can map the location and size information of the freezing pipe leakage through the freezing pipe leakage features;

[0008] S3: During the on-site freezing method construction, pressure sensors are installed at both ends of the freezing pipe to monitor whether there is leakage in the freezing pipe and the location and size of the leakage;

[0009] S4: monitor the negative pressure wave signal of the frozen pipe through the pressure sensor, perform denoising and feature extraction on the monitoring signal, import it into the LSTM model trained in step S2 in real time, output the frozen pipe status information, and realize real-time monitoring of frozen pipe leakage;

[0010] S5: Based on the monitoring results and comprehensive multi-factor analysis, appropriate construction treatment measures are taken for frozen pipe ruptures.

[0011] Furthermore, in step S1, the indoor test is a freezing pipe leakage monitoring simulation test. The test device used in the indoor test includes a refrigerator and a freezing pipe. The freezing pipe is connected to the refrigerator from end to end. Multiple ball valves are installed on the freezing pipe to simulate different leakage states of the pores. At least two pressure sensors are provided on the freezing pipe. The two pressure sensors are respectively located on the front and rear sides of the ball valve. The pressure sensors are communicatively connected to the data acquisition system.

[0012] Preferably, the pressure sensor is a MIK-P300G pressure sensor.

[0013] Furthermore, in step S2, the denoising process is based on the EEMD-PRT algorithm, including the original signal being subjected to inverse Fourier transform processing to obtain a random phase, the obtained multi-scale signal being subjected to EEMD processing, and then being subjected to a bandpass filter and a Hilbert transform, and finally reconstructing the signal to complete the denoising process; the feature extraction uses a stacked autoencoder to extract the frozen pipe leakage features in the denoising signal.

[0014] Further, in step S3, the method for installing pressure sensors at both ends of the freezing pipe is: after the freezing pipe is installed and before the refrigerator starts working, the pressure sensor is welded to both ends of the freezing pipe, and after the installation is completed, the line is connected and debugged to ensure that the pressure sensor works normally.

[0015] Furthermore, in step S4, the denoising and feature extraction are the same as those in step S2, and the processed data is imported into the trained LSTM model in the computer in real time via a wired manner. The trained LSTM model realizes the mapping of the frozen pipe leakage characteristics to the frozen pipe leakage position and size information through learning and analysis.

[0016] Furthermore, in step S5, the comprehensive multi-factor analysis includes but is not limited to analyzing and considering the freezing pipe leakage position, the freezing pipe crack size, the pipeline liquid supply pressure and the leakage amount.

[0017] Furthermore, the comprehensive multi-factor analysis also includes predictive calculation of single-point leakage, and taking appropriate construction treatment measures for frozen pipe breakage according to the single-point leakage; including: if during the construction process, the predicted value of the single-point leakage is less than the preset value, it can be preliminarily judged that the impact on the thickness and strength of the frozen wall is small, and it is recommended to maintain the original freezing design; if the predicted value of the single-point leakage exceeds the preset value, it is recommended to stop construction, take casing measures for the frozen pipe that has been broken, and resume freezing construction after the treatment is completed.

[0018] Preferably, the preset value range of the single-point leakage volume is 10 to 30L.

[0019] The beneficial effects of the present invention are:

[0020] The present invention monitors the frozen pipe leakage data through a pressure sensor, and applies a deep learning network LSTM (Long Short-Term Memory Network) to the frozen pipe leakage status monitoring. It can retain distant context information when processing data, and can better realize the comprehensive processing and analysis of signals, help the project to monitor the frozen pipe status in real time, quickly discover the location and size of the leakage point, improve the efficiency of tunnel excavation operations, and ensure the safety of tunnel construction.

[0021] In the present invention, the denoising process is based on the EEMD-PRT algorithm, which can improve the signal-to-noise ratio and correlation coefficient of the noisy signal, reduce the mean square error and mean absolute error, and better capture the local characteristics and detail information of the signal compared with the traditional denoising method.

[0022] The present invention comprehensively analyzes and considers multiple factors based on the monitoring results, including judging the leakage position of the freezing pipe, the size of the crack in the freezing pipe, the pipeline liquid supply pressure, the leakage amount and the predicted value of the single-point leakage amount of the freezing pipe, and gives actual construction suggestions. It takes into account both construction safety and actual construction progress factors, and has guiding significance for construction using the freezing method. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0024] Figure 1 The present invention is a construction process flow chart of the method.

