An acoustic wave monitoring system based on fiber optic sensing

Abstract

The application discloses a kind of acoustic wave monitoring systems based on optical fiber sensing, it is related to optical fiber sensing acoustic wave monitoring technical field, contain multiple modules, optical fiber sensing module uses the novel microstructure optical fiber with unique porous structure, surface is coated with intelligent composite material, also fill nanofluid, enhance response to acoustic wave, signal conversion module combines the advantage of multiple photoelectric conversion, according to light intensity dynamic switching mode, signal processing module uses hybrid filtering and adaptive quantization technology, data analysis module is based on the network of deep learning, fusion multi-scale feature, early warning module is based on risk assessment dynamic early warning, comprehensive multiple factor adjustment threshold and level, each module is coordinated, realize the efficient monitoring, analysis and early warning of acoustic wave signal.The application has excellent performance, high sensitivity, wide dynamic range;It can be intelligent and accurate analysis early warning, judge acoustic wave characteristics and timely alarm;Data storage is safe and efficient, provides advanced and reliable scheme for acoustic wave monitoring.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensing acoustic wave monitoring technology, and in particular to an acoustic wave monitoring system based on fiber optic sensing. Background Technology

[0002] In the development of various industries today, acoustic monitoring is becoming increasingly important as a crucial means of acquiring environmental and equipment information. However, traditional acoustic monitoring methods have many limitations and are unable to meet the complex and diverse application needs of today.

[0003] Traditional acoustic monitoring equipment mostly relies on conventional technologies such as piezoelectric sensors. These devices are not very sensitive and have limited ability to capture weak acoustic signals. For example, in the early stages of earthquake monitoring, weak seismic wave signals are difficult to detect accurately, leading to delayed warnings and potentially missing the optimal response time. Furthermore, their dynamic range is narrow, making them prone to saturation distortion when faced with strong acoustic signals, resulting in incomplete recording of acoustic information. Moreover, traditional equipment has weak anti-interference capabilities; in complex environments, such as industrial sites with significant electromagnetic interference, the accuracy of monitoring data is severely affected, making the monitoring results difficult to accurately reflect the actual situation.

[0004] Furthermore, with the advent of the Internet of Things and the era of big data, various industries are placing increasingly higher demands on the integration and intelligence of acoustic monitoring systems. However, existing acoustic monitoring systems often operate in isolation, making effective integration with other systems difficult. In industrial production, the lack of data interaction between equipment management systems and acoustic monitoring systems prevents timely adjustments to production plans and maintenance strategies based on acoustic signals generated during equipment operation, leading to low production efficiency and increased equipment maintenance costs. Moreover, traditional systems lack intelligent analysis capabilities, simply recording acoustic data without extracting valuable information from massive amounts of data. They struggle to accurately determine the source, nature, and development trend of acoustic signals, failing to provide strong support for decision-making.

[0005] In specialized fields such as marine acoustic monitoring and underground resource exploration, higher performance requirements are placed on acoustic monitoring systems. Traditional systems, due to technological limitations, cannot adapt to complex environmental conditions and diverse monitoring needs. Complex hydrological conditions in the ocean and the inhomogeneity of underground media pose significant challenges to acoustic monitoring. Therefore, there is an urgent need for an innovative acoustic monitoring system that can overcome the limitations of traditional technologies, achieving high sensitivity, wide dynamic range, strong anti-interference capabilities, and intelligent monitoring and analysis to meet the growing application demands of various fields. Summary of the Invention

[0006] The present invention proposes an acoustic wave monitoring system based on fiber optic sensing to solve the problems mentioned in the prior art.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] An acoustic wave monitoring system based on fiber optic sensing includes the following modules:

[0009] Fiber optic sensing module: Employs microstructured optical fibers with an internal porous structure, coated with a smart composite material. The material properties are adjusted according to the formula: E adjust =E0+k1·A+k2·f, where E adjust E0 is the initial elastic modulus, A is the acoustic amplitude, f is the acoustic frequency, and k1 and k2 are material property adjustment coefficients.

