CN114861737A - Remote disturbance feature extraction method and system of distributed optical fiber sensing system - Google Patents

Abstract

The invention belongs to the technical field of signal processing of a distributed optical fiber sensing system, and discloses a remote disturbance feature extraction method of the distributed optical fiber sensing system, which comprises the following steps: dividing distance dimensional intervals of a signal Rayleigh scattering spectrum in the optical fiber channel, establishing corresponding dictionary learning models for the intervals, and learning and observing an original dictionary to form a polymorphic equivalent dictionary; based on the polymorphic equivalent dictionary, establishing a joint sparse representation model under the cascade of the polymorphic equivalent dictionaries to form joint sparse representation of the multi-echo scattering spectrum; and constructing a combined optimization reconstruction algorithm based on the polymorphic cascade dictionary, and extracting the characteristics of the remote disturbance signal. The problem that far-end interference signals in a distributed optical fiber sensing system are submerged due to inherent attenuation of Rayleigh scattering spectrums can be solved. The accuracy of extracting the characteristics of the disturbance signals is greatly improved.

Description

Translated fromChinese

分布式光纤传感系统的远端扰动特征提取方法及系统Remote disturbance feature extraction method and system for distributed optical fiber sensing system

技术领域technical field

本发明属于分布式光纤传感系统的信号处理技术领域，本发明涉及一种分布式光纤传感系统的远端扰动特征提取方法及系统。The invention belongs to the technical field of signal processing of a distributed optical fiber sensing system, and relates to a method and system for extracting remote disturbance features of a distributed optical fiber sensing system.

背景技术Background technique

光纤传感器凭借其不受电磁干扰、灵活性高以及容易组网的优势，在安防、监测与勘测领域中得到了广泛的应用。分布式光纤传感系统将光纤上的点均作为独立的传感单元，利用相位敏感(OTDR)技术，相比于传统的点式光纤传感系统，在测量范围、测量灵敏度以及响应能力上均有显著的提高。但同样的，分布式光纤传感系统的高灵敏度与快响应速度会造成系统对噪声与环境扰动的敏感性。同时，在长距离传输探测场景下，由于瑞利散射谱自身的衰落特性，将导致远端干扰信号被近端回波信号所淹没，从而影响对干扰信号的准确提取。Optical fiber sensors have been widely used in the fields of security, monitoring and surveying due to their advantages of being immune to electromagnetic interference, high flexibility and easy networking. The distributed optical fiber sensing system uses the points on the optical fiber as independent sensing units, and uses phase sensitive (OTDR) technology. Compared with the traditional point-based optical fiber sensing system, the measurement range, measurement sensitivity and response ability are all There is a significant improvement. But in the same way, the high sensitivity and fast response speed of the distributed optical fiber sensing system will make the system sensitive to noise and environmental disturbance. At the same time, in the long-distance transmission detection scenario, due to the fading characteristics of the Rayleigh scattering spectrum itself, the far-end interference signal will be submerged by the near-end echo signal, thus affecting the accurate extraction of the interference signal.

申请号201810359506X公开了一种基于扰动信号特征提取的分布式光纤振动传感定位方法及装置，将待检测光缆划分为若干个区间，并在每个区间分别接收振动产生的脉冲峰信号后记录脉冲峰个数，其中，第i帧第j段光缆上的脉冲峰个数记为N(i，j)；分别统计光缆每个区间内脉冲峰个数的平均值和方差；在每个光缆区间内，将所述脉冲峰个数的平均值和方差和预设参数进行比较，根据比较结果确认当前区间的光缆是否存在扰动信号，以此类推，确认所述待检测光缆存在扰动信号的各个区间；获取存在扰动信号的光缆区间，再分别确认所述区间内扰动点位置。该方法基于统计得到，需要有一定量的数据。实际操作繁琐，在数据量较小时，往往无法得到准确的结果。Application No. 201810359506X discloses a distributed optical fiber vibration sensing positioning method and device based on feature extraction of disturbance signals. The optical cable to be detected is divided into several sections, and the pulse peak signals generated by vibration are respectively received in each section. The number of peaks, among which, the number of pulse peaks on the jth segment of the optical cable in the i-th frame is recorded as N(i, j); the average and variance of the number of pulse peaks in each interval of the optical cable are counted separately; in each optical cable interval , compare the average value and variance of the number of pulse peaks with the preset parameters, and confirm whether there is a disturbance signal in the optical cable in the current interval according to the comparison result, and so on, confirm that the optical cable to be detected has a disturbance signal in each interval ; Obtain the optical cable section where the disturbance signal exists, and then confirm the position of the disturbance point in the interval respectively. This method is based on statistics and requires a certain amount of data. The actual operation is cumbersome, and when the amount of data is small, it is often impossible to obtain accurate results.

