Pipeline remote monitoring and early warning method and system based on Internet of Things

Abstract

The invention discloses a pipeline remote monitoring and early warning method and system based on the Internet of Things, and relates to the technical field of pipeline leakage monitoring. The method comprises the following steps: acquiring sound signal data of a pipeline in real time, and inputting a de-noised sound signal into a pre-trained pipeline leakage identification model to obtain an identification result; if the identification result is leakage, leakage positioning is carried out according to a subsequent sound signal, and position information of a leakage point is obtained; and generating an early warning report according to the identification result and the position information. By combining the Internet of Things technology with a sound signal acquisition and intelligent identification model, real-time and remote monitoring and accurate positioning of pipeline leakage are realized, especially, the system has higher identification sensitivity to tiny cracks and chronic leakage, the defects of traditional flow and pressure detection means are overcome, the operation and maintenance efficiency is improved, and potential safety hazards are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of pipeline leakage detection, and particularly relates to a pipeline remote monitoring and early warning method and system based on the Internet of Things. Background Art

[0002] With the acceleration of the industrialization and urbanization processes, pipelines have become an indispensable infrastructure in modern society and are widely used in multiple fields such as oil, natural gas, water supply, and heating. The safe operation of pipelines is directly related to energy supply, environmental protection, and the safety of people's lives and property. However, since pipelines are usually distributed over a vast geographical area and some are buried underground, traditional monitoring and maintenance methods often cannot comprehensively and in real time grasp the operating status of pipelines. Faults such as pipeline leaks and ruptures are difficult to detect and repair in a timely manner, which can easily lead to serious environmental pollution, safety accidents, and economic losses.

[0003] Flow and pressure detection are common technical means in traditional pipeline monitoring. Usually, by monitoring the flow rate changes and pressure fluctuations of the fluid in the pipeline, it is determined whether a leak has occurred. For large-scale leaks (such as pipeline ruptures or large-scale leakage incidents), the flow rate and pressure changes are usually more significant, and the monitoring system can relatively easily detect abnormalities. However, for small-scale leaks, especially tiny cracks or chronic leaks, the flow rate and pressure changes are often not obvious, or even show no change at all, resulting in the system being difficult to detect leakage problems in a timely manner and unable to issue early warnings at the initial stage of leakage, affecting the repair efficiency. Summary of the Invention

[0004] The purpose of the present invention is to solve the problem of difficult timely detection of leaks mentioned in the above background art, and to propose a pipeline remote monitoring and early warning method and system based on the Internet of Things.

[0005] In the first aspect of the implementation of the present invention, a pipeline remote monitoring and early warning method based on the Internet of Things is provided. The method includes: Real-time collecting acoustic signal data of a target section of pipeline through multiple sensors, and extracting a first acoustic signal in a first target time period; Preprocessing the first acoustic signal to obtain a second acoustic signal; Inputting the second acoustic signal into a pre-trained pipeline leakage identification model to obtain an identification result; the identification result includes normal and leakage; If the identification result is leakage, obtaining the acoustic signal in a second target time period, and performing leakage location based on the acoustic signal in the second target time period to obtain the position information of the leakage point; the second target time period is a time period of a preset duration after the first target time period; Generating an early warning report according to the identification result and the position information.

[0006] Optionally, the preprocessing of the first acoustic signal to obtain a second acoustic signal includes: Performing variational mode decomposition on the first acoustic signal data to obtain a plurality of mode components; Calculating the envelope entropy of each mode component, and denoting the mode components with envelope entropy less than a preset threshold as effective components; Performing wavelet threshold denoising on each effective component to obtain denoised effective components; Performing signal reconstruction based on a plurality of denoised effective components to obtain a second acoustic signal.