[0010] Signal conversion module: Utilizing multimodal photoelectric conversion technology, combining PIN photodiodes and avalanche photodiodes (APDs), it dynamically switches conversion modes based on the intensity of the optical signal. The formula is as follows: Among them G switch For the gain after switching, G APD For APD gain, G PIN P is the gain of the PIN photodiode, P is the input optical power, P threshold Power threshold;

[0011] Signal processing module: Employs a hybrid filtering algorithm, combining traditional bandpass filters with wavelet transform filtering. The decomposition and reconstruction process of the wavelet transform is based on an improved Mallat algorithm, with the iterative formula: c j+1 [n]=∑ k h[k-2n]c j [k], d j+1 [n]=∑ k g[k-2n]c j [k], where c j [n] is the approximation coefficient, d j [n] represents the detail coefficients, and h[k] and g[k] represent the wavelet filter coefficients;

[0012] Data Analysis Module: This module introduces a deep learning-based multi-scale feature fusion network to analyze acoustic signals. This network combines the advantages of convolutional neural networks and long short-term memory networks, fusing features from different scales through a multi-scale feature fusion layer. The network loss function employs an improved cross-entropy loss function. Where y ij For real labels, p ij To predict probabilities, N is the sample size, M is the number of classes, λ is the regularization coefficient, and W... l This is the weight matrix;

[0013] Early warning module: Employs a dynamic early warning strategy based on risk assessment, quantifying risk through fuzzy comprehensive evaluation, using the following formula: Where R is the risk assessment value, w i For the weight, μ i As a membership degree, when the risk assessment value exceeds a certain threshold, an early warning is triggered and relevant personnel are notified.

[0014] Furthermore, the system also includes a calibration module that employs an adaptive calibration method based on a virtual reference signal. During system operation, a virtual reference signal is generated. By comparing the difference between the monitored signal and the virtual reference signal, an adaptive algorithm is used to automatically adjust the system parameters. The adaptive adjustment formula is as follows: Where θ n Let J(θ) be the parameter vector for the nth iteration, μ be the learning rate, and J(θ) be the parameter vector for the nth iteration. n The loss function is denoted by , and a calibration history database is established to record and analyze the parameters and results of each calibration.

[0015] Furthermore, it also includes a data storage module. Preliminary data processing is performed on edge devices at the monitoring site to extract key features. The processed data is then encrypted and stored on distributed storage nodes. Blockchain technology is used, and smart contracts enable data access control and sharing. The redundancy of data storage is dynamically adjusted based on the importance and real-time nature of the data, using the formula: R redundancy = α·I + β·T, where P redundancy Here, I represents redundancy, T represents data importance, and α and β represent adjustment coefficients.

[0016] Furthermore, in the fiber optic sensing module, a nanofluid with acoustic amplification function is filled into the porous structure of the optical fiber. When sound waves act on the optical fiber, the nanoparticles in the nanofluid vibrate and scatter, enhancing the photoacoustic interaction inside the optical fiber. The acoustic amplification factor is related to the nanoparticle concentration and the sound wave frequency according to the formula: A amplify =A0·(1+k·C·f), where A amplify A0 represents the amplified sound wave amplitude, C represents the nanoparticle concentration, f represents the sound wave frequency, and k represents the amplification factor.

[0017] Furthermore, in the signal conversion module, a photoelectric conversion enhancement structure based on a photonic crystal is adopted. This structure consists of a photonic crystal and a photoelectric conversion material. By optimizing the structural parameters of the photonic crystal, the conversion of optical signals of different frequencies is achieved. The relationship between the light absorption efficiency and the structural parameters of the photonic crystal is fitted through numerical simulation and experimental data to obtain an empirical formula: η absorb = a·L+b·r+c, where η absorbdenoted as light absorption efficiency, L as lattice constant, r as filling ratio, and a, b, and c as fitting coefficients.