申请号2017103708538公开了一种相位敏感光时域反射分布式光纤传感系统快速定位方法，构建多个光脉冲对应的瑞利散射光数字信号矩阵、在信号矩阵上间隔一定长度选择测试窗口和测试列、获得各个测试列的相位、根据相邻测试窗口测试列相位对扰动源区间粗略定位、提取包含扰动源的区间信号进行扰动精确定位。该方法没有解决由于瑞利散射谱自身的衰落特性，导致远端干扰信号被近端回波信号所淹没的技术问题，无法提取得到准确的扰动特征。Application No. 2017103708538 discloses a fast positioning method for a phase-sensitive optical time-domain reflectance distributed optical fiber sensing system, constructing a digital signal matrix of Rayleigh scattered light corresponding to multiple optical pulses, selecting a test window at a certain length on the signal matrix, and testing column, obtain the phase of each test column, roughly locate the disturbance source interval according to the phase of the adjacent test window test column, extract the interval signal containing the disturbance source for precise disturbance location. This method does not solve the technical problem that the far-end interference signal is submerged by the near-end echo signal due to the fading characteristics of the Rayleigh scattering spectrum itself, and cannot extract accurate disturbance features.

发明内容SUMMARY OF THE INVENTION

本发明的目的在于提供一种分布式光纤传感系统的远端扰动特征提取方法及系统，针对不同的瑞利散射谱关注区间，对原有的字典进行学习观测形成多态等效字典，随后基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示，最后，通过优化重构算法完成对远端扰动信号的识别与提取。提高了扰动信号特征提取的准确性。The purpose of the present invention is to provide a remote disturbance feature extraction method and system for a distributed optical fiber sensing system. According to different Rayleigh scattering spectrum attention intervals, the original dictionary is learned and observed to form a polymorphic equivalent dictionary, and then a polymorphic equivalent dictionary is formed. Based on the polymorphic equivalent dictionary, a joint sparse representation model under the cascade of polymorphic equivalent dictionaries is established to form a joint sparse representation of the multi-echo scattering spectrum. extract. The accuracy of feature extraction of disturbance signal is improved.

实现本发明目的的技术解决方案为：The technical solution that realizes the object of the present invention is:

一种分布式光纤传感系统的远端扰动特征提取方法，包括以下步骤：A method for extracting remote disturbance features of a distributed optical fiber sensing system, comprising the following steps:

S01：对光纤信道内信号瑞利散射谱的距离维区间进行划分，并针对各区间建立相应的字典学习模型，对原有的字典进行学习观测形成多态等效字典；S01: Divide the distance dimension interval of the Rayleigh scattering spectrum of the signal in the fiber channel, establish a corresponding dictionary learning model for each interval, and learn and observe the original dictionary to form a polymorphic equivalent dictionary;

S02：基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示；S02: Based on the polymorphic equivalent dictionary, a joint sparse representation model under the cascade of polymorphic equivalent dictionaries is established to form a joint sparse representation of the multi-echo scattering spectrum;

S03：构建基于多态级联字典的联合优化重构算法，提取远端扰动信号特征。S03: Construct a joint optimization and reconstruction algorithm based on a polymorphic cascade dictionary to extract features of remote disturbance signals.

优选的技术方案中，所述步骤S01中形成多态等效字典的方法包括：In a preferred technical solution, the method for forming a polymorphic equivalent dictionary in the step S01 includes:

S11：针对所有的探测距离构造一个本征特征字典：S11: Construct an intrinsic feature dictionary for all detection distances:

Ψ＝[ψ₁,ψ₂,…,ψ_L]；Ψ=[ψ₁ ,ψ₂ ,...,ψ_L ];

其中，ψ_l表示在第l个探测距离上的单位响应；Among them, ψ_l represents the unit response at the l-th detection distance;

得到基态谱

表示为：get ground state spectrum

Expressed as:

其中，η为基态特征系数；Among them, η is the ground state characteristic coefficient;

S12：针对第n个扰动事件ζ_n，其扰动特征空间为

其扰动特征互补空间为

基于

得到对应的特征系数向量

为：S12: For the nth disturbance event ζ_n , its disturbance feature space is

Its perturbation feature complementary space is

based on

Get the corresponding eigencoefficient vector

for:

其中，η_l与

分别表示η和

的第l个元素；where η_l and

denote η and

the lth element of ;

S13：针对扰动事件ζ_n，构建扰动模态观测矩阵：S13: For the disturbance event ζ_n , construct the disturbance modal observation matrix:

其中，diag(z)表示以向量z为对角元素构成矩阵，β⁽ⁿ⁾为权系数向量；Among them, diag(z) represents a matrix with the vector z as the diagonal element, and β⁽ⁿ⁾ is the weight coefficient vector;

S14：基于上述扰动模态观测矩阵，对本征特征字典进行学习观测，形成扰动事件ζ_n所对应的模态字典：S14: Based on the above disturbance modal observation matrix, learn and observe the eigenfeature dictionary to form a modal dictionary corresponding to the disturbance event_ζn :

Ψ⁽ⁿ⁾＝Φ⁽ⁿ⁾Ψ；Ψ⁽ⁿ⁾ = Φ⁽ⁿ⁾ Ψ;

将所有扰动事件N所对应的模态字典与本征特征字典进行级联组合，得到多态等效字典：The modal dictionaries corresponding to all disturbance events N are combined with the eigenfeature dictionaries in cascade to obtain a polymorphic equivalent dictionary:

优选的技术方案中，所述步骤S02中形成多回波散射谱的联合稀疏表示的方法包括：In a preferred technical solution, the method for forming a joint sparse representation of multiple echo scattering spectra in step S02 includes:

将M个回波光脉冲组成的联合回波矩阵表示为：The joint echo matrix composed of M echo light pulses is expressed as:

其中，P_i表示第i个回波光脉冲对应的采样瑞利曲线，[·]^T表示向量与矩阵的转置；Among them, P_i represents the sampled Rayleigh curve corresponding to the ith echo light pulse, [ ]^T represents the transpose of the vector and the matrix;

基于多态等效字典，得到联合回波矩阵的联合稀疏表示模型为：Based on the polymorphic equivalent dictionary, the joint sparse representation model of the joint echo matrix is obtained as:

其中，Θ为联合特征矩阵，同时也是一个列稀疏矩阵，其非零列表征了相应多态字典所对应的扰动事件。Among them, Θ is the joint feature matrix, which is also a column sparse matrix, and its non-zero column represents the perturbation events corresponding to the corresponding polymorphic dictionary.

优选的技术方案中，所述步骤S03中提取远端扰动信号特征的方法包括：In a preferred technical solution, the method for extracting the features of the remote disturbance signal in the step S03 includes:

S31：对联合回波矩阵进行向量化得到r＝vec(R)，进而得到扩展联合稀疏表示：S31: Vectorize the joint echo matrix to obtain r=vec(R), and then obtain the extended joint sparse representation:

r＝Γθr=Γθ

其中，

I_(N+1)×(N+1)表示维度为(N+1)×(N+1)的单位阵，

表示克罗内克积，θ为重构特征向量；in,

I_(N+1)×(N+1) represents a unit matrix of dimension (N+1)×(N+1),

represents the Kronecker product, and θ is the reconstructed eigenvector;

S32：初始化重构特征向量θ⁽⁰⁾＝0，残差ε⁽⁰⁾＝r；索引集

设置终止迭代门限ξ；t＝1，Θ₀为空矩阵；S32: Initialize reconstruction feature vector θ⁽⁰⁾ = 0, residual ε⁽⁰⁾ = r; index set

Set the termination iteration threshold ξ; t=1, Θ₀ is an empty matrix;

S33：将Γ中各模态字典Ψ⁽ⁿ⁾与第l列抽取组成字矩阵Γ^l，将每个Γ^l,l＝1,2…L与残差ε相乘，找出积为最大值所对应的索引为λ_t，即

S33: Extract each modal dictionary Ψ⁽ⁿ⁾ and the l-th column in Γ to form a word matrix Γ^l , multiply each Γ^l , l=1, 2...L with the residual ε, and find the product as the maximum value The corresponding index is λ_t , that is,