[0007] Optionally, the performing wavelet threshold denoising on each effective component to obtain denoised effective components includes: Using wavelet transform to decompose the target mode component into detail coefficients and approximation coefficients at multiple scales; the target mode component is any one of the effective components; Performing threshold processing on the detail coefficients at multiple scales using a preset threshold function to obtain a plurality of target detail coefficients; the threshold function is: ; where, is the target detail coefficient after threshold processing; W j,k is the detail coefficient at the j-th layer scale, λ j is its corresponding wavelet threshold; sgn() is the sign function; p is a constant; Performing inverse wavelet transform on the approximation coefficient and the plurality of target detail coefficients to obtain denoised effective components.

[0008] Optionally, the pipeline leakage identification model includes an initial convolutional network, a dense convolutional network, and a fully connected network; where: The initial convolutional network uses three consecutive one-dimensional convolutional layers to perform local feature extraction on the input acoustic signal, and then uses a max pooling layer for spatial downsampling to obtain a first feature map; The dense convolutional network uses four consecutive dense blocks to enhance the feature representation of the first feature map, and then uses a max pooling layer for spatial downsampling to obtain a second feature map; The fully connected network uses three fully connected layers to map the second feature map to a leakage probability score. If the leakage probability score is greater than a preset threshold, the output identification result is leakage.

[0009] Optionally, the obtaining the acoustic signal in the second target time period and performing leakage positioning based on the acoustic signal in the second target time period to obtain the position information of the leakage point includes: Mark the sensors corresponding to the acoustic signal data identified as leaks as the first target sensors, and obtain the acoustic signals collected by the multiple first target sensors during the second target period; According to the position information of each first target sensor, mark two adjacent sensors as a sensor group, and determine the time difference threshold according to their spacing; Calculate the propagation time difference of the acoustic signal within each sensor group; Take the pipeline section corresponding to the sensor group with a propagation time difference less than the time difference threshold as the leakage section; mark the sensors of this sensor group as the upstream sensor and the downstream sensor; Calculate the position of the leakage point according to the spacing, propagation time difference and acoustic wave velocity of the upstream sensor and the downstream sensor.

[0010] In the second aspect of the implementation of the present invention, a pipeline remote monitoring and early warning system based on the Internet of Things is provided. The system includes: A data acquisition module, configured to collect acoustic signal data of the target section of the pipeline in real time through multiple sensors, and extract the first acoustic signal during the first target period; A preprocessing module, configured to preprocess the first acoustic signal to obtain a second acoustic signal; A leakage identification module, configured to input the second acoustic signal into a pre-trained pipeline leakage identification model to obtain an identification result; the identification result includes normal and leakage; A leakage location module, configured to, if the identification result is leakage, obtain the acoustic signal during the second target period, and perform leakage location according to the acoustic signal during the second target period to obtain the position information of the leakage point; the second target period is a preset duration period after the first target period; An early warning module, configured to generate an early warning report according to the identification result and the position information.

[0011] Optionally, the preprocessing module includes: A modal decomposition module, configured to perform variational modal decomposition on the first acoustic signal data to obtain multiple modal components; A noise discrimination module, configured to calculate the envelope entropy of each modal component, and mark the modal components with an envelope entropy less than a preset threshold as effective components; A wavelet denoising module, configured to perform wavelet threshold denoising on each effective component to obtain a denoised effective component; A signal reconstruction module, configured to perform signal reconstruction according to multiple denoised effective components to obtain a second acoustic signal.

[0012] Optionally, the wavelet denoising module includes: Wavelet transform module, which is used to decompose the target modal component into detail coefficients and approximation coefficients at multiple scales by using wavelet transform; the target modal component is any valid component; Threshold processing module, which is used to perform threshold processing on the detail coefficients at multiple scales by using a preset threshold function to obtain multiple target detail coefficients; the threshold function is: ; where is the target detail coefficient after threshold processing; W j,k is the detail coefficient at the j-th layer scale, and λ j is its corresponding wavelet threshold; sgn() is the sign function; p is a constant; Inverse wavelet transform module, which is used to perform inverse wavelet transform on the approximation coefficient and multiple target detail coefficients to obtain a denoised valid component.