[0025] Figure 2 It is a schematic diagram of the structure of the indoor test device in the present invention.

[0026] Figure numerals: 1-refrigeration machine, 2-freezing pipe, 3-ball valve, 4-pressure sensor, 5-data acquisition system. DETAILED DESCRIPTION

[0027] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians of the art without making creative work are within the scope of protection of the present application.

[0028] An intelligent monitoring method for freezing pipe leakage during freezing construction, such as Figure 1 As shown, the following steps are included:

[0029] S1: Through indoor testing, collect the negative pressure wave signal generated near the freezing pipe leakage point and its corresponding freezing pipe leakage point location and size information. The indoor test is to conduct a freezing pipe leakage monitoring simulation test. The test device used in the indoor test is as follows: Figure 2 As shown, it mainly includes a refrigerator 1 and a freezing pipe 2. The freezing pipe 2 is connected to the refrigerator 1 from end to end. A plurality of ball valves 3 are installed on the freezing pipe 2 to simulate different leakage states of the pores. At least two pressure sensors 4 are provided on the freezing pipe 1. The two pressure sensors 4 are respectively located at the front and rear sides of the ball valve 3. The pressure sensors 4 are communicatively connected to a data acquisition system 5. Preferably, the pressure sensor 4 is a MIK-P300G pressure sensor.

[0030] S2: De-noise and extract features of the collected negative pressure wave signal, build and train an LSTM (long short-term memory network) model. The trained LSTM model can map the location and size information of the frozen pipe leakage through the frozen pipe leakage characteristics.

[0031] The denoising process is based on the EEMD-PRT algorithm. The main process is that the original signal is processed with a random phase through an inverse Fourier transform, and the obtained multi-scale signal is processed with EEMD, and then through a bandpass filter and a Hilbert transform, and finally the signal is reconstructed to complete the denoising process. The feature extraction uses a stacked autoencoder to extract the frozen pipe leakage features in the processed signal. The LSTM (long short-term memory network) model used in this application is a recurrent neural network for processing and predicting time series data. It controls the flow of information through a gating mechanism, so that the network can learn the long-term dependencies in the sequence, and is highly applicable to the monitoring research of such frozen pipe leakage.

[0032] S3: During the on-site freezing method construction, pressure sensors are installed at both ends of the freezing pipe to monitor whether there is leakage in the freezing pipe and the location and size of the leakage. The method of installing the pressure sensor at both ends of the freezing pipe is as follows: after the freezing pipe is installed and before the refrigerator starts working, the pressure sensor is directly welded to both ends of the freezing pipe, and after the installation is completed, the line is connected and debugged to ensure that the pressure sensor works normally.

[0033] S4: Monitor the signal of the frozen pipe through the pressure sensor, denoise and extract features of the monitoring signal, import the trained LSTM model in real time, output the frozen pipe status information, and realize real-time monitoring of frozen pipe leakage.

[0034] The denoising and feature extraction in this step are the same as those in step S2. The processed data is imported into the trained LSTM model in the computer in real time via wired means. The trained LSTM model realizes the mapping of the freezing pipe leakage characteristics to the freezing pipe leakage position and size information through learning and analysis.

[0035] S5: Based on the monitoring results and comprehensive multi-factor analysis, appropriate construction treatment measures are taken for frozen pipe ruptures.

[0036] The comprehensive multi-factor analysis and consideration takes appropriate construction treatment measures for the frozen pipe break. The multi-factors include but are not limited to analyzing and considering factors such as the frozen pipe leakage location, the size of the frozen pipe crack, the pipeline liquid supply pressure and the leakage amount, and also includes predicting and calculating the single-point leakage. If the single-point leakage prediction value is less than the preset value during the construction process, it can be preliminarily judged that the impact on the thickness and strength of the frozen wall is small, and it is recommended to maintain the original freezing design; if the single-point leakage prediction value exceeds the preset value, it is recommended to stop the construction and take casing measures for the frozen pipe that has broken, that is, put a bottom-sealed and slightly smaller diameter frozen pipe under the mother pipe to reconstruct a new freezer. After the treatment is completed, the freezing construction can be resumed. The single-point leakage preset value can be preset according to the actual construction environment and situation. The single-point leakage preset value ranges from 10 to 30L. Preferably, the single-point leakage preset value is set to 20L.

[0037] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, technicians familiar with the field may make various corresponding changes and deformations based on the present invention, but these corresponding changes and deformations should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. An intelligent monitoring method for freezing pipe leakage during freezing construction, characterized in that: The steps include: S1: Through indoor tests, collect the negative pressure wave signal generated near the freezing pipe leakage point and its corresponding freezing pipe leakage point location and size information; S2: De-noise and extract features of the collected negative pressure wave signal, build and train an LSTM model, and the trained LSTM model can map the location and size information of the freezing pipe leakage through the freezing pipe leakage features; S3: During the on-site freezing method construction, pressure sensors are installed at both ends of the freezing pipe to monitor whether there is leakage in the freezing pipe and the location and size of the leakage; S4: monitor the negative pressure wave signal of the frozen pipe through the pressure sensor, perform denoising and feature extraction on the monitoring signal, import it into the LSTM model trained in step S2 in real time, output the frozen pipe status information, and realize real-time monitoring of frozen pipe leakage; S5: Based on the monitoring results and comprehensive multi-factor analysis, appropriate construction treatment measures are taken for frozen pipe ruptures.

2. The intelligent monitoring method for freezing pipe leakage in freezing construction according to claim 1 is characterized in that: In step S1, the indoor test is a freezing pipe leakage monitoring simulation test. The test device used in the indoor test includes a refrigerator and a freezing pipe. The freezing pipe is connected to the refrigerator from end to end. Multiple ball valves are installed on the freezing pipe to simulate different leakage states of the pores. At least two pressure sensors are provided on the freezing pipe. The two pressure sensors are respectively located on the front and rear sides of the ball valve. The pressure sensors are communicated and connected to the data acquisition system.

3. The intelligent monitoring method for freezing pipe leakage in freezing construction according to claim 2 is characterized in that: The pressure sensor is a MIK-P300G pressure sensor.

4. The intelligent monitoring method for freezing pipe leakage during freezing construction according to claim 1 is characterized in that: In step S2, the denoising process is based on the EEMD-PRT algorithm, including the original signal being subjected to inverse Fourier transform to process the random phase, the obtained multi-scale signal being subjected to EEMD processing, and then being subjected to a bandpass filter and Hilbert transform, and finally reconstructing the signal to complete the denoising process; the feature extraction uses a stacked autoencoder to extract the frozen pipe leakage features in the denoising signal.

5. The intelligent monitoring method for freezing pipe leakage during freezing construction according to claim 1 is characterized in that: In step S3, the method for installing pressure sensors at both ends of the freezing pipe is as follows: after the freezing pipe is installed and before the refrigerator starts working, the pressure sensors are welded to both ends of the freezing pipe, and after the installation is completed, the lines are connected and debugged to ensure that the pressure sensors work normally.

6. The intelligent monitoring method for freezing pipe leakage during freezing construction according to claim 4 is characterized in that: In step S4, the denoising and feature extraction are the same as those in step S2. The processed data is imported into the trained LSTM model in the computer in real time via a wired manner. The trained LSTM model realizes the mapping of the frozen pipe leakage characteristics to the frozen pipe leakage position and size information through learning and analysis.

7. The intelligent monitoring method for freezing pipe leakage in freezing construction according to claim 1 is characterized by: In step S5, the comprehensive multi-factor analysis includes but is not limited to analyzing and considering the freezing pipe leakage position, the size of the freezing pipe crack, the pipeline liquid supply pressure and the leakage amount.

8. The intelligent monitoring method for freezing pipe leakage in freezing construction according to claim 7 is characterized in that: The comprehensive multi-factor analysis also includes predicting and calculating single-point leakage, and taking appropriate construction treatment measures for frozen pipe fractures based on the single-point leakage; including: if during the construction process, the predicted value of the single-point leakage is less than the preset value, it can be preliminarily determined that the impact on the thickness and strength of the frozen wall is small, and it is recommended to maintain the original freezing design; if the predicted value of the single-point leakage exceeds the preset value, it is recommended to stop construction, take casing measures for the frozen pipe that has already broken, and resume freezing construction after the treatment is completed.

9. The intelligent monitoring method for freezing pipe leakage during freezing construction according to claim 8, characterized in that: The preset value range of the single point leakage volume is 10 to 30L.

Images