S34：更新索引集Λ_t＝Λ_t-1∪{λ_t}，记录找到的多态等效字典中与残差相关度最高的列组合，并重建原子集合为γ_t＝[γ_t-1,Γ^l]；S34: Update the index set Λ_t = Λ_t-1 ∪{λ_t }, record the column combination with the highest degree of correlation with the residual in the found polymorphic equivalent dictionary, and reconstruct the atomic set as γ_t = [γ_t-1 , Γ^l ];

S35：求解得到θ^(t)＝argmin||r-γ_tθ^(t)||₂，更新残差ε^(t)＝r-γ_tθ^(t)，t＝t+1；S35: Solve to obtain θ^(t) = argmin||r-γ_t θ^(t) ||₂ , update the residual ε^(t) =r-γ_t θ^(t) , t=t+1;

S36：判断残差ε^(t)是否小于ξ，如是，则迭代停止，输出重构特征向量θ；反之，则跳转至步骤S33继续执行；S36: determine whether the residual ε^(t) is less than ξ, if so, stop the iteration, and output the reconstructed feature vector θ; otherwise, jump to step S33 to continue execution;

S37：由重构特征向量θ重构得到联合特征矩阵Θ。S37: Reconstruction from the reconstructed feature vector θ to obtain a joint feature matrix Θ.

优选的技术方案中，所述步骤S03之后还包括，将重构得到的联合特征矩阵Θ的前L列置零，随后确定剩余的非零列，获得非零列索引向量υ，由索引向量υ指向扰动事件类型。In a preferred technical solution, after the step S03, it further includes: setting the first L columns of the reconstructed joint feature matrix Θ to zero, then determining the remaining non-zero columns, and obtaining a non-zero column index vector υ, which is determined by the index vector υ Points to the disturbance event type.

本发明还公开了一种分布式光纤传感系统的远端扰动特征提取系统，包括：The invention also discloses a remote disturbance feature extraction system of the distributed optical fiber sensing system, comprising:

多态等效字典构建模块，对光纤信道内信号瑞利散射谱的距离维区间进行划分，并针对各区间建立相应的字典学习模型，对原有的字典进行学习观测形成多态等效字典；The polymorphic equivalent dictionary building module divides the distance dimension interval of the Rayleigh scattering spectrum of the signal in the fiber channel, establishes a corresponding dictionary learning model for each interval, and learns and observes the original dictionary to form a polymorphic equivalent dictionary;

多回波散射谱的联合稀疏表示模块，基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示；The joint sparse representation module of multiple echo scattering spectra, based on the polymorphic equivalent dictionary, establishes a joint sparse representation model under the cascade of polymorphic equivalent dictionaries to form a joint sparse representation of multiple echo scattering spectra;

扰动信号特征提取模块，构建基于多态级联字典的联合优化重构算法，提取远端扰动信号特征。The disturbance signal feature extraction module constructs a joint optimization and reconstruction algorithm based on polymorphic cascade dictionary to extract the remote disturbance signal features.

优选的技术方案中，所述形成多态等效字典的方法包括：In a preferred technical solution, the method for forming a polymorphic equivalent dictionary includes:

S11：针对所有的探测距离构造一个本征特征字典：S11: Construct an intrinsic feature dictionary for all detection distances:

Ψ＝[ψ₁,ψ₂,…,ψ_L]；Ψ=[ψ₁ ,ψ₂ ,...,ψ_L ];

其中，ψ_l表示在第l个探测距离上的单位响应；Among them, ψ_l represents the unit response at the l-th detection distance;

得到基态谱

表示为：get ground state spectrum

Expressed as:

其中，η为基态特征系数；Among them, η is the ground state characteristic coefficient;

S12：针对第n个扰动事件ζ_n，其扰动特征空间为

其扰动特征互补空间为

基于

得到对应的特征系数向量

为：S12: For the nth disturbance event ζ_n , its disturbance feature space is

Its perturbation feature complementary space is

based on

Get the corresponding eigencoefficient vector

for:

其中，η_l与

分别表示η和

的第l个元素；where η_l and

denote η and

the lth element of ;

S13：针对扰动事件ζ_n，构建扰动模态观测矩阵：S13: For the disturbance event ζ_n , construct the disturbance modal observation matrix:

其中，diag(z)表示以向量z为对角元素构成矩阵，β⁽ⁿ⁾为权系数向量；Among them, diag(z) represents a matrix with the vector z as the diagonal element, and β⁽ⁿ⁾ is the weight coefficient vector;

S14：基于上述扰动模态观测矩阵，对本征特征字典进行学习观测，形成扰动事件ζ_n所对应的模态字典：S14: Based on the above disturbance modal observation matrix, learn and observe the eigenfeature dictionary to form a modal dictionary corresponding to the disturbance event_ζn :

Ψ⁽ⁿ⁾＝Φ⁽ⁿ⁾Ψ；Ψ⁽ⁿ⁾ = Φ⁽ⁿ⁾ Ψ;

将所有扰动事件N所对应的模态字典与本征特征字典进行级联组合，得到多态等效字典：The modal dictionaries corresponding to all disturbance events N are combined with the eigenfeature dictionaries in cascade to obtain a polymorphic equivalent dictionary:

优选的技术方案中，所述多回波散射谱的联合稀疏表示的方法包括：In a preferred technical solution, the method for joint sparse representation of multiple echo scattering spectra includes:

将M个回波光脉冲组成的联合回波矩阵表示为：The joint echo matrix composed of M echo light pulses is expressed as:

其中，P_i表示第i个回波光脉冲对应的采样瑞利曲线，[·]^T表示向量与矩阵的转置；Among them, P_i represents the sampled Rayleigh curve corresponding to the ith echo light pulse, [ ]^T represents the transpose of the vector and the matrix;

基于多态等效字典，得到联合回波矩阵的联合稀疏表示模型为：Based on the polymorphic equivalent dictionary, the joint sparse representation model of the joint echo matrix is obtained as:

其中，Θ为联合特征矩阵，同时也是一个列稀疏矩阵，其非零列表征了相应多态字典所对应的扰动事件。Among them, Θ is the joint feature matrix, which is also a column sparse matrix, and its non-zero column represents the perturbation events corresponding to the corresponding polymorphic dictionary.

优选的技术方案中，所述提取远端扰动信号特征的方法包括：In a preferred technical solution, the method for extracting features of remote disturbance signals includes:

S31：对联合回波矩阵进行向量化得到r＝vec(R)，进而得到扩展联合稀疏表示：S31: Vectorize the joint echo matrix to obtain r=vec(R), and then obtain the extended joint sparse representation:

r＝Γθr=Γθ

其中，

I_(N+1)×(N+1)表示维度为(N+1)×(N+1)的单位阵，

表示克罗内克积，θ为重构特征向量；in,

I_(N+1)×(N+1) represents a unit matrix of dimension (N+1)×(N+1),

represents the Kronecker product, and θ is the reconstructed eigenvector;

S32：初始化重构特征向量θ⁽⁰⁾＝0，残差ε⁽⁰⁾＝r；索引集

设置终止迭代门限ξ；t＝1，Θ₀为空矩阵；S32: Initialize reconstruction feature vector θ⁽⁰⁾ = 0, residual ε⁽⁰⁾ = r; index set

Set the termination iteration threshold ξ; t=1, Θ₀ is an empty matrix;

S33：将Γ中各模态字典Ψ⁽ⁿ⁾与第l列抽取组成字矩阵Γ^l，将每个Γ^l,l＝1,2…L与残差ε相乘，找出积为最大值所对应的索引为λ_t，即

S33: Extract each modal dictionary Ψ⁽ⁿ⁾ and the l-th column in Γ to form a word matrix Γ^l , multiply each Γ^l , l=1, 2...L with the residual ε, and find the product as the maximum value The corresponding index is λ_t , that is,

S34：更新索引集Λ_t＝Λ_t-1∪{λ_t}，记录找到的多态等效字典中与残差相关度最高的列组合，并重建原子集合为γ_t＝[γ_t-1,Γ^l]；S34: Update the index set Λ_t = Λ_t-1 ∪{λ_t }, record the column combination with the highest degree of correlation with the residual in the found polymorphic equivalent dictionary, and reconstruct the atomic set as γ_t = [γ_t-1 , Γ^l ];

S35：求解得到

更新残差ε^(t)＝r-γ_tθ^(t)，t＝t+1；S35: Solve to get

Update residual ε^(t) = r-γ_t θ^(t) , t=t+1;

S36：判断残差ε^(t)是否小于ξ，如是，则迭代停止，输出重构特征向量θ；反之，则跳转至步骤S33继续执行；S36: determine whether the residual ε^(t) is less than ξ, if so, stop the iteration, and output the reconstructed feature vector θ; otherwise, jump to step S33 to continue execution;

S37：由重构特征向量θ重构得到联合特征矩阵Θ。S37: Reconstruction from the reconstructed feature vector θ to obtain a joint feature matrix Θ.