[0013] Optionally, the pipeline leakage identification model includes an initial convolutional network, a dense convolutional network, and a fully connected network; where: The initial convolutional network uses three consecutive one-dimensional convolutional layers to extract local features of the input acoustic signal, and then uses a max pooling layer for spatial downsampling to obtain a first feature map; The dense convolutional network uses four consecutive dense blocks to enhance the feature representation of the first feature map, and then uses a max pooling layer for spatial downsampling to obtain a second feature map; The fully connected network uses three fully connected layers to map the second feature map to a leakage probability score. If the leakage probability score is greater than a preset threshold, the output recognition result is leakage.

[0014] Optionally, the leakage location module includes: Region determination module, which is used to mark the sensors corresponding to the acoustic signal data with the recognition result of leakage as the first target sensors, and obtain the acoustic signals collected by multiple first target sensors in the second target time period; Longest time difference determination module, which is used to mark two adjacent sensors as a sensor group according to the position information of each first target sensor, and determine the time difference threshold according to their spacing; Time difference calculation module, which is used to calculate the propagation time difference of the acoustic signal within each sensor group; Pipeline segment determination module, which is used to take the pipeline segment corresponding to the sensor group with the propagation time difference less than the time difference threshold as the leakage segment; mark the sensors of this sensor group as the upstream sensor and the downstream sensor; Position determination module, which is used to calculate the position of the leakage point according to the spacing, propagation time difference, and acoustic wave velocity of the upstream sensor and the downstream sensor.

[0015] Advantages of the present invention: The present invention provides a method for remote monitoring and early warning of pipelines based on the Internet of Things. The method includes: collecting acoustic signal data of a target section of the pipeline in real time through multiple sensors, and extracting a first acoustic signal in a first target period; preprocessing the first acoustic signal to obtain a second acoustic signal; inputting the second acoustic signal into a pre-trained pipeline leakage identification model to obtain an identification result; the identification result includes normal and leakage; if the identification result is leakage, obtain the acoustic signal in a second target period, and perform leakage positioning based on the acoustic signal in the second target period to obtain the position information of the leakage point; the second target period is a preset duration period after the first target period; generating an early warning report according to the identification result and the position information.

[0016] By combining Internet of Things technology with acoustic signal acquisition and intelligent identification models, real-time, remote monitoring and precise positioning of pipeline leakage are realized. In particular, higher identification sensitivity is achieved for micro cracks and chronic leaks, making up for the deficiencies of traditional flow and pressure detection methods, improving operation and maintenance efficiency, reducing potential safety hazards, and significantly enhancing the safety and intelligence level of pipeline operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 FIG. Figure 2 is a flowchart of a method for remote monitoring and early warning of pipelines based on the Internet of Things provided by an embodiment of the present invention; Figure 3 FIG. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] An embodiment of the present invention provides a method for remote monitoring and early warning of pipelines based on the Internet of Things. Refer to Figure 1 , Figure 1 FIG. S101, collect acoustic signal data of a target section of the pipeline in real time through multiple sensors, and extract a first acoustic signal in a first target period.

[0020] S102. Preprocess the first acoustic signal to obtain a second acoustic signal.

[0021] S103. Input the second acoustic signal into a pre-trained pipeline leak identification model to obtain an identification result.

[0022] S104. If the identification result is a leak, obtain the acoustic signal in the second target period, and perform leak localization based on the acoustic signal in the second target period to obtain the location information of the leak point.

[0023] S105. Generate a warning report based on the identification result and the location information.

[0024] Wherein, the target section of the pipeline is a straight pipeline; the identification result includes normal and leak; the second target period is a period of preset duration after the first target period.

