Cable-free seabed observation platform based on underwater acoustic communication
By employing intelligent discrimination and adaptive compression coding through multi-channel sensors and embedded processing units, the problems of rapid deployment and real-time data transmission for cableless seabed observation platforms have been solved, enabling near real-time acquisition and long-term stable observation of seismic data.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHONGQING GEOLOGICAL INSTR FACTORY
- Filing Date
- 2026-01-04
- Publication Date
- 2026-05-12
AI Technical Summary
Existing cableless seabed observation platforms are difficult to deploy quickly in vast, deep sea areas or areas prone to sudden geological events. Data can only be obtained after physical recovery of equipment. They lack effective underwater acoustic communication capabilities, making it difficult to meet the timeliness requirements for earthquake early warning and tsunami rapid reporting. Their sensor configuration is also weak, making it difficult to transmit effective data stably over a long period of time.
The system employs multi-channel sensors to synchronously acquire seabed signals, combined with an embedded processing unit for intelligent discrimination and adaptive compression encoding, and achieves near real-time data transmission through underwater acoustic communication. The platform design features low-power sleep monitoring and modular pressure resistance, and supports network collaboration.
It achieves near real-time transmission of earthquake data, meets the timeliness requirements of earthquake early warning and tsunami rapid reporting, reduces energy consumption, supports long-term continuous operation and rapid deployment, and is suitable for observation in vast and deep sea areas and areas prone to sudden geological events.
Smart Images
Figure CN122017954A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of marine observation and underwater communication technology, specifically to a cableless seabed observation platform based on underwater acoustic communication. Background Technology
[0002] Currently, seabed geophysical observation mainly relies on two types of systems: one is the cabled seabed observation network, such as Japan's DONET, Canada's NEPTUNE, and China's National Seabed Scientific Observation Network; the other is the self-contained uncabled seabed seismograph, among which the floating OBS is widely used.
[0003] Cabled observation networks achieve continuous power supply and high-bandwidth data transmission through submarine optical-electric composite cables, and have advantages such as good real-time performance and long observation cycle. However, their construction and maintenance costs are extremely high, their deployment flexibility is poor, and they are difficult to deploy quickly in vast and deep seas or areas with sudden geological changes.
[0004] In contrast, submersible OBSs are battery-powered and cableless, relying on gravity to sink to the seabed during deployment and rising to the surface for recovery after mission completion via acoustic commands or timed ballast release. These devices are widely used for short-term seismic observation during scientific expeditions; however, they have fundamental limitations, such as not supporting real-time or near-real-time transmission of seismic data: all seismic waveform data is stored in internal solid-state storage and can only be retrieved after physical recovery of the device, making it unsuitable for time-sensitive operational scenarios such as earthquake early warning and tsunami forecasting; and lacking underwater acoustic communication capabilities or possessing only minimal signaling functions: while some newer submersible OBSs integrate underwater acoustic transponders, these are only used to receive ballast release commands or send power and deployment success messages, making it difficult to transmit any valid seismic waveforms or event characteristics.
[0005] The limitations of existing technologies include at least the following issues: They are difficult to deploy rapidly in vast, deep-sea areas or regions prone to sudden geological events; all data can only be acquired after physical retrieval of equipment, making it difficult to meet the timeliness requirements for earthquake early warning and tsunami reporting; they lack effective underwater acoustic communication capabilities, making it difficult to transmit effective seismic waveforms or event characteristics; sensor configurations are weak, with most lacking hydrophones; the platform structure is not optimized for long-term operational use, making it difficult to support long-term continuous operation and lacking networking and collaborative capabilities. Furthermore, efficient compression and robust coding schemes have not been designed specifically for the characteristics of seismic signals, and the hardware modules are independently designed with high power consumption, making it difficult to stably transmit effective data over long periods under cableless, low-bandwidth conditions. Summary of the Invention
[0006] To address the shortcomings of existing technologies, this invention provides a cableless seabed observation platform based on underwater acoustic communication, which solves the problems of insufficient effective communication and difficulty in long-term stable transmission in existing technologies.
[0007] To achieve the above objectives, the present invention provides the following technical solution: a cableless seabed observation platform based on underwater acoustic communication, comprising: a battery compartment unit, a seismic unit, an acoustic communication unit, a communication sphere unit, and an embedded processing unit; the battery compartment unit provides stable power to each unit; the main platform shell provides deep-sea pressure-resistant protection for the internal units; the seismic unit is used to synchronously acquire multi-channel raw seabed signals; the embedded processing unit is the core control module of the platform, specifically including a low-power sleep monitoring module, a signal processing module, and an intelligent event discrimination module: the signal processing module receives multi-channel raw seabed signals and performs feature extraction processing to generate data to be discriminated; the intelligent event discrimination module analyzes the data to be discriminated using a lightweight neural network model to generate target data; the acoustic communication unit performs adaptive compression and channel coding processing on the target data to generate a transmission bitstream of compressed data. The communication sphere unit has a built-in underwater acoustic transducer that modulates the transmitted bitstream and compressed data to generate a modulated signal. And transmit it to the target receiver.