[0018] Furthermore, in the signal processing module, an adaptive noise cancellation technique based on genetic algorithm optimization is adopted. This technique searches for the optimal filter coefficients using a genetic algorithm. The fitness function of the genetic algorithm is defined as the reciprocal of the sum of squared errors between the monitored signal and the noise estimation signal. The population is evolved through selection, crossover, and mutation operations. The fitness function formula is: Where F is the fitness value, s i To monitor signals, The noise estimation signal is given by N, where N is the signal length.

[0019] Furthermore, in the data analysis module, a knowledge fusion method based on transfer learning is adopted. When facing new monitoring scenarios, existing relevant model knowledge is used for transfer learning, and the model parameters are fine-tuned to adapt to the new task. At the same time, knowledge from different data sources is fused. The weight allocation of knowledge fusion is dynamically adjusted according to the reliability and relevance of the data sources, as shown in the formula: Where w source For data source weights, R source C is a reliability metric for the data source. source This is a correlation metric between the data source and the current task, where m is the number of data sources.

[0020] Furthermore, the early warning module employs a collaborative early warning mechanism based on multi-sensor fusion. This mechanism processes and analyzes multi-sensor data using a data fusion algorithm. The data fusion algorithm utilizes a Kalman filter, and its state prediction equation is as follows: The update equation is: in Let F be the state estimate, B be the state transition matrix, u be the control input matrix, z be the measurement value, H be the measurement matrix, and K be the Kalman gain.

[0021] Furthermore, the calibration module employs a real-time calibration method based on online learning. During system operation, new calibration data is collected, and the calibration model is updated in real time using an online learning algorithm. The online learning algorithm uses stochastic gradient descent, and its parameter update formula is as follows: Where θ n Let μ be the parameter vector for the nth iteration, μ be the learning rate, L be the loss function, and x be the parameter vector for the nth iteration. n For the nth calibration data sample, y n This corresponds to the actual value.

[0022] Furthermore, in the data storage module, intelligent caching is set up on edge devices. Data is stored in a hierarchical manner based on its access frequency and importance. The cache replacement strategy adopts the Least Recently Used (LRU) algorithm to improve data access efficiency. The relationship between cache hit rate, cache size, and data access pattern is derived from an empirical formula through simulation experiments: H cache =α·S cache +b·M access +c, where H cache For cache hit rate, S cache M is the cache size. access Here, is a data access pattern indicator, and a, b, and c are fitting coefficients.

[0023] Compared with existing technologies, the beneficial effects of this invention are:

[0024] In terms of monitoring performance, the system exhibits exceptional sensitivity and a wide dynamic range. The combination of novel microstructured optical fibers and intelligent composite materials, along with the acoustic amplification capabilities of nanofluids, significantly enhances the ability to capture weak acoustic signals, enabling precise detection of even extremely subtle changes in sound waves. Simultaneously, it operates stably even with strong acoustic signals, without saturation distortion, and can completely record acoustic information. Furthermore, its anti-interference capability is greatly improved; the special optical fiber structure and signal processing technology effectively resist external electromagnetic interference and other influences, ensuring accurate and reliable monitoring data acquisition even in complex environments.

[0025] In terms of data analysis and early warning, the system achieves intelligence and precision. A multi-scale feature fusion network based on deep learning can deeply analyze acoustic signals, accurately identify signal characteristics, and quickly determine the source, nature, and development trend of acoustic waves. Based on a dynamic early warning strategy of risk assessment and a collaborative early warning mechanism of multi-sensor fusion, considering multiple factors, it can issue timely and accurate early warning information, allowing relevant personnel to prepare in advance and effectively reduce losses caused by disasters and accidents.

[0026] In terms of system integration and data management, the architecture combining distributed encrypted storage and edge computing achieves efficient data storage and secure management. The application of blockchain technology ensures data immutability and security, while intelligent caching strategies improve data access efficiency. Simultaneously, the system possesses excellent integrability, enabling data interaction and collaborative work with other systems to meet the diverse needs of different industries. The innovative calibration module method guarantees the long-term accuracy and stability of the system, providing strong assurance for the reliability of monitoring results. In conclusion, this system provides an advanced and reliable solution for the acoustic wave monitoring needs of various industries, promoting the development and application of acoustic wave monitoring technology. Attached Figure Description

[0027] Figure 1 This is a schematic block diagram of the acoustic wave monitoring system based on fiber optic sensing proposed in this invention.

[0028] Figure 2 A bar chart comparing the sensitivity of traditional monitoring systems and this system;

[0029] Figure 3 This is a line graph showing the change in the early warning accuracy of this system over five consecutive time periods.

[0030] Figure 4 This is a pie chart showing the proportion of different types of data (raw monitoring data, processed data, calibration data, early warning data, and other data) stored in this system within a certain time period. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0033] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0034] Reference Figures 1 to 4 A fiber optic sensing-based acoustic wave monitoring system includes the following modules:

[0035] Fiber Optic Sensing Module: Utilizing an innovative microstructured optical fiber, this module incorporates a unique and precise porous structure. These porous structures are not randomly distributed; their pore sizes range from nanometers to micrometers, fabricated through precise photolithography and etching processes to form a highly ordered array distribution. Based on advanced fluid-solid coupling theory, the size, shape, and spacing of the porous structure are repeatedly optimized through simulation. By combining numerical calculations and experimental verification, structural parameters are precisely adjusted to ensure maximum enhancement of the interaction between the optical fiber and sound waves under different environments, significantly improving sensitivity to various frequency sound waves. Simultaneously, the intelligent composite material coated on the optical fiber surface is prepared from various functional nanoparticles and polymers through a special blending process. These nanoparticles include piezoelectric nanoparticles and magnetostrictive nanoparticles, uniformly dispersed within a polymer matrix. This composite material possesses adaptive acoustic properties; when external sound waves are applied, the internal nanoparticles interact with the sound waves. Its elastic modulus and damping coefficient can be dynamically adjusted according to the intensity and frequency of the external sound waves, following a precisely derived formula E. adjust =E0+k1·A+k2·f, further enhancing the optical fiber's response to sound waves, achieving efficient capture and accurate sensing of sound wave signals, where E adjust E0 is the adjusted elastic modulus, A is the acoustic amplitude, f is the acoustic frequency, and k1 and k2 are material property adjustment coefficients.

[0036] Signal Conversion Module: Utilizing cutting-edge multimodal photoelectric conversion technology, this module deeply integrates the unique advantages of PIN photodiodes and avalanche photodiodes (APDs). This technology monitors the light signal intensity in real time through a high-precision signal detection circuit. In terms of hardware design, it carefully selects PIN photodiodes with ultra-low noise characteristics, fabricated using high-purity semiconductor materials. Through advanced epitaxial growth processes, it precisely controls the impurity concentration and crystal structure of the material, enabling high-precision linear conversion even under strong light signal input due to its excellent linearity. For the avalanche photodiode (APD), a special semiconductor structure design is employed, constructing multiple heterojunctions internally. By precisely controlling the thickness and doping concentration of each layer, it can achieve high-gain conversion under weak light signal conditions by utilizing the avalanche multiplication effect. The system monitors the input light power P in real time and compares it with a pre-set power threshold P. threshold The comparison is performed. Gain adjustment during the conversion process follows the formula: Among them G switch For the gain after switching, G APDG is the gain of APD. PIN This represents the gain of the PIN photodiode. When P <P threshold When the optical signal is weak, it automatically switches to APD working mode, utilizing its high gain characteristic G. APD Amplify and convert the optical signal; when P ≥ P threshold When the light signal is strong, it switches to the PIN photodiode operating mode, utilizing its linear conversion characteristic G PIN This ensures high accuracy in the conversion and effectively expands the dynamic range of the signal conversion.

[0037] Signal processing module: Employs an innovative hybrid filtering algorithm, organically integrating the advantages of traditional bandpass filters and wavelet transform filtering. The bandpass filter section utilizes high-quality electronic components and precisely sets the filter's center frequency and bandwidth based on the expected frequency range of the signal, effectively filtering out interference signals outside the frequency band and achieving preliminary frequency selection. The wavelet transform decomposition and reconstruction process is based on an improved Mallat algorithm, with the following iterative formula:

[0038] c j+1 [n]=∑ k h[k-2n]c j [k],

[0039] d j+1 [n]=∑ k g[k-2n]c j [k], where c j [n] is the approximation coefficient, d j [n] represents the detail coefficients, and h[k] and g[k] represent the wavelet filter coefficients. After multi-stage amplification, adaptive quantization is used to quantize the signal, dynamically adjusting the quantization step size based on the signal's statistical characteristics to reduce quantization errors.

[0040] Data Analysis Module: This module introduces a deep learning-based multi-scale feature fusion network (MSFF-Net) to analyze acoustic signals. This network combines the advantages of convolutional neural networks (CNNs) and long short-term memory networks (LSTMs). The CNN employs a carefully customized multi-layer convolutional structure. The kernel size varies from 3×3 to 5×5, rationally configured according to the frequency characteristics and local details of the acoustic signal. Through different levels of convolution operations, it can keenly capture the local features of the acoustic signal in both the time and frequency domains, such as intensity changes and subtle waveform fluctuations within specific frequency bands. After each convolutional layer, batch normalization is applied to accelerate network convergence and improve training stability. The LSTM is designed to accurately capture the temporal characteristics of acoustic signals. Its internal memory units are carefully tuned, and the weight parameters of the input, forget, and output gates are continuously optimized through training with a large amount of acoustic data. It can effectively handle long-term dependencies in the time dimension of the signal, accurately grasping information such as the signal's start, duration, and periodic changes. A multi-scale feature fusion layer is used to fuse features from different scales, improving the classification and recognition capabilities of acoustic signals. The network's loss function employs an improved cross-entropy loss function. Where y ij For real labels, p ij To predict probabilities, N is the sample size, M is the number of classes, λ is the regularization coefficient, and W... l Let be the weight matrix of the l-th layer of the network.

[0041] Early Warning Module: Employing a dynamic early warning strategy based on risk assessment, this module comprehensively and meticulously considers multiple factors to construct a risk assessment model. For acoustic signal intensity, high-precision sensors are used for real-time monitoring, dividing the signal into different intervals based on decibel levels. Regarding frequency, algorithms such as Fast Fourier Transform are used for precise analysis to determine whether the signal falls within the low-frequency, mid-frequency, or high-frequency range. Duration is accurately calculated by recording the start and end times using timestamps. Time and location information are obtained from a supporting time synchronization system and positioning device. When quantifying risk using fuzzy comprehensive evaluation, the formula is applied... Calculate the risk assessment value R, w i μ is the weight of the i-th factor. i Let represent the membership degree of the i-th factor. The warning threshold and warning level are dynamically adjusted based on the risk assessment value. When the risk assessment value exceeds a certain threshold, a high-level warning is triggered, and relevant personnel are promptly notified through various means (such as SMS, voice, and images).

[0042] This invention also includes a calibration module. This module employs an adaptive calibration method based on a virtual reference signal. During system operation, a virtual reference signal is generated in real time, simulating the characteristics of a standard acoustic signal. By comparing the difference between the monitored signal and the virtual reference signal, an adaptive algorithm automatically adjusts system parameters, such as gain and filter coefficients. The adaptive adjustment formula is: Where θ n Let J(θ) be the parameter vector for the nth iteration, μ be the learning rate, and J(θ) be the parameter vector for the nth iteration. n The loss function is used to measure the difference between the monitored signal and the virtual reference signal. Simultaneously, a calibration history database is established to record and analyze the parameters and results of each calibration, in order to optimize the calibration strategy.

[0043] This invention also includes a data storage module. This module employs an architecture combining distributed encrypted storage and edge computing. Preliminary data processing and analysis are performed on edge devices at the monitoring site to extract key features. The processed data is then encrypted and stored on distributed storage nodes. Blockchain technology is used to ensure data security and immutability, and smart contracts are used to implement data access control and sharing. The redundancy of data storage is dynamically adjusted based on the importance and real-time nature of the data, using the formula: R redundancy = α·I + β·T, where R redundancy Here, I represents redundancy, T represents data importance, and α and β represent adjustment coefficients.

[0044] In this invention, a nanofluid with acoustic amplification function is filled into the porous structure of the optical fiber in the fiber optic sensing module. This nanofluid consists of nanoparticles and a base fluid; the type and concentration of the nanoparticles are optimized to improve the amplification effect on sound waves. When sound waves act on the optical fiber, the nanoparticles in the nanofluid vibrate and scatter, enhancing the photoacoustic interaction within the fiber. The acoustic amplification factor is related to the nanoparticle concentration and the sound wave frequency according to the formula: A amplify =A0·(1+k·C·f), where A amplify A0 is the amplified sound wave amplitude, C is the nanoparticle concentration, f is the sound wave frequency, and k is the amplification factor.

[0045] In this invention, a photoelectric conversion enhancement structure based on a photonic crystal is employed in the signal conversion module. This structure consists of a photonic crystal and a photoelectric conversion material. The photonic crystal possesses unique photonic bandgap characteristics, enabling it to localize light of specific frequencies within the photoelectric conversion material, thereby improving both absorption and conversion efficiency. By optimizing the structural parameters of the photonic crystal, such as lattice constant and fill ratio, efficient conversion of optical signals of different frequencies is achieved. The relationship between light absorption efficiency and the photonic crystal structural parameters can be fitted using numerical simulations and experimental data to obtain an empirical formula: η absorb = a·L+b·r+c, where η absorb denoted as light absorption efficiency, L as lattice constant, r as filling ratio, and a, b, and c as fitting coefficients.

[0046] In this invention, the signal processing module employs an adaptive noise cancellation technique based on a genetic algorithm. This technique begins with the random generation of an initial population of filter coefficients, which determine the filter's ability to suppress noise at different frequencies. The core of the genetic algorithm lies in the fitness function, which is defined as... Monitoring signals are acquired through high-precision sensors and advanced signal acquisition circuits. i The noise estimation signal is obtained using a noise estimation model based on statistical analysis and machine learning. N is determined by both the signal acquisition duration and the sampling frequency. During algorithm iteration, the selection operation uses fitness values ​​and strategies such as roulette wheel selection to choose superior individuals, giving filter coefficients with high fitness a greater probability of entering the next generation. The crossover operation exchanges gene segments between individuals through single-point crossover, multi-point crossover, etc., generating new potentially superior combinations. The mutation operation randomly changes some genes of individuals with a small probability to maintain population diversity. By continuously repeating these operations, the population continues to evolve until the optimal filter coefficients are found, achieving efficient noise cancellation.

[0047] In this invention, the data analysis module employs a knowledge fusion method based on transfer learning. When faced with new monitoring scenarios or data types, existing relevant model knowledge is utilized for transfer learning, allowing for rapid adaptation to new tasks through fine-tuning of model parameters. Simultaneously, knowledge from different data sources is fused, such as historical monitoring data and expert experience, to improve the analysis and understanding of acoustic signals. The weight allocation for knowledge fusion is dynamically adjusted based on the reliability and relevance of the data sources, using the following formula: Where w source For the weights of the data source, R source C is a reliability metric for the data source. source This is a correlation metric between the data source and the current task, where m is the number of data sources.

[0048] In this invention, the early warning module employs a collaborative early warning mechanism based on multi-sensor fusion. In addition to the acoustic signal monitored by the fiber optic sensing module, data from other types of sensors, such as vibration sensors and pressure sensors, are also fused. A data fusion algorithm is used to process and analyze the multi-sensor data, improving the accuracy and reliability of the early warning. The data fusion algorithm employs the Kalman filter algorithm, and its state prediction equation is: The update equation is: in Let F be the state estimate, B be the state transition matrix, u be the control input matrix, z be the measurement value, H be the measurement matrix, and K be the Kalman gain.

[0049] In this invention, the calibration module employs a real-time calibration method based on online learning to ensure the long-term stable and accurate operation of the system. During system operation, the data acquisition unit uses a high-precision sensor as its front end, coupled with a high-speed data transmission line, to continuously collect various new calibration data. This data covers multiple parameters during system operation, such as the impact of ambient temperature and humidity on measurement results. The online learning algorithm used is the stochastic gradient descent algorithm, which has been specifically optimized. The learning rate μ is not a fixed value but is dynamically adjusted based on the changes in the loss function in each iteration through an adaptive adjustment strategy. The loss function L is constructed by comprehensively considering multiple error factors, such as the absolute error and relative error between the measured value and the true value. In each iteration, based on the nth calibration data sample x... n and its corresponding true value y n Update formula according to parameters For the parameter vector θ in the nth iteration n The system is updated and the calibration model is corrected in real time, which effectively reduces errors and ensures the accuracy and stability of the system in complex and ever-changing environments.

[0050] In this invention, the data storage module employs a high-efficiency data storage strategy based on intelligent caching. Intelligent caching is set up on edge devices, and data is stored hierarchically according to its access frequency and importance. Frequently accessed and important data is preferentially stored in the high-speed cache; infrequently used data is stored on distributed storage nodes. The cache replacement strategy uses the Least Recently Used (LRU) algorithm to improve data access efficiency. The relationship between cache hit rate, cache size, and data access pattern can be obtained through simulation experiments using an empirical formula: H... cache =a·S cache +b·M access +c, where H cache For cache hit rate, S cache M is the cache size. access Here, is a data access pattern indicator, and a, b, and c are fitting coefficients.

[0051] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An optical fiber sensing based acoustic wave monitoring system, characterized by, The following modules are included: The optical fiber sensing module adopts a microstructure optical fiber with a porous structure inside, and an intelligent composite material is coated on the surface of the optical fiber, and the material property adjustment follows the formula: wherein is the adjusted elastic modulus, is the initial elastic modulus, A is the amplitude of the acoustic wave, f is the frequency of the acoustic wave, and is the material property adjustment coefficient; Signal conversion module: use multi-mode photoelectric conversion technology, combined with PIN photodiode and avalanche photodiode APD, according to the intensity of optical signal dynamic switching conversion mode, the formula is: Wherein The gain after switching, APD gain, PIN photodiode gain, P is the input optical power, Power threshold; Signal processing module: a hybrid filtering algorithm is adopted, which combines the traditional band-pass filter with the wavelet transform filter; the decomposition and reconstruction process of the wavelet transform is based on the improved Mallat algorithm, and the iteration formula is as follows: , wherein is an approximate coefficient, is a detail coefficient, and h[k] and g[k] are wavelet filter coefficients; Data analysis module: a multi-scale feature fusion network based on deep learning is introduced to analyze the sound wave signal, which combines the advantages of convolutional neural network and long short-term memory network, and fuses the features of different scales through a multi-scale feature fusion layer. The network loss function adopts an improved cross-entropy loss function: wherein is a true label, is a prediction probability, N is the number of samples, and M is the number of categories, is a regularization coefficient, is a weight matrix; Early warning module: adopt dynamic early warning strategy based on risk assessment, quantify risk through fuzzy comprehensive evaluation method, formula is: Wherein R is risk assessment value, is weight, is membership degree, when risk assessment value exceeds certain threshold, trigger early warning and notify relevant personnel; In the optical fiber sensing module, the nano-fluid with acoustic amplification function is filled in the porous structure of the optical fiber. When the acoustic wave acts on the optical fiber, the nano-particles in the nano-fluid vibrate and scatter, enhancing the light-acoustic interaction in the optical fiber. The relationship between the acoustic amplification multiple and the nano-particle concentration and the acoustic wave frequency follows the formula: wherein is the amplified acoustic wave amplitude, is the initial acoustic wave amplitude, C is the nano-particle concentration, f is the acoustic wave frequency, and k is the amplification coefficient.

2. The acoustic wave monitoring system based on optical fiber sensing according to claim 1, characterized in that, Also included is a calibration module that uses an adaptive calibration method based on a virtual reference signal. During system operation, a virtual reference signal is generated, and by comparing the difference between the monitoring signal and the virtual reference signal, the parameters of the system are automatically adjusted using an adaptive algorithm. The adaptive adjustment formula is: wherein is the parameter vector of the nth iteration, is the learning rate, is the loss function, and at the same time, a calibration history database is established to record and analyze the parameters and results of each calibration.

3. The acoustic wave monitoring system based on optical fiber sensing according to claim 1, characterized in that, Also includes a data storage module, preliminary data processing on the edge device of the monitoring site, extracts key features, and then stores the processed data in encrypted form on the distributed storage node, uses blockchain technology, realizes the access control and sharing of data through smart contract, the redundancy of data storage is dynamically adjusted according to the importance and real-time of data, the formula is: Wherein The redundancy, I is the data importance index, T is the data real-time index, And Adjustment coefficient.

4. The acoustic wave monitoring system based on optical fiber sensing according to claim 1, characterized in that, In the signal conversion module, a light-electricity conversion enhancement structure based on photonic crystal is adopted, which is composed of photonic crystal and photoelectric conversion material. By optimizing the structure parameters of the photonic crystal, conversion of different frequency optical signals is realized. The relationship between the light absorption efficiency and the structure parameters of the photonic crystal is fitted through numerical simulation and experimental data to obtain an empirical formula: Wherein is the light absorption efficiency, L is the lattice constant, r is the filling ratio, a, b and c are fitting coefficients.

5. The acoustic wave monitoring system based on fiber optic sensing of claim 1, wherein, In the signal processing module, the adaptive noise cancellation technology based on genetic algorithm optimization is adopted. The technology searches the optimal filter coefficients through genetic algorithm. The fitness function of genetic algorithm is defined as the reciprocal of the sum of squares of errors between the monitoring signal and the noise estimation signal. The population is evolved through selection, crossover and mutation operations. The fitness function formula is: where F is the fitness value, is the monitoring signal, is the noise estimation signal, and N is the signal length.

6. The acoustic wave monitoring system based on fiber optic sensing of claim 1, wherein, In the data analysis module, a knowledge fusion method based on transfer learning is adopted. When facing a new monitoring scene, transfer learning is used to utilize the existing related model knowledge. The model parameters are fine-tuned to adapt to the new task. At the same time, the knowledge of different data sources is fused. The weight distribution of knowledge fusion is dynamically adjusted according to the reliability and correlation of the data sources. The formula is: wherein is the data source weight, is the reliability index of the data source, is the correlation index of the data source and the current task, and m is the number of data sources.

7. The acoustic wave monitoring system based on fiber optic sensing of claim 1, wherein, In the early warning module, a cooperative early warning mechanism based on multi-sensor fusion is adopted, and multi-sensor data are processed and analyzed through a data fusion algorithm. The Kalman filtering algorithm is adopted in the data fusion algorithm, and the state prediction equation is: The update equation is: Wherein is the state estimation value, F is the state transition matrix, B is the control input matrix, u is the control input, z is the measurement value, H is the measurement matrix, and K is the Kalman gain.

8. The acoustic wave monitoring system based on fiber optic sensing of claim 2, wherein, In the calibration module, an online learning based real-time calibration method is adopted. During the system running, new calibration data is collected, and the calibration model is updated in real time by using the online learning algorithm. The online learning algorithm adopts the stochastic gradient descent algorithm, and the parameter update formula is: wherein is the parameter vector of the n th iteration, is the learning rate, and L is the loss function, is the n th calibration data sample, is the corresponding true value.

9. The acoustic wave monitoring system based on optical fiber sensing according to claim 3, characterized in that, In the data storage module, an intelligent cache is set on the edge device, the data is stored hierarchically according to the access frequency and importance of the data, the replacement strategy of the cache adopts the least recently used LRU algorithm, the access efficiency of the data is improved, and the relationship between the cache hit rate and the cache size and the data access mode is obtained through an empirical formula by simulation experiment: wherein is the cache hit rate, is the cache size, is the data access mode index, and a, b and c are fitting coefficients.

Images