[0025] Based on the pipeline remote monitoring and warning method based on the Internet of Things provided by the embodiments of the present invention, through the Internet of Things technology combined with acoustic signal acquisition and intelligent identification models, real-time, remote monitoring and precise positioning of pipeline leaks are realized. In particular, it has higher identification sensitivity for micro cracks and chronic leaks, makes up for the deficiencies of traditional flow and pressure detection means, improves operation and maintenance efficiency, reduces potential safety hazards, and significantly enhances the safety and intelligent level of pipeline operation.

[0026] In one implementation, the first target period is a period of time before the current moment, such as the period from 1.2 seconds before the current time to 1 second before the current time. The first acoustic signal is the acoustic signal collected by any sensor during this period. By performing quasi-real-time analysis with a one-second delay, it is possible to avoid the situation where only the nearby sensors detect the leak while the distant sensors have not yet sensed it due to the delay in signal transmission and processing between sensors.

[0027] In one embodiment, step S102 includes: Step 1. Perform variational mode decomposition on the first acoustic signal data to obtain multiple mode components.

[0028] Step 2. Calculate the envelope entropy of each mode component, and record the mode components with envelope entropy less than the preset threshold as effective components.

[0029] Step 3. Perform wavelet threshold denoising on each effective component to obtain the denoised effective components.

[0030] Step 4. Perform signal reconstruction based on multiple denoised effective components to obtain the second acoustic signal.

[0031] In this embodiment, the acoustic signal is decomposed into multiple modal components through variational mode decomposition, effectively extracting the signal features. The effective components are screened out by calculating the envelope entropy, removing the irrelevant noise, and enhancing the information related to leakage. Subsequently, the wavelet threshold denoising technique is used to remove the residual noise, further improving the clarity and stability of the signal. Finally, the signal is reconstructed based on the denoised effective components to obtain the second acoustic signal. This process significantly improves the signal quality, provides a more accurate input for subsequent leakage identification, and enhances the reliability and sensitivity of the detection system.

[0032] In one implementation, the preset threshold can be determined by statistical methods. For example, calculate the mean and standard deviation of the envelope entropy of all modal components, and then the threshold can be set to the mean plus three times the standard deviation. This statistical threshold setting method can reasonably filter out the noise signals, reduce the misidentification rate, while retaining the true leakage signals, and improve the overall robustness of the system.

[0033] In one embodiment, performing wavelet threshold denoising on each effective component to obtain the denoised effective components includes: Using wavelet transform to decompose the target modal component into detail coefficients and approximation coefficients at multiple scales; the target modal component is any one of the effective components; Using a preset threshold function to perform threshold processing on the detail coefficients at multiple scales to obtain multiple target detail coefficients; the threshold function is: ; where is the target detail coefficient after threshold processing; W j,k is the detail coefficient at the j-th scale, λ j is its corresponding wavelet threshold; sgn() is the sign function; p is a constant; Performing inverse wavelet transform on the approximation coefficient and the multiple target detail coefficients to obtain the denoised effective component.

[0034] In one implementation, wavelet transform can decompose the signal into detail coefficients and approximation coefficients at multiple scales, thereby analyzing the signal at different scales, effectively removing the noise in each frequency band, and retaining the key leakage feature information. By performing threshold processing on the detail coefficients, fine adjustment can be made for the noise at different scales, optimizing the signal quality and reducing the influence of interfering noise.

[0035] In one implementation, the db8 wavelet basis function can be used when performing wavelet transform, and the decomposition level is 4 layers. The threshold calculation formula is: ; where λ j is the wavelet threshold at the j-th layer; MED j is the median of the detail coefficients at the j-th layer; N j is the length of the detail coefficients at the j-th layer.

[0036] In one implementation, traditional threshold functions are divided into hard threshold functions and soft threshold functions. The hard threshold function can obtain a high peak signal-to-noise ratio by setting wavelet coefficients below the threshold to zero. However, due to the discontinuity of the threshold function, the processed signal may exhibit oscillations or discontinuities at the signal edges. In contrast, the soft threshold function has a continuous structure, enabling the processed signal to be smoother. However, soft threshold processing also introduces a constant bias, resulting in a certain degree of distortion of the high-frequency information of the signal, especially at the detailed level of the signal. For the threshold function introduced in this embodiment, the parameter p can be set to 0.5. The function is an odd function. Taking the positive half-axis as an example. When W j,k is greater than the threshold, the larger it is, the closer the corrected value is to the original value, and it is continuous at the threshold, avoiding the reconstruction oscillation problem brought by the hard threshold function and also avoiding the reconstruction bias problem brought by the soft threshold function.

[0037] In one implementation, 30 data sets are randomly selected from the experimental data, and different threshold functions (hard threshold function, soft threshold function, improved threshold function) are used for denoising processing. The signal-to-noise ratio (SNR) and root mean square error (RMSE) of each data set after denoising are calculated respectively. Finally, the denoising effects of different methods are evaluated by taking the average value. The comparison results are shown in Table 1.

[0038] Table 1:

[0039] The increase in the signal-to-noise ratio indicates that the method can more effectively increase the proportion of effective components in the signal and reduce the influence of noise. The decrease in the root mean square error indicates that the deviation between the denoised signal and the original signal is smaller, and the denoising effect is better.

[0040] In one embodiment, refer to Figure 2 , Figure 2 which is the network structure diagram of a leakage identification model provided by an embodiment of the present invention. The pipeline leakage identification model includes an initial convolutional network, a dense convolutional network, and a fully connected network; where: The initial convolutional network uses three consecutive one-dimensional convolutional layers to extract local features of the input acoustic signal, and then uses a max pooling layer for spatial downsampling to obtain the first feature map. Specifically, each convolutional layer uses the ReLU activation function; the pooling window size of the max pooling layer is 2, and the stride is 2.

[0041] The dense convolutional network enhances the feature representation of the first feature map by using four consecutive dense blocks, and then performs spatial downsampling using a max pooling layer to obtain the second feature map. Specifically, each dense block includes four cascaded one-dimensional convolutional layers, and the convolutional layer in the subsequent position receives the outputs of all convolutional layers in the previous position as inputs. Each convolutional layer uses the ReLU activation function.

[0042] The fully connected network uses three fully connected layers to map the second feature map to the leakage probability score. If the leakage probability score is greater than the preset threshold, the output recognition result is leakage. Specifically, the second feature map is flattened into a one-dimensional vector and input into the first fully connected layer. The first two fully connected layers use the ReLU activation function. The last fully connected layer uses the Sigmoid function to obtain the probability score.

[0043] In one implementation, the dense convolutional network enhances feature flow and reuse through multi-layer dense connections, improves the model's expression ability and generalization ability for leakage features, and reduces the problem of gradient disappearance. The overall structure of the model efficiently integrates shallow and deep features, significantly improving the recognition accuracy of tiny leaks and the robustness of the model.

[0044] In one implementation, to verify the effectiveness of the pipeline leakage detection method proposed in the present invention, the proposed improved model was compared with the mainstream 1D CNN model and LSTM model in terms of performance. The evaluation metrics included accuracy, precision, recall, and F1 score. The specific comparison results are shown in Table 2.

[0045] Table 2:

[0046] As can be seen from Table 2, the improved model is superior to the comparison models in all evaluation metrics. The higher accuracy and precision indicate that the overall prediction results are closest to the real situation, reducing false alarms and missed detections. The higher recall indicates stronger leakage detection ability for actual leaks, lower missed detection rate, and improved security guarantee ability of the system.

[0047] In one embodiment, step S104 includes: Mark the sensors corresponding to the acoustic signal data with the recognition result of leakage as the first target sensors, and obtain the acoustic signals collected by multiple first target sensors in the second target time period; According to the position information of each first target sensor, mark two adjacent sensors as a sensor group, and determine the time difference threshold according to their spacing; Calculate the propagation time difference of the acoustic signal within each sensor group; Take the pipeline section corresponding to the sensor group with a propagation time difference less than the time difference threshold as the leakage section; denote the sensors in this sensor group as the upstream sensor and the downstream sensor; Calculate the position of the leakage point based on the distance between the upstream sensor and the downstream sensor, the propagation time difference, and the sound wave velocity.

