CN-121676892-B - Fire-fighting pipeline water leakage identification method and system based on multi-mode data processing

CN121676892BCN 121676892 BCN121676892 BCN 121676892BCN-121676892-B

Abstract

The application discloses a fire-fighting pipeline water leakage identification method and system based on multi-mode data processing, and relates to the field of pipeline water leakage identification. On the basis, a space topology construction technology based on GPS is introduced, and a closed loop feedback mechanism of acoustic and geographic information is constructed by taking an accurate physical path provided by geographic coordinates as a space constraint vector. By applying spatial constraints to the adaptive group velocity matching under the effect of dispersion, the actual group velocities of different frequency bands propagating within the tube are dynamically calculated. And finally, combining a multi-mode weighting fusion strategy, and integrating the corrected distance differences of each frequency band to obtain a precise positioning result, thereby realizing effective compensation and correction of the frequency errors in the heterogeneous medium by utilizing multi-mode data.

Inventors

CHEN QIUHUANG
MIN YINGHAO
Xue Xiaorao
TAO YUQING

Assignees

杭州速利科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260210

Claims (9)

1. The method for identifying the water leakage of the fire-fighting pipeline based on multi-mode data processing is characterized by comprising the following steps of: s1, performing time domain alignment and space label binding on an original sensor data stream acquired by each node of a message pipe network to generate an alignment data set containing synchronous sound vibration waveforms and corresponding geographic coordinate data; s2, carrying out VMD-based signal frequency domain layered decomposition on the synchronous acoustic waveforms in the aligned data set to obtain an intrinsic mode component set; S3, performing GPS constraint-based pipe network space topology construction on geographic coordinate data in the aligned data set to obtain space constraint vectors, wherein the GPS constraint-based pipe network space topology construction comprises the steps of analyzing longitude and latitude information of sensor nodes in the aligned data set, performing spherical coordinate system conversion processing to obtain standardized radian coordinate pairs, performing spherical great circle ranging calculation on the standardized radian coordinate pairs to obtain earth surface Euclidean distances representing mapping straight line distances between the nodes, and performing topology mapping and physical path correction on the earth surface Euclidean distances to obtain the space constraint vectors; s4, performing self-adaptive group velocity matching on the eigenvector component set and the space constraint vector under the dispersion effect to obtain a self-adaptive velocity matrix; S5, carrying out hierarchical cross-correlation calculation and distance conversion on the intrinsic mode component set based on the self-adaptive speed matrix to obtain a dispersion correction distance difference set aiming at different frequency bands; s6, carrying out multi-mode weighting fusion positioning calculation on the dispersion correction distance difference set according to the energy duty ratio of the intrinsic mode component set to obtain a final leakage point positioning result.
2. The method for identifying water leakage of a fire fighting pipeline based on multi-modal data processing according to claim 1, wherein step S1 includes: Performing PPS edge detection and gating analog-to-digital conversion on the original sensor data stream to obtain a digital original sound vibration sequence and a GPS message cache to be analyzed; Analyzing standard positioning sentences in a GPS message cache to be analyzed, and extracting world coordination time and longitude and latitude coordinates from the standard positioning sentences to obtain geographic space-time labels containing space-time attributes; And writing the geographic space-time labels into the associated structure of the digitized original sound vibration sequence as metadata to obtain an aligned data set.
3. The method for identifying water leakage of a fire fighting pipeline based on multi-modal data processing according to claim 1, wherein step S2 includes: performing Hilbert transform and spectrum initialization on the synchronous sound vibration waveforms in the aligned data set to obtain an initialization variation parameter set, and simultaneously separating a metadata cache from the aligned data set; Carrying out iterative optimization on the initialized variation parameter set by an alternate direction multiplier method to obtain an optimized modal spectrum matrix and a final center frequency vector; And carrying out inverse Fourier transform on the optimized modal spectrum matrix to restore each narrow-band time domain signal, and carrying out structural recombination on the narrow-band time domain signal, the final center frequency vector and the metadata cache to obtain an intrinsic modal component set.
4. The method for identifying water leakage in a fire-fighting pipeline based on multi-modal data processing according to claim 1, wherein step S4 includes: extracting the modal characteristic frequency of the intrinsic modal component set to obtain a modal effective frequency vector; carrying out frequency dispersion model parameter matching on the space constraint vector to obtain a frequency dispersion characteristic parameter set; And performing group velocity adaptive calculation on the modal effective frequency vector and the frequency dispersion characteristic parameter set to obtain an adaptive velocity matrix.
5. The method for identifying water leakage in a fire pipe based on multi-modal data processing as set forth in claim 4 wherein the group velocity adaptive calculation of the modal effective frequency vector and the dispersion characteristic parameter set to obtain an adaptive velocity matrix comprises the group velocity adaptive calculation of the modal effective frequency vector and the dispersion characteristic parameter set with the following formula: ; Wherein, the For the theoretical speed of sound, Is the coefficient of the attenuation of the frequency dispersion, Is the effective group frequency of the kth eigenmode component, Is the loop frequency or cut-off frequency of the pipe, Is the inner diameter of the pipeline, For the wall thickness of the pipe, For the water volume modulus, Is the modulus of elasticity of the material, Is of frequency of Group velocity of the sonic wave packet traveling within the tube.
6. The method for identifying water leakage in a fire pipe based on multi-modal data processing as set forth in claim 1, wherein step S5 includes: Performing cross-power spectrum sharpening and time domain inverse transformation on the intrinsic mode component set to obtain a generalized cross-correlation function set; Performing peak search and sub-sampling level precision fitting on the generalized cross-correlation function set to obtain a layered time delay vector; and carrying out frequency dispersion correction distance difference calculation on the layered time delay vector and the self-adaptive speed matrix to obtain a frequency dispersion correction distance difference set.
7. The method for identifying water leakage in a fire pipe based on multi-modal data processing as set forth in claim 1, wherein step S5 includes: Performing cross-power spectrum sharpening and time domain inverse transformation on the intrinsic mode component set to obtain a generalized cross-correlation function set; Subsampling time delay estimation based on Sinc kernel function reconstruction and Newton iteration is carried out on the generalized cross-correlation function set to obtain a layered time delay vector; and carrying out frequency dispersion correction distance difference calculation on the layered time delay vector and the self-adaptive speed matrix to obtain a frequency dispersion correction distance difference set.
8. The method for identifying water leakage in a fire pipe based on multi-modal data processing as set forth in claim 1, wherein step S6 includes: performing energy duty ratio evaluation and normalization processing on the intrinsic mode component set to obtain a weighting coefficient vector; Carrying out weighted decision fusion of multi-band distance differences on the weighted coefficient vector and the dispersion correction distance difference set to obtain fusion distance differences; And combining the space constraint vector and reference node information in the alignment data set, and performing geometric inversion and absolute coordinate mapping on the fusion distance difference to obtain a final leakage point positioning result.
9. A fire pipeline water leakage identification system based on multi-modal data processing, comprising: The sensor data stream processing module is used for carrying out time domain alignment and space label binding on the original sensor data stream collected by each node of the fire control pipe network so as to generate an alignment data set containing synchronous sound vibration waveforms and corresponding geographic coordinate data; the intrinsic mode component generating module is used for carrying out VMD-based signal frequency domain layered decomposition on the synchronous sound vibration waveforms in the aligned data set to obtain an intrinsic mode component set; The space constraint module is used for carrying out GPS constraint-based pipe network space topology construction on geographic coordinate data in the alignment data set to obtain a space constraint vector, and comprises the steps of analyzing longitude and latitude information of a sensor node from the alignment data set, carrying out spherical coordinate system conversion processing to obtain a standardized radian coordinate pair, carrying out spherical great circle ranging calculation on the standardized radian coordinate pair to obtain a surface Euclidean distance representing a mapping linear distance between the nodes, and carrying out topology mapping and physical path correction on the surface Euclidean distance to obtain the space constraint vector; The self-adaptive group velocity matching module is used for carrying out self-adaptive group velocity matching on the eigenvalue component set and the space constraint vector under the dispersion effect so as to obtain a self-adaptive velocity matrix; The frequency dispersion correction distance difference set generation module is used for carrying out layered cross-correlation calculation and distance conversion on the intrinsic mode component set based on the self-adaptive speed matrix so as to obtain frequency dispersion correction distance difference sets aiming at different frequency bands; The positioning result generation module is used for carrying out multi-mode weighting fusion positioning calculation on the dispersion correction distance difference set according to the energy duty ratio of the intrinsic mode component set so as to obtain a final leakage point positioning result.

Description

Fire-fighting pipeline water leakage identification method and system based on multi-mode data processing Technical Field The application relates to the field of pipeline water leakage identification, in particular to a method and a system for identifying water leakage of a fire-fighting pipeline based on multi-mode data processing. Background With the acceleration of the urban process, the fire-fighting pipe network is used as a life line for guaranteeing public safety, the coverage range of the fire-fighting pipe network is increasingly wide, and the topology structure is increasingly complex. Fire control pipeline is usually laid by carbon steel, galvanized pipe and some Polyethylene (PE) pipe mix, and long-term buried or hidden operation makes little leakage be difficult to be perceived in time, not only causes the water waste, more probably causes serious result because of the water pressure is insufficient when the conflagration takes place. In modern pipe network monitoring, along with the development of the technology of the Internet of things, acquired data presents obvious multi-mode characteristics, and not only contains acoustic vibration waveform data reflecting hydrodynamic characteristics, but also contains geographic coordinate data reflecting the spatial position of the pipe network. How to effectively fuse the heterogeneous data so as to meet the accurate positioning requirement under the complex pipe environment becomes a technical problem to be solved in the current industry. However, existing pipeline water leakage identification schemes mostly rely on a single acoustic signal processing technology, and are generally set based on an idealized physical model, namely, assuming that the propagation speed of sound waves in a pipeline is a constant, such as standard underwater sound or general wave speed in a steel pipe. This fixed sound velocity model ignores non-uniformities in the physical medium of the fire pipe network. In practice, when the sound wave propagates in a bounded solid or liquid-solid coupling pipeline, a significant dispersion effect is generated, that is, the propagation speeds of sound waves of different frequency components are not consistent, and the phase speed is separated from the group speed. After long distance propagation, broadband noise generated by water leakage can change the time difference between high-frequency and low-frequency components reaching the sensor due to the speed difference, so that the wave packet is widened and the waveform is distorted. If the calculation is performed using a single speed, systematic accumulated errors that are difficult to correct are generated. Furthermore, the prior art has serious spatio-temporal fracturing problems when multi-modal data applications are involved. Geolocation system (GPS) data is often used only to mark the final leak location on a map and is not truly involved in the physical process of acoustic resolution. In fact, the precise euclidean distance provided by the geographic coordinates is a key constraint for reverse calibrating unknown dispersion characteristics within the pipe, but existing schemes fail to build such an acoustic-geographic closed loop feedback mechanism. Conventional algorithms tend to crudely average the full-band signal, masking the true phase delay information in a particular dominant band. Therefore, there is a need for a water leakage identification method that can deeply fuse spatio-temporal multi-modal data and perform adaptive speed matching with respect to the dispersion effect. Disclosure of Invention The application provides a fire fighting pipeline water leakage identification method based on multi-mode data processing, which aims at the problems existing in the prior art and comprises the steps of S1, carrying out time domain alignment and space label binding on original sensor data streams collected by nodes of a fire fighting pipeline to generate an aligned data set containing synchronous acoustic vibration waveforms and corresponding geographic coordinate data, S2, carrying out signal frequency domain layered decomposition on the synchronous acoustic vibration waveforms in the aligned data set to obtain an intrinsic mode component set, S3, carrying out pipeline network space topology construction on the geographic coordinate data in the aligned data set based on GPS constraint to obtain a space constraint vector, S4, carrying out self-adaptive group velocity matching on the intrinsic mode component set and the space constraint vector under a frequency dispersion effect to obtain a self-adaptive velocity matrix, S5, carrying out layered cross-correlation calculation and distance conversion on the intrinsic mode component set to obtain a frequency dispersion correction distance difference set aiming at different frequency bands, and S6, carrying out multi-mode weighted fusion positioning calculation on the frequency dispersion correction distance difference set according