[0008] Furthermore, the seismic unit includes a three-channel velocity sensor, a three-channel acceleration sensor, and a one-channel hydrophone to simultaneously acquire several channels of raw signals, the raw signals of which include velocity signals, acceleration signals, and hydrophone signals.
[0009] Furthermore, the specific steps for generating the data to be judged are as follows: preprocessing the multi-channel raw seabed signals; performing wavelet transform processing on the preprocessed multi-channel raw seabed signals to extract the time-frequency features of each channel's raw seabed signals, and extracting the energy features of each channel's raw seabed signals within a preset dominant frequency band; and constructing a fused feature vector based on the time-frequency features and energy features of each channel's raw seabed signals. They are then marked as data to be judged.
[0010] Furthermore, the specific steps for generating the target data are as follows: The data to be judged is input into a lightweight neural network model to generate the earthquake probability, i.e. Determine whether the earthquake probability is higher than a preset earthquake probability threshold. If the probability exceeds the preset earthquake probability threshold, the data to be judged is time-stamped to locate the preprocessed multi-channel seabed raw signal corresponding to the data to be judged, and it is marked as the target data.
[0011] Furthermore, the generated transmission bitstream is used to compress the data. The specific steps are as follows: Adaptive decomposition of the target data is performed using wavelet packet transform to generate compressed data; error correction coding is performed on the compressed data based on short LDPC codes to generate a transmission bitstream for the compressed data. .
[0012] Further, a modulated signal is generated. The specific steps for transmitting the data to the target receiver are as follows: Based on Orthogonal Frequency Division Multiplexing (OFDM), the transmitted bit stream is compressed into data. Modulation processing is performed to generate a modulated signal. The underwater acoustic transducer modulates the signal. The conversion process is performed and then transmitted to the target receiver.
[0013] Furthermore, it also includes: the modulated signal after conversion processing at the target receiver. Perform reverse recovery processing; and convert the modulated signal after reverse recovery processing... The data is compiled into earthquake characteristic data that can be used for earthquake phase identification, magnitude estimation, or tsunami warning.
[0014] Furthermore, the low-power sleep monitoring module of the embedded processing unit adopts a dynamic power management mechanism: during non-event periods, the main processor enters a deep sleep state, and only the low-power coprocessor listens for trigger signals.
[0015] The beneficial effects of this invention are as follows:
[0016] By integrating multi-channel sensors into the seismic unit, more comprehensive data is collected, providing a rich foundation for subsequent analysis. Through local intelligent discrimination of the embedded processing unit and efficient compression encoding of the acoustic unit, the low bandwidth characteristics of the underwater acoustic channel are adapted to achieve near real-time transmission of seismic data. Key information can be obtained without retrieving the equipment, meeting the timeliness requirements of earthquake early warning and tsunami early warning. The low-power sleep monitoring mechanism significantly reduces energy consumption. Combined with the modular pressure-resistant platform design, it supports rapid deployment and long-term continuous operation in the deep sea. The platform has networking and collaborative potential, which can flexibly adapt to the observation needs of the vast deep sea and areas with sudden geological events. It effectively solves the pain points of existing technologies such as limited deployment, data lag, high energy consumption, and difficulty in long-term operation, providing an efficient and reliable solution for marine geological observation.
[0017] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description
[0018] Figure 1 This is a core workflow diagram of a cableless seabed observation platform based on underwater acoustic communication according to the present invention. Detailed Implementation
[0019] Please see Figure 1This invention provides a technical solution: a cableless seabed observation platform based on underwater acoustic communication, comprising: a battery compartment unit (equipped with a large-capacity battery pack and a low-power design to ensure continuous operation for more than 6 months), a seismic unit, an acoustic communication unit, a communication sphere unit, and an embedded processing unit; the battery compartment unit provides stable power to each unit; the seismic unit is used to synchronously acquire multi-channel raw seabed signals; the embedded processing unit is the core control module of the platform, specifically including a low-power sleep monitoring module, a signal processing module, and an intelligent event discrimination module: the low-power sleep monitoring module maintains low-power operation of the system while ensuring continuous acquisition of raw signals by the seismic unit; the signal processing module receives multi-channel raw seabed signals and performs feature extraction processing to generate data to be discriminated; the intelligent event discrimination module analyzes the data to be discriminated using a lightweight neural network model to generate target data; the acoustic communication unit performs adaptive compression and channel coding processing on the target data to generate the final transmission bitstream of compressed data. The communication sphere unit has a built-in underwater acoustic transducer that modulates the compressed data of the final transmitted bitstream to generate a modulated signal. And transmit it to the target receiver.
[0020] It also includes: the modulated signal after conversion processing at the target receiver. Perform reverse recovery processing, namely demodulation, decoding, and reconstruction; and convert the modulated signal after reverse recovery processing... The data is compiled into earthquake characteristic data that can be used for earthquake phase identification, magnitude estimation, or tsunami warning.
[0021] The low-power sleep monitoring module of the embedded processing unit adopts a dynamic power management mechanism: during non-event periods, the main processor enters a deep sleep state, and only the low-power coprocessor listens for the trigger signal, which reduces the overall power consumption of the platform by more than 80%.
[0022] Specifically, the seismic unit includes a three-channel velocity sensor, a three-channel acceleration sensor, and a one-channel hydrophone. Several (e.g., 7) channels of raw signals are synchronously acquired at a fixed sampling rate of 100Hz. These 7 channels of raw signals include 3 velocity signals, 3 acceleration signals, and 1 hydrophone signal. The signals are obtained through the triaxial velocity sensor within the seismic unit. Triaxial accelerometer and hydrophone With a fixed sampling rate Simultaneously acquire raw signals from 7 channels.
[0023] The specific steps for generating the data to be judged are as follows: Preprocess the multi-channel raw seabed signals (e.g., remove the mean, filter, etc.); perform wavelet transform (WT) processing on the preprocessed multi-channel raw seabed signals to extract the time-frequency features of each channel's raw seabed signal, i.e. ;in Indicates the first Channel signal, For wavelet basis functions, The scale parameter is used to extract the energy characteristics of the original seabed signals from each channel within the preset dominant frequency band (1-6Hz P-wave), i.e. ;
[0024] Based on the time-frequency and energy characteristics of the original seabed signals from each channel, a fused feature vector is constructed. And marked as data to be judged, that is: ;in This represents the time difference between the first arrival of the P-wave in the hydrophone and the vertical velocity channel. The average coherence coefficient among the six solid channels. and These are the signal-to-noise ratios of the hydrophone and the vertical velocity channel, respectively.
[0025] The specific steps for generating target data are as follows: Input the data to be judged into a lightweight neural network model to generate earthquake probabilities, i.e.: Determine whether the earthquake probability is higher than a preset earthquake probability threshold. If the probability exceeds the preset earthquake probability threshold, i.e. If the value is 0.85, it is considered a suspected earthquake event, and the data to be identified is time-stamped to locate the preprocessed multi-channel raw seabed signal for the suspected earthquake event period corresponding to the data to be identified, and it is marked as the target data; if it is not higher than the preset earthquake probability threshold, it is considered a suspected earthquake event. If so, it will maintain a low-power listening state and will not generate target data.
[0026] In this implementation scheme, the seismic unit synchronously acquires 7 channels of signals at a fixed sampling rate of 100Hz, compensating for the deficiency of single acquisition dimension and providing comprehensive data support for subsequent analysis. The preprocessing stage effectively filters out noise, and wavelet transform accurately extracts time-frequency features. Combined with energy extraction of the 1-6Hz P-wave dominant frequency band, it focuses on key seismic signals and improves feature effectiveness. The fusion feature vector integrates time-frequency, energy features, and key parameters such as P-wave first arrival time difference, channel coherence coefficient, and signal-to-noise ratio, strengthening the event discrimination criteria from multiple dimensions and solving the problem of insufficient discrimination accuracy in traditional methods. A lightweight neural network generates earthquake probabilities and sets a threshold of 0.85 to achieve accurate screening of suspected seismic events, avoiding invalid data from consuming resources. When the threshold is not reached, a low-power state is maintained. The overall process takes into account data comprehensiveness, feature effectiveness, and discrimination accuracy, laying a high-quality foundation for subsequent data compression and transmission while reducing energy consumption.
[0027] Specifically, generating the final transmission bitstream for compressed data The specific steps are as follows: Adaptive decomposition of the target data is performed using wavelet packet transform (WPT), that is: ;in Different sub-bands are represented, and the optimal decomposition tree is selected to maximize the compression ratio. Furthermore, the wavelet coefficients in the wavelet packet transform are adaptively quantized based on a quantizer, as shown in the following formula: ;in To quantize the step size, the signal-to-noise ratio is dynamically adjusted to generate compressed data;
[0028] Error correction encoding is performed on the compressed data using short-code LDPC (Low-Density Parity-Check) codes to generate the final transmission bitstream. ,Right now: .
[0029] Generate modulated signal The specific steps for transmitting the data to the target receiver are as follows: Based on Orthogonal Frequency Division Multiplexing (OFDM), the final transmitted bit stream is compressed. Modulation processing is performed to generate a modulated signal. ,Right now: ;in For modulation basis functions, For signed numbers;
[0030] underwater acoustic transducers modulate signals Perform conversion processing (convert the modulated signal) It is converted into sound waves and propagated through the seawater medium to the target receiving end (such as a buoy or shore-based node).
[0031] In this implementation scheme, wavelet packet transform adaptive decomposition and selection of the optimal decomposition tree, combined with adaptive quantization based on dynamic adjustment of the quantization step size according to the signal-to-noise ratio, maximizes the compression ratio while accurately preserving key seismic phase information, avoiding the distortion effect of traditional compression on non-stationary seismic signals; short-code LDPC error correction coding significantly improves the data's anti-interference capability, adapts to the high bit error rate problem caused by multipath and Doppler effects in underwater acoustic channels, and ensures the reliability of data transmission after encoding; OFDM modulation technology adapts to the characteristics of underwater acoustic channels, realizing efficient signal carrying and transmission, and completes electro-acoustic conversion and seawater medium propagation with underwater acoustic transducers, without relying on submarine optical cables, realizing cableless near real-time transmission of target data, providing key technical support for real-time earthquake data backhaul and tsunami early warning.
[0032] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0033] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A cableless seabed observation platform based on underwater acoustic communication, characterized in that, include: Battery compartment unit, seismic unit, acoustic unit, communication ball unit, and embedded processing unit; The battery compartment unit provides a stable power supply to each unit; The seismic unit is used to synchronously acquire multi-channel raw seabed signals; The embedded processing unit includes a low-power sleep monitoring module, a signal processing module, and an intelligent event discrimination module. The signal processing module receives raw signals from multiple channels on the seabed, performs feature extraction processing, and generates data to be judged. The intelligent event discrimination module analyzes the data to be discriminated using a lightweight neural network model to generate target data; The acoustic communication unit performs adaptive compression and channel coding processing on the target data, generating a transmission bitstream of compressed data. ; The communication ball unit has a built-in underwater acoustic transducer that modulates the transmitted bitstream and compressed data to generate a modulated signal. And transmit it to the target receiver.
2. The cableless seabed observation platform based on underwater acoustic communication according to claim 1, characterized in that, The seismic unit includes a three-channel velocity sensor, a three-channel acceleration sensor, and a one-channel hydrophone to simultaneously acquire several channels of raw signals. The raw signals of the channels include velocity signals, acceleration signals, and hydrophone signals.
3. The cableless seabed observation platform based on underwater acoustic communication according to claim 2, characterized in that, The specific steps for generating the data to be judged are as follows: Preprocessing of raw multi-channel seabed signals; Wavelet transform processing is performed on the preprocessed multi-channel raw seabed signals to extract the time-frequency characteristics of each channel's raw seabed signals, and the energy characteristics of each channel's raw seabed signals within the preset dominant frequency band are also extracted. Based on the time-frequency and energy characteristics of the original seabed signals from each channel, a fused feature vector is constructed. They are then marked as data to be judged.
4. The cableless seabed observation platform based on underwater acoustic communication according to claim 1, characterized in that, The specific steps for generating the target data are as follows: The data to be judged is input into a lightweight neural network model to generate earthquake probabilities, i.e. ; Determine if the earthquake probability is higher than a preset earthquake probability threshold. ; If the earthquake probability exceeds the preset threshold, the data to be judged is time-stamped to locate the preprocessed multi-channel seabed raw signal corresponding to the data to be judged, and it is marked as the target data.
5. The cableless seabed observation platform based on underwater acoustic communication according to claim 4, characterized in that, Generate a transmission bitstream for compressed data The specific steps are as follows: Wavelet packet transform is used to adaptively decompose the target data and generate compressed data. Error correction encoding of compressed data is performed based on short code LDPC codes to generate a transmission bitstream of compressed data. .
6. The cableless seabed observation platform based on underwater acoustic communication according to claim 5, characterized in that, Generate modulated signal The specific steps for transmitting the signal to the target receiver are as follows: Based on Orthogonal Frequency Division Multiplexing (OFDM), the transmitted bit stream is compressed into data. Modulation processing is performed to generate a modulated signal. ; The underwater acoustic transducer modulates the signal. The conversion process is performed and then transmitted to the target receiver.
7. The cableless seabed observation platform based on underwater acoustic communication according to claim 1, characterized in that, Also includes: The target receiver processes the modulated signal after conversion. Perform reverse recovery processing; The modulated signal after reverse recovery processing The data is compiled into earthquake characteristic data that can be used for earthquake phase identification, magnitude estimation, or tsunami warning.
8. The cableless seabed observation platform based on underwater acoustic communication according to claim 1, characterized in that, The low-power sleep monitoring module of the embedded processing unit adopts a dynamic power management mechanism: during non-event periods, the main processor enters a deep sleep state, and only the low-power coprocessor listens for trigger signals.