[0048] In one implementation, the time difference threshold TM = D / v. Where D is the distance between the two sensors; v is the sound wave velocity. Ensure that the propagation time difference is within a reasonable range, effectively eliminate abnormal data, and improve the positioning accuracy.

[0049] In one implementation, the cross-correlation method can be used to calculate the propagation time difference within the sensor group. By analyzing the propagation time difference of the sound signal between adjacent sensors, precise positioning of the leakage point is achieved.

[0050] In one implementation, the calculation formula for the position of the leakage point is: ; where x is the distance from the upstream sensor, that is, the position of the leakage point; d is the distance between the upstream sensor and the downstream sensor; v is the sound wave velocity; is the time difference, and when the upstream sensor is closer to the leakage point, the time difference is positive.

[0051] The embodiment of the present invention provides a pipeline remote monitoring and early warning system based on the Internet of Things. Refer to Figure 3 , Figure 3 is the architecture diagram of a pipeline remote monitoring and early warning system provided by the embodiment of the present invention. The system includes: A data acquisition module, which is used to collect the sound signal data of the target section of the pipeline in real time through multiple sensors and extract the first sound signal in the first target time period.

[0052] A preprocessing module, which is used to preprocess the first sound signal to obtain the second sound signal.

[0053] A leakage identification module, which is used to input the second sound signal into a pre-trained pipeline leakage identification model to obtain the identification result.

[0054] A leakage positioning module, which is used to, if the identification result is leakage, obtain the sound signal in the second target time period and perform leakage positioning based on the sound signal in the second target time period to obtain the position information of the leakage point.

[0055] An early warning module, which is used to generate an early warning report based on the identification result and the position information.

[0056] Among them, the target section of the pipeline is a straight pipeline; the identification result includes normal and leakage; the second target time period is a preset duration of time after the first target time period.

[0057] Based on an Internet of Things-based pipeline remote monitoring and early warning system provided by an embodiment of the present invention, through the combination of Internet of Things technology, acoustic signal acquisition, and an intelligent recognition model, real-time, remote monitoring, and precise positioning of pipeline leaks are achieved. In particular, it has higher recognition sensitivity for micro cracks and chronic leaks, makes up for the deficiencies of traditional flow and pressure detection means, improves operation and maintenance efficiency, reduces potential safety hazards, and significantly enhances the safety and intelligence level of pipeline operation.

[0058] In one embodiment, the preprocessing module includes: A variational mode decomposition module for performing variational mode decomposition on the first acoustic signal data to obtain multiple mode components.

[0059] A noise discrimination module for calculating the envelope entropy of each mode component and marking the mode components with envelope entropy less than a preset threshold as effective components.

[0060] A wavelet denoising module for performing wavelet threshold denoising on each effective component to obtain denoised effective components.

[0061] A signal reconstruction module for reconstructing a signal based on multiple denoised effective components to obtain a second acoustic signal.

[0062] In one embodiment, the wavelet denoising module includes: A wavelet transform module for decomposing a target mode component into detail coefficients and approximation coefficients at multiple scales by using wavelet transform; the target mode component is any one of the effective components.

[0063] A threshold processing module for performing threshold processing on the detail coefficients at multiple scales by using a preset threshold function to obtain multiple target detail coefficients; the threshold function is: ; Wherein, is the target detail coefficient after threshold processing; W j,k is the detail coefficient at the j-th layer scale, λ j is its corresponding wavelet threshold; sgn() is the sign function; p is a constant.

[0064] A wavelet inverse transform module for performing wavelet inverse transform on the approximation coefficient and multiple target detail coefficients to obtain denoised effective components.

[0065] In one embodiment, the leak location module includes: A region determination module for marking the sensors corresponding to the acoustic signal data with an identified leak as first target sensors and acquiring the acoustic signals collected by the multiple first target sensors during a second target period; The longest time difference determination module is used to mark two adjacent sensors as a sensor group according to the position information of each first target sensor, and determine a time difference threshold according to their spacing; The time difference calculation module is used to calculate the propagation time difference of the acoustic signal within each sensor group; The pipeline segment determination module is used to take the pipeline segment corresponding to the sensor group with a propagation time difference less than the time difference threshold as the leakage segment; mark the sensors of this sensor group as the upstream sensor and the downstream sensor; The position determination module is used to calculate the position of the leakage point according to the spacing between the upstream sensor and the downstream sensor, the propagation time difference and the acoustic wave velocity.

[0066] It should be noted that in this article, terms such as "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such a process, method, article or device.

[0067] The above has described the embodiments of the present invention in detail, but the content described is only the preferred embodiments of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. A pipeline remote monitoring and early warning method based on the Internet of Things, characterized in that, The method includes: Collecting acoustic signal data of the pipeline in the target section in real time through multiple sensors, and extracting the first acoustic signal in the first target period; Preprocessing the first acoustic signal to obtain a second acoustic signal; Inputting the second acoustic signal into a pre-trained pipeline leakage identification model to obtain an identification result; the identification result includes normal and leakage; If the identification result is leakage, obtain the acoustic signal in the second target period, and perform leakage positioning according to the acoustic signal in the second target period to obtain the position information of the leakage point; the second target period is a preset duration period after the first target period; Generating a warning report according to the identification result and the position information; The preprocessing the first acoustic signal to obtain a second acoustic signal includes: Performing variational mode decomposition on the first acoustic signal data to obtain multiple modal components; Calculating the envelope entropy of each modal component, and recording the modal components with envelope entropy less than a preset threshold as effective components; Performing wavelet threshold denoising on each effective component to obtain a denoised effective component; Performing signal reconstruction according to multiple denoised effective components to obtain a second acoustic signal; The performing wavelet threshold denoising on each effective component to obtain a denoised effective component includes: Using wavelet transform to decompose the target modal component into detail coefficients and approximation coefficients of multiple scales; the target modal component is any one of the effective components; Performing threshold processing on the detail coefficients of multiple scales using a preset threshold function to obtain multiple target detail coefficients; the threshold function is: ； Among them, is the target detail coefficient after threshold processing; W j,k is the detail coefficient at the j-th scale, λ j is its corresponding wavelet threshold; sgn( ) is the sign function; p is a constant; Performing inverse wavelet transform on the approximation coefficient and multiple target detail coefficients to obtain a denoised effective component.

2. The method for remotely monitoring and warning of pipelines based on the Internet of Things according to claim 1, characterized in that The pipeline leakage identification model includes an initial convolutional network, a dense convolutional network, and a fully connected network; where: The initial convolutional network uses three consecutive one-dimensional convolutional layers to extract local features of the input acoustic signal, and then uses a max pooling layer for spatial downsampling to obtain a first feature map; The dense convolutional network uses four consecutive dense blocks to enhance the feature representation of the first feature map, and then uses a max pooling layer for spatial downsampling to obtain a second feature map; The fully connected network uses three fully connected layers to map the second feature map to a leakage probability score. If the leakage probability score is greater than a preset threshold, the output identification result is leakage.

3. A method for remote monitoring and early warning of pipelines based on the Internet of Things according to claim 1, characterized in that, The obtaining the acoustic signal in the second target period, and performing leakage positioning according to the acoustic signal in the second target period to obtain the position information of the leakage point includes: Marking the sensors corresponding to the acoustic signal data with an identification result of leakage as first target sensors, and obtaining the acoustic signals collected by multiple first target sensors in the second target period; According to the position information of each first target sensor, marking two adjacent sensors as a sensor group, and determining a time difference threshold according to their spacing; Calculating the propagation time difference of the acoustic signal in each sensor group; Taking the pipeline section corresponding to the sensor group with a propagation time difference less than the time difference threshold as the leakage section; marking the sensors of this sensor group as the upstream sensor and the downstream sensor; Calculate the location of the leakage point based on the distance between the upstream sensor and the downstream sensor, the propagation time difference, and the acoustic wave velocity.