优选的技术方案中，还包括扰动事件类型识别模块，将重构得到的联合特征矩阵Θ的前L列置零，随后确定剩余的非零列，获得非零列索引向量υ，由索引向量υ指向扰动事件类型。In a preferred technical solution, it also includes a disturbance event type identification module, which sets the first L columns of the reconstructed joint feature matrix Θ to zero, then determines the remaining non-zero columns, and obtains a non-zero column index vector υ, which is determined by the index vector υ Points to the disturbance event type.

本发明与现有技术相比，其显著优点为：Compared with the prior art, the present invention has the following significant advantages:

本发明可以解决针对分布式光纤传感系统中远端干扰信号因瑞利散射谱固有衰减而导致的被淹没问题。针对不同的瑞利散射谱关注区间，对原有的字典进行学习观测形成多态等效字典，随后基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示，最后，通过优化重构算法完成对远端扰动信号的识别与提取。大大提高了扰动信号特征提取的准确性，还可以对扰动事件类型进行识别。The invention can solve the problem of submersion caused by the inherent attenuation of the Rayleigh scattering spectrum for the remote interference signal in the distributed optical fiber sensing system. For different attention intervals of Rayleigh scattering spectrum, the original dictionary is learned and observed to form a polymorphic equivalent dictionary, and then based on the polymorphic equivalent dictionary, a joint sparse representation model under the cascade of polymorphic equivalent dictionaries is established to form a polymorphic equivalent dictionary. The joint sparse representation of the multi-echo scattering spectrum, and finally, the identification and extraction of the far-end disturbance signal is completed by optimizing the reconstruction algorithm. The accuracy of feature extraction of disturbance signals is greatly improved, and the types of disturbance events can also be identified.

附图说明Description of drawings

图1为较佳实施例的分布式光纤传感系统的远端扰动特征提取方法的流程图；1 is a flowchart of a method for extracting remote disturbance features of a distributed optical fiber sensing system according to a preferred embodiment;

图2为较佳实施例的分布式光纤传感系统的远端扰动特征提取系统的原理框图。FIG. 2 is a schematic block diagram of a remote disturbance feature extraction system of a distributed optical fiber sensing system according to a preferred embodiment.

具体实施方式Detailed ways

本发明的原理是：针对现有通用夹具和专用夹具的不足，设计一种简易装夹、不损伤精密零件、能完整检测各项形位误差，并且在同类型零件测量和检测对象尺寸存在偏差时具有通用性的夹具，可减少夹具拆卸和不必要的重复定位，从而减轻工人的劳动强度，提高测量效率。The principle of the invention is: aiming at the deficiencies of the existing general fixtures and special fixtures, a simple fixture is designed, which does not damage the precision parts, can completely detect various shape and position errors, and has deviations in the measurement of the same type of parts and the size of the detection objects It is a universal fixture, which can reduce fixture disassembly and unnecessary repeated positioning, thereby reducing the labor intensity of workers and improving measurement efficiency.

实施例1：Example 1:

如图1所示，一种分布式光纤传感系统的远端扰动特征提取方法，包括以下步骤：As shown in Figure 1, a method for extracting remote disturbance features of a distributed optical fiber sensing system includes the following steps:

S01：对光纤信道内信号瑞利散射谱的距离维区间进行划分，并针对各区间建立相应的字典学习模型，对原有的字典进行学习观测形成多态等效字典；S01: Divide the distance dimension interval of the Rayleigh scattering spectrum of the signal in the fiber channel, establish a corresponding dictionary learning model for each interval, and learn and observe the original dictionary to form a polymorphic equivalent dictionary;

S02：基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示；S02: Based on the polymorphic equivalent dictionary, a joint sparse representation model under the cascade of polymorphic equivalent dictionaries is established to form a joint sparse representation of the multi-echo scattering spectrum;

S03：构建基于多态级联字典的联合优化重构算法，提取远端扰动信号特征。S03: Construct a joint optimization and reconstruction algorithm based on a polymorphic cascade dictionary to extract features of remote disturbance signals.

需要说明的是：对于已完成布线的光缆，其在未收扰动情况下的瑞利散射谱具有固定的分量组合，本发明将上述固定的分量组合称为光缆瑞利散射谱的基态谱，简称基态谱，若考虑光缆中分布式传感器的间隔为Δ，同时探测总长为LΔ，为方便后续内容中公式的简洁性，本发明中假设Δ＝1，上述假设不会影响Δ在其他取值情况下本发明内容的一般性。It should be noted that: for the optical cable that has been wired, its Rayleigh scattering spectrum in the case of no disturbance has a fixed combination of components, and the present invention refers to the above fixed component combination as the ground state spectrum of the optical cable Rayleigh scattering spectrum, referred to as the ground state spectrum of the optical cable Rayleigh scattering spectrum. For the ground state spectrum, if the interval of the distributed sensors in the optical cable is considered to be Δ, and the total detection length is LΔ, in order to facilitate the simplicity of the formula in the subsequent content, it is assumed that Δ=1 in the present invention, and the above assumption will not affect other values of Δ. The generality of the present disclosure follows.

一较佳的实施例，步骤S01中形成多态等效字典的方法包括：A preferred embodiment, the method for forming a polymorphic equivalent dictionary in step S01 includes:

S11：针对所有的探测距离构造一个本征特征字典：S11: Construct an intrinsic feature dictionary for all detection distances:

Ψ＝[ψ₁,ψ₂,…,ψ_L]；Ψ=[ψ₁ ,ψ₂ ,...,ψ_L ];

其中，ψ_l表示在第l个探测距离上的单位响应；Among them, ψ_l represents the unit response at the l-th detection distance;

得到基态谱

表示为：get ground state spectrum

Expressed as:

其中，η为基态特征系数；Among them, η is the ground state characteristic coefficient;

S12：针对第n个扰动事件ζ_n，其扰动特征空间为

其扰动特征互补空间为

基于

得到对应的特征系数向量

为：S12: For the nth disturbance event ζ_n , its disturbance feature space is

Its perturbation feature complementary space is

based on

Get the corresponding eigencoefficient vector

for:

其中，η_l与

分别表示η和

的第l个元素；where η_l and

denote η and

the lth element of ;

S13：针对扰动事件ζ_n，构建扰动模态观测矩阵：S13: For the disturbance event ζ_n , construct the disturbance modal observation matrix:

其中，diag(z)表示以向量z为对角元素构成矩阵，β⁽ⁿ⁾为权系数向量；Among them, diag(z) represents a matrix with the vector z as the diagonal element, and β⁽ⁿ⁾ is the weight coefficient vector;

S14：基于上述扰动模态观测矩阵，对本征特征字典进行学习观测，形成扰动事件ζ_n所对应的模态字典：S14: Based on the above disturbance modal observation matrix, learn and observe the eigenfeature dictionary to form a modal dictionary corresponding to the disturbance event_ζn :

Ψ⁽ⁿ⁾＝Φ⁽ⁿ⁾Ψ；Ψ⁽ⁿ⁾ = Φ⁽ⁿ⁾ Ψ;

将所有扰动事件N所对应的模态字典与本征特征字典进行级联组合，得到多态等效字典：The modal dictionaries corresponding to all disturbance events N are combined with the eigenfeature dictionaries in cascade to obtain a polymorphic equivalent dictionary:

一较佳的实施例，步骤S02中形成多回波散射谱的联合稀疏表示的方法包括：In a preferred embodiment, the method for forming a joint sparse representation of multiple echo scattering spectra in step S02 includes:

将M个回波光脉冲组成的联合回波矩阵表示为：The joint echo matrix composed of M echo light pulses is expressed as:

其中，P_i表示第i个回波光脉冲对应的采样瑞利曲线，[·]^T表示向量与矩阵的转置；Among them, P_i represents the sampled Rayleigh curve corresponding to the ith echo light pulse, [ ]^T represents the transpose of the vector and the matrix;

基于多态等效字典，得到联合回波矩阵的联合稀疏表示模型为：Based on the polymorphic equivalent dictionary, the joint sparse representation model of the joint echo matrix is obtained as:

其中，Θ为联合特征矩阵，同时也是一个列稀疏矩阵，其非零列表征了相应多态字典所对应的扰动事件。Among them, Θ is the joint feature matrix, which is also a column sparse matrix, and its non-zero column represents the perturbation events corresponding to the corresponding polymorphic dictionary.

一较佳的实施例，步骤S03中提取远端扰动信号特征的方法包括：In a preferred embodiment, the method for extracting the features of the remote disturbance signal in step S03 includes:

S31：对联合回波矩阵进行向量化得到r＝vec(R)，进而得到扩展联合稀疏表示：S31: Vectorize the joint echo matrix to obtain r=vec(R), and then obtain the extended joint sparse representation:

r＝Γθr=Γθ

其中，

I_(N+1)×(N+1)表示维度为(N+1)×(N+1)的单位阵，

表示克罗内克积，θ为重构特征向量；in,

I_(N+1)×(N+1) represents a unit matrix of dimension (N+1)×(N+1),

represents the Kronecker product, and θ is the reconstructed eigenvector;

S32：初始化重构特征向量θ⁽⁰⁾＝0，残差ε⁽⁰⁾＝r；索引集

设置终止迭代门限ξ；t＝1，Θ₀为空矩阵；S32: Initialize reconstruction feature vector θ⁽⁰⁾ = 0, residual ε⁽⁰⁾ = r; index set

Set the termination iteration threshold ξ; t=1, Θ₀ is an empty matrix;

S33：将Γ中各模态字典Ψ⁽ⁿ⁾与第l列抽取组成字矩阵Γ^l，将每个Γ^l,l＝1,2…L与残差ε相乘，找出积为最大值所对应的索引为λ_t，即

S33: Extract each modal dictionary Ψ⁽ⁿ⁾ and the l-th column in Γ to form a word matrix Γ^l , multiply each Γ^l , l=1, 2...L with the residual ε, and find the product as the maximum value The corresponding index is λ_t , that is,

S34：更新索引集Λ_t＝Λ_t-1∪{λ_t}，记录找到的多态等效字典中与残差相关度最高的列组合，并重建原子集合为γ_t＝[γ_t-1,Γ^l]；S34: Update the index set Λ_t = Λ_t-1 ∪{λ_t }, record the column combination with the highest degree of correlation with the residual in the found polymorphic equivalent dictionary, and reconstruct the atomic set as γ_t = [γ_t-1 , Γ^l ];

S35：求解得到θ^(t)＝argmin||r-γ_tθ^(t)||₂，更新残差ε^(t)＝r-γ_tθ^(t)，t＝t+1；S35: Solve to obtain θ^(t) = argmin||r-γ_t θ^(t) ||₂ , update the residual ε^(t) =r-γ_t θ^(t) , t=t+1;

S36：判断残差ε^(t)是否小于ξ，如是，则迭代停止，输出重构特征向量θ；反之，则跳转至步骤S33继续执行；S36: determine whether the residual ε^(t) is less than ξ, if so, stop the iteration, and output the reconstructed feature vector θ; otherwise, jump to step S33 to continue execution;

S37：由重构特征向量θ重构得到联合特征矩阵Θ。S37: Reconstruction from the reconstructed feature vector θ to obtain a joint feature matrix Θ.

另一实施例中，步骤S03之后还包括，将重构得到的联合特征矩阵Θ的前L列置零，随后确定剩余的非零列，获得非零列索引向量υ，由索引向量υ指向扰动事件类型。从而可以对扰动事件类型进行识别。In another embodiment, after step S03, it further includes: setting the first L columns of the reconstructed joint feature matrix Θ to zero, then determining the remaining non-zero columns, and obtaining a non-zero column index vector υ, which points to the disturbance from the index vector υ. Event type. Thus, the disturbance event type can be identified.

又一实施例中，如图2所示，本发明还公开了一种分布式光纤传感系统的远端扰动特征提取系统，包括：In yet another embodiment, as shown in FIG. 2 , the present invention also discloses a remote disturbance feature extraction system of a distributed optical fiber sensing system, including:

多态等效字典构建模块，对光纤信道内信号瑞利散射谱的距离维区间进行划分，并针对各区间建立相应的字典学习模型，对原有的字典进行学习观测形成多态等效字典；The polymorphic equivalent dictionary building module divides the distance dