Monitoring device for clogging condition of longitudinal drainage system of in-service highway tunnel
Through the signal acquisition and processing module combined with the multimodal neural network model, the blockage of highway tunnel drainage system is monitored in real time, solving the defects of the existing technology being unable to detect blockage problems in a timely manner, and achieving safe and low-cost long-term monitoring and maintenance.
Patent Information
- Application Number
- CN202411027261.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-12
AI Technical Summary
The prior art cannot monitor and promptly deal with the blockage of longitudinal drainage systems in highway tunnels in real time, especially in water-rich karst tunnels, resulting in high risk of poor drainage and lining structure cracking.
The signal acquisition module is used to emit sound waves through a micro-acoustic wave generator, and the signal processing module performs feature extraction. Combined with the multimodal neural network model, the degree of pipeline blockage is predicted to achieve real-time monitoring and evaluation.
Real-time monitoring of the blockage status of highway tunnel drainage systems is realized, timely discovering and handling blockage problems, avoiding safety accidents, reducing maintenance costs, and improving monitoring safety.
Smart Images

Figure CN118934026A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of monitoring highway tunnel drainage systems, and in particular relates to a monitoring device for the clogging condition of a longitudinal drainage system of an in-service highway tunnel. Background Art
[0002] During the construction of highway mountain tunnels, since the strata often contain mudstone or limestone, after the tunnel is built, a large amount of mud or calcium carbonate sediments will often be deposited in the central longitudinal drainage pipe or the road side ditch, resulting in blockage of the drainage pipe or the side ditch. This blockage not only affects the normal operation of the drainage system, but also may cause the back of the lining structure to bear a large water pressure due to the inability to effectively drain the groundwater, which will eventually lead to cracks and seepage in the lining, endangering the safety of the structure. At present, the drainage system of highway tunnels is a concealed project. The central drainage ditch is generally buried 100 cm below the road surface in the tunnel. Usually, there is an inspection well every 200 meters to clear the blockage in the central drainage ditch. However, existing detection methods such as pipeline robots, industrial endoscopes, and sonar methods cannot achieve real-time monitoring, and cleaning is difficult, especially in water-rich karst tunnels. It is impossible to timely understand the blockage status of the drainage pipe or side ditch, which may lead to poor drainage during heavy rains, and the water pressure on the back of the lining increases sharply, causing the tunnel lining structure to collapse and sudden water inrush accidents. Summary of the invention
[0003] In order to solve the above technical problems, the present invention provides a monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel, comprising:
[0004] A signal acquisition module is used to collect sound wave signals to determine the blockage of drainage pipes;
[0005] A signal processing module, connected to the signal acquisition module, for extracting features from the acoustic wave signal to obtain a feature signal;
[0006] The clogging monitoring module is connected to the signal processing module and is used to input the characteristic signal into a multimodal neural network model to predict the degree of pipeline clogging and obtain clogging monitoring results in real time.
[0007] Preferably, the signal acquisition module includes a signal transmitting unit, a signal receiving unit, and a time calculation unit;
[0008] The signal transmitting unit is used to emit sound waves of a specified frequency and a specified direction through a micro sound wave generator;
[0009] The signal receiving unit is used to receive sound waves reflected from pipeline obstacles;
[0010] The time calculation unit is used to calculate the time interval between the micro-sound wave generator sending and receiving the sound wave.
[0011] Preferably, the micro-sound wave generators are arranged on the survey line at preset intervals and have a physical positioning function, which is achieved through position coding;
[0012] The measuring lines are symmetrically arranged on the top and both sides of the drainage pipe;
[0013] A controller is provided at one end of the measuring line, and the controller is used to control the sound wave emission, reception and interpretation of the micro-sound wave generator.
[0014] Preferably, the signal processing module includes a preprocessing unit and a feature extraction unit;
[0015] The preprocessing unit is used to sequentially remove the mean, remove the linear trend, perform waveform pinch-out and perform band-pass filtering on the acoustic wave signal;
[0016] The feature extraction unit is used to extract time-frequency domain features of the preprocessed sound wave signal through wavelet transformation.
[0017] Preferably, the preprocessing unit includes a mean removal unit, a linear trend removal unit, a waveform pinch-out unit, and a bandpass filtering unit;
[0018] The de-averaging unit is used to remove the average value of the waveform data, eliminate the DC component of the data, and make the average amplitude of the processed data zero;
[0019] The delinear trend unit is used to eliminate the linear deflection in the data, fit the data into a straight line, and then subtract the linear trend represented by the straight line from the data;
[0020] The waveform pinch-out unit is used to gradually reduce the beginning and end of the waveform data from the original value to zero;
[0021] The bandpass filtering unit is used to retain the waveform within a specific frequency range, shield the signals in other frequency bands, obtain the effective signal frequency band and suppress noise.
[0022] Preferably, the congestion monitoring module includes a model building unit, a model training unit, a fusion unit, and a congestion monitoring unit;
[0023] The model building unit is used to build a convolutional neural network model and a recurrent neural network model;
[0024] The model training unit is used to train the convolutional neural network model and the recurrent neural network model respectively through the signal image data and the time series data;
[0025] The fusion unit is used to fuse the trained convolutional neural network model and the recurrent neural network model to obtain a multimodal neural network model; and is also used to integrate the signal image data and the time series data to obtain an integrated data set;
[0026] The blockage monitoring unit is used to analyze the integrated data through a multimodal neural network model to obtain the overall blockage result of the pipeline.
[0027] Preferably, the model construction unit includes a first network construction unit and a second network construction unit;
[0028] The first network construction unit is used to construct a convolutional neural network and generate a convolutional neural network group based on the convolutional neural network;
[0029] The second network construction unit is used to construct a recurrent neural network, and introduce an attention mechanism into the recurrent neural network for optimization to generate an improved recurrent neural network.
[0030] Preferably, the model training unit includes a first training unit, a second training unit, and a third training unit;
[0031] The first training unit includes an iterative calculation unit, which is used to iterate the convolutional neural network group through a genetic algorithm until the optimal model is obtained and the iteration is terminated to obtain a genetic neural network;
[0032] The second training unit includes a pre-training unit, an iterative training unit, and a transfer learning unit;
[0033] The pre-training unit is used to pre-train the improved recurrent neural network through the Transformer model to generate a pre-training model; the improved recurrent neural network generates an improved recurrent neural network by introducing the constructed attention mechanism module and residual module into the recurrent neural network;
[0034] The iterative training unit is used to iteratively train the pre-trained model through a sample set, and then optimize the trained model through an optimizer to generate an optimized model;
[0035] The transfer learning unit is used to freeze the bottom-level parameters of the optimization model, optimize the upper-level parameters of the optimization model, and generate a target recurrent neural network;
[0036] The third training unit is used to train the target recurrent neural network through the output of the genetic neural network to generate a multimodal neural network model.
[0037] Preferably, the fusion unit includes a fusion unit and a data fusion unit;
[0038] The data fusion unit comprises:
[0039] A weight generating unit, used for assigning weights to the signal image data and the time series data based on the importance of the data, to obtain image weight data and time series weight data;
[0040] A weight calculation unit, used for multiplying the image weight data and the time series weight data with the corresponding weights to obtain a weighted value;
[0041] The summing and integrating unit is used to add the weighted values to obtain the integrated data set.
[0042] Preferably, the blockage monitoring unit includes a data interpretation unit and a blockage assessment unit;
[0043] The data interpretation unit is used to calculate the distance between the liquid surface and each sound wave generator according to the time interval between the micro-sound wave generator sending and receiving the sound wave and the frequency wavelength of the micro-sound wave generator;
[0044] The blockage assessment unit is used to connect the distances obtained by each sound wave generator, estimate the liquid level in the pipeline, calculate the ratio of the liquid level to the pipeline height and input it into the multimodal neural network model, and predict the blockage degree of the pipeline based on the multimodal neural network model.
[0045] Compared with the prior art, the present invention has the following advantages and technical effects:
[0046] The present invention can monitor the blockage status of the drainage system in real time, discover and deal with the blockage problem in time, and avoid safety accidents caused by poor drainage.
[0047] The present invention can accurately monitor the height of the liquid level in the pipeline and accurately evaluate the blockage degree of the pipeline through the sound wave reflection recognition technology.
[0048] The device of the present invention is easy to install and operate, has low maintenance cost, and is suitable for long-term monitoring of in-service highway tunnels.
[0049] The present invention adopts a non-contact monitoring method, thereby avoiding secondary damage to the tunnel structure and improving the safety of monitoring.
[0050] The monitoring device of the present invention can effectively monitor the clogging condition of the longitudinal drainage system of an in-service highway tunnel, timely discover and deal with the clogging problem, and ensure the safe operation of the tunnel. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0052] Figure 1 A schematic diagram of the system structure of an embodiment of the present invention;
[0053] Figure 2 This is a schematic diagram of the internal measurement line structure of a pipeline according to an embodiment of the present invention;
[0054] Figure 3 Schematic diagram of the measurement line structure of an embodiment of the present invention. DETAILED DESCRIPTION
[0055] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0056] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0057] like Figure 1-3 As shown, in this embodiment, a monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel is provided, comprising:
[0058] A signal acquisition module is used to collect sound wave signals to determine the blockage of drainage pipes;
[0059] The signal processing module is connected to the signal acquisition module and is used to extract the characteristics of the sound wave signal to obtain a characteristic signal;
[0060] The blockage monitoring module is connected to the signal processing module and is used to input the characteristic signal into the multimodal neural network model to predict the degree of pipeline blockage and obtain the blockage monitoring results in real time.
[0061] Furthermore, the signal acquisition module includes a signal transmitting unit, a signal receiving unit, and a time calculation unit;
[0062] The signal transmitting unit is used to emit sound waves of a specified frequency and a specified direction through a micro sound wave generator;
[0063] The signal receiving unit is used to receive the sound waves reflected by the pipeline obstacles;
[0064] The time calculation unit is used to calculate the time interval between the micro-sound wave generator sending and receiving the sound wave.
[0065] Furthermore, if Figure 2 As shown, the micro-sound wave generators are arranged on the survey line at preset intervals and have a physical positioning function, which is achieved through position coding;
[0066] The measuring lines are symmetrically arranged at a certain distance on the top of the drainage pipe and on both sides;
[0067] like Figure 3 As shown, a controller is provided at one end of the measuring line, and the controller is used to control the sound wave emission, reception and interpretation of the micro-sound wave generator.
[0068] Further, the signal processing module includes a preprocessing unit and a feature extraction unit;
[0069] The preprocessing unit is used to perform mean removal, linear trend removal, waveform extinction and band-pass filtering on the acoustic wave signal in sequence;
[0070] The feature extraction unit is used to extract time-frequency domain features of the preprocessed sound wave signal through wavelet transformation.
[0071] Furthermore, the preprocessing unit includes a mean removal unit, a linear trend removal unit, a waveform pinch-out unit, and a bandpass filtering unit;
[0072] The de-averaging unit is used to remove the average value of the waveform data, eliminate the DC component of the data, and make the average amplitude of the processed data zero;
[0073] The delinear trend unit is used to eliminate linear skewness in the data by fitting the data to a straight line and then subtracting the linear trend represented by the straight line from the data;
[0074] The waveform pinch-out unit is used to gradually reduce the beginning and end of the waveform data from the original value to zero;
[0075] The bandpass filter unit is used to retain the waveform within a specific frequency range, shield the signals in other frequency bands, obtain the effective signal frequency band, suppress noise, and improve the signal-to-noise ratio of the signal.
[0076] Further, the congestion monitoring module includes a model building unit, a model training unit, a fusion unit, and a congestion monitoring unit;
[0077] The model building unit is used to build a convolutional neural network model and a recurrent neural network model;
[0078] The model training unit is used to train the convolutional neural network model and the recurrent neural network model respectively through the signal image data and the time series data;
[0079] The fusion unit is used to fuse the trained convolutional neural network model and the recurrent neural network model to obtain a multimodal neural network model; it is also used to integrate the signal image data and the time series data to obtain an integrated data set;
[0080] The blockage monitoring unit is used to analyze the integrated data through a multimodal neural network model to obtain the overall blockage results of the pipeline.
[0081] Further, the model construction unit includes a first network construction unit and a second network construction unit;
[0082] The first network construction unit is used to construct a convolutional neural network and generate a convolutional neural network group based on the convolutional neural network;
[0083] The second network construction unit is used to construct a recurrent neural network, and introduce an attention mechanism into the recurrent neural network for optimization to generate an improved recurrent neural network;
[0084] The model training unit includes a first training unit, a second training unit, and a third training unit;
[0085] The first training unit includes an iterative calculation unit, which is used to perform multiple iterations on the convolutional neural network group through a genetic algorithm until the optimal model is obtained and the iteration is terminated to obtain a genetic neural network;
[0086] The second training unit includes a pre-training unit, an iterative training unit, and a transfer learning unit;
[0087] A pre-training unit is used to pre-train the improved recurrent neural network through a Transformer model to generate a pre-training model; the improved recurrent neural network generates an improved recurrent neural network by introducing the constructed attention mechanism module and residual module into the recurrent neural network;
[0088] An iterative training unit is used to iteratively train the pre-trained model through a sample set, and then optimize the trained model through an optimizer to generate an optimized model;
[0089] The transfer learning unit is used to freeze the underlying parameters of the optimization model, optimize the upper parameters of the optimization model, and generate the target recurrent neural network;
[0090] A third training unit is used to train the target recurrent neural network through the output of the genetic neural network to generate a multimodal neural network model;
[0091] The fusion unit includes a fusion unit and a data fusion unit;
[0092] The data fusion unit includes:
[0093] A weight generating unit, used for assigning weights to the signal image data and the time series data based on the importance of the data, to obtain image weight data and time series weight data;
[0094] A weight calculation unit, used for multiplying the image weight data and the time series weight data with the corresponding weights to obtain a weighted value;
[0095] The summing and integrating unit is used to add the weighted values to obtain an integrated data set.
[0096] Further, the blockage monitoring unit includes a data interpretation unit and a blockage assessment unit;
[0097] The data interpretation unit is used to calculate the distance between the liquid surface and each sound wave generator according to the time interval between the micro-sound wave generator sending and receiving the sound wave and the specific frequency wavelength of the micro-sound wave generator;
[0098] The blockage assessment unit is used to connect the distances obtained by each sound wave generator, estimate the liquid level in the pipeline, calculate the ratio of the liquid level to the pipeline height and input it into the multimodal neural network model, and predict the blockage degree of the pipeline based on the multimodal neural network model.
[0099] Furthermore, the iterative calculation unit includes:
[0100] A network structure encoding unit is used to encode the structural parameters of the convolutional neural network group into chromosomes of the genetic algorithm to obtain the network chromosome structure;
[0101] A function definition unit, used to construct a fitness function based on detection rate, false detection rate and positioning accuracy;
[0102] An iteration unit, used to randomly generate a set of initial CNN chromosome structures through the network chromosome structure;
[0103] The genetic unit is used to iterate the initial CNN chromosome structure based on the fitness function, select the CNN structure corresponding to the individual with the highest fitness in the last generation of iteration and decode it to obtain a genetic neural network;
[0104] The pre-trained units include:
[0105] Module building unit, used to build attention mechanism modules and residual modules;
[0106] An attention mechanism introduction unit is used to introduce the attention mechanism module into the first three layers of the hidden layer of the recurrent neural network to obtain an attention classification model;
[0107] The residual introduction unit is used to introduce the residual module into the attention classification model to generate an improved recurrent neural network;
[0108] Genetic units include:
[0109] A selection subunit is used to select the initial CNN chromosome structure through a fitness function to obtain a selected parent CNN chromosome structure;
[0110] The crossover subunit is used to perform a crossover operation on the selected parent CNN chromosome structure to generate a new individual;
[0111] The mutation subunit is used to introduce randomness, mutate new individuals, and obtain mutant individuals;
[0112] The evolutionary iterative subunit is used to iterate and evolve the CNN chromosome structure multiple times through the selection subunit, crossover subunit and mutation subunit to select the individual with the highest fitness;
[0113] The decoding subunit is used to decode the individual with the highest fitness to obtain a genetic neural network.
[0114] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel, characterized in that: include: A signal acquisition module is used to collect sound wave signals to determine the blockage of drainage pipes; A signal processing module, connected to the signal acquisition module, for extracting features from the acoustic wave signal to obtain a feature signal; The blockage monitoring module is connected to the signal processing module and is used to input the characteristic signal into a multimodal neural network model to predict the degree of pipeline blockage and obtain blockage monitoring results in real time.
2. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 1 is characterized in that: The signal acquisition module includes a signal transmitting unit, a signal receiving unit, and a time calculation unit; The signal transmitting unit is used to emit sound waves of a specified frequency and a specified direction through a micro sound wave generator; The signal receiving unit is used to receive sound waves reflected from pipeline obstacles; The time calculation unit is used to calculate the time interval between the micro-sound wave generator sending and receiving the sound wave.
3. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 2 is characterized in that: The micro-sound wave generators are arranged on the survey line at preset intervals and have a physical positioning function, which is achieved through position coding; The measuring lines are symmetrically arranged on the top and both sides of the drainage pipe; A controller is provided at one end of the measuring line, and the controller is used to control the sound wave emission, reception and interpretation of the micro-sound wave generator.
4. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 1 is characterized in that: The signal processing module includes a preprocessing unit and a feature extraction unit; The preprocessing unit is used to sequentially remove the mean, remove the linear trend, eliminate the waveform pinch-out and perform band-pass filtering on the acoustic wave signal; The feature extraction unit is used to extract time-frequency domain features of the preprocessed sound wave signal through wavelet transformation.
5. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 4 is characterized in that: The preprocessing unit includes a mean removal unit, a linear trend removal unit, a waveform pinch-out unit, and a bandpass filtering unit; The de-averaging unit is used to remove the average value of the waveform data, eliminate the DC component of the data, and make the average amplitude of the processed data zero; The delinear trend unit is used to eliminate the linear deflection in the data, fit the data into a straight line, and then subtract the linear trend represented by the straight line from the data; The waveform pinch-out unit is used to gradually reduce the beginning and end of the waveform data from the original value to zero; The bandpass filtering unit is used to retain the waveform within a specific frequency range, shield the signals in other frequency bands, obtain the effective signal frequency band and suppress noise.
6. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 1 is characterized in that: The congestion monitoring module includes a model building unit, a model training unit, a fusion unit, and a congestion monitoring unit; The model building unit is used to build a convolutional neural network model and a recurrent neural network model; The model training unit is used to train the convolutional neural network model and the recurrent neural network model respectively through the signal image data and the time series data; The fusion unit is used to fuse the trained convolutional neural network model and the recurrent neural network model to obtain a multimodal neural network model; and is also used to integrate the signal image data and the time series data to obtain an integrated data set; The blockage monitoring unit is used to analyze the integrated data through a multimodal neural network model to obtain the overall blockage result of the pipeline.
7. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 6 is characterized in that: The model construction unit includes a first network construction unit and a second network construction unit; The first network construction unit is used to construct a convolutional neural network and generate a convolutional neural network group based on the convolutional neural network; The second network construction unit is used to construct a recurrent neural network, and introduce an attention mechanism into the recurrent neural network for optimization to generate an improved recurrent neural network.
8. The device for monitoring the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 6 is characterized in that: The model training unit includes a first training unit, a second training unit, and a third training unit; The first training unit includes an iterative calculation unit, which is used to iterate the convolutional neural network group through a genetic algorithm until the optimal model is obtained and the iteration is terminated to obtain a genetic neural network; The second training unit includes a pre-training unit, an iterative training unit, and a transfer learning unit; The pre-training unit is used to pre-train the improved recurrent neural network through the Transformer model to generate a pre-training model; the improved recurrent neural network generates an improved recurrent neural network by introducing the constructed attention mechanism module and residual module into the recurrent neural network; The iterative training unit is used to iteratively train the pre-trained model through a sample set, and then optimize the trained model through an optimizer to generate an optimized model; The transfer learning unit is used to freeze the bottom-level parameters of the optimization model, optimize the upper-level parameters of the optimization model, and generate a target recurrent neural network; The third training unit is used to train the target recurrent neural network through the output of the genetic neural network to generate a multimodal neural network model.
9. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 6 is characterized in that: The fusion unit includes a fusion unit and a data fusion unit; The data fusion unit comprises: A weight generating unit, used for assigning weights to the signal image data and the time series data based on the importance of the data, to obtain image weight data and time series weight data; A weight calculation unit, used for multiplying the image weight data and the time series weight data with the corresponding weights to obtain a weighted value; The summing and integrating unit is used to add the weighted values to obtain the integrated data set.
10. The monitoring device for the clogging condition of the longitudinal drainage system of an in-service highway tunnel according to claim 6, characterized in that: The blockage monitoring unit includes a data interpretation unit and a blockage assessment unit; The data interpretation unit is used to calculate the distance between the liquid surface and each sound wave generator according to the time interval between the micro-sound wave generator sending and receiving the sound wave and the frequency wavelength of the micro-sound wave generator; The blockage assessment unit is used to connect the distances obtained by each sound wave generator, estimate the liquid level in the pipeline, calculate the ratio of the liquid level to the pipeline height and input it into the multimodal neural network model, and predict the blockage degree of the pipeline based on the multimodal neural network model.
Citation Information
Cited By
Real-time open channel clogging diagnosis system and method based on voiceprint and flow velocity fusion
CN121634111A