4. A pipeline remote monitoring and early warning system based on the Internet of Things, characterized in that The system includes: A data acquisition module for collecting acoustic signal data of the target section pipeline in real time through multiple sensors and extracting the first acoustic signal in the first target period; A preprocessing module for preprocessing the first acoustic signal to obtain a second acoustic signal; A leakage identification module for inputting the second acoustic signal into a pre-trained pipeline leakage identification model to obtain an identification result; the identification result includes normal and leakage; A leakage location module for, if the identification result is leakage, acquiring the acoustic signal in the second target period and performing leakage location based on the acoustic signal in the second target period to obtain the location information of the leakage point; the second target period is a period of a preset duration after the first target period; An early warning module for generating an early warning report according to the identification result and the location information; The preprocessing module includes: A modal decomposition module for performing variational modal decomposition on the first acoustic signal data to obtain a plurality of modal components; A noise discrimination module for calculating the envelope entropy of each modal component and recording the modal components with envelope entropy less than a preset threshold as effective components; A wavelet denoising module for performing wavelet threshold denoising on each effective component to obtain a denoised effective component; A signal reconstruction module for performing signal reconstruction according to a plurality of denoised effective components to obtain a second acoustic signal; The wavelet denoising module includes: A wavelet transform module for decomposing the target modal component into detail coefficients and approximation coefficients of multiple scales by using wavelet transform; the target modal component is any one of the effective components; A threshold processing module for performing threshold processing on the detail coefficients of multiple scales by using a preset threshold function to obtain a plurality of target detail coefficients; the threshold function is: ； Among them, is the target detail coefficient after threshold processing; W j,k is the detail coefficient at the j-th layer scale, λ j is its corresponding wavelet threshold; sgn( ) is the sign function; p is a constant; A wavelet inverse transform module for performing wavelet inverse transform on the approximation coefficient and the plurality of target detail coefficients to obtain a denoised effective component.

5. The pipeline remote monitoring and early warning system based on the Internet of Things according to claim 4, wherein, The pipeline leakage identification model includes an initial convolutional network, a dense convolutional network, and a fully connected network; where: The initial convolutional network uses three consecutive one-dimensional convolutional layers to extract local features of the input acoustic signal, and then uses a max pooling layer for spatial downsampling to obtain a first feature map; The dense convolutional network uses four consecutive dense blocks to enhance the feature representation of the first feature map, and then uses a max pooling layer for spatial downsampling to obtain a second feature map; The fully connected network uses three fully connected layers to map the second feature map to a leakage probability score, and if the leakage probability score is greater than a preset threshold, outputs the identification result as leakage.

6. The pipeline remote monitoring and early warning system based on the Internet of Things according to claim 4, characterized in that The leakage location module includes: A region determination module for marking the sensors corresponding to the acoustic signal data with an identification result of leakage as the first target sensors and acquiring the acoustic signals collected by the plurality of first target sensors in the second target period; A longest time difference determination module for, according to the position information of each first target sensor, marking two adjacent sensors as a sensor group and determining a time difference threshold according to their distance; A time difference calculation module, which is used to calculate the propagation time difference of the acoustic signal within each sensor group; A pipeline section determination module, which is used to take the pipeline section corresponding to the sensor group with a propagation time difference less than the time difference threshold as the leakage section; and mark the sensors of this sensor group as the upstream sensor and the downstream sensor; A position determination module, which is used to calculate the position of the leakage point according to the distance between the upstream sensor and the downstream sensor, the propagation time difference and the acoustic wave velocity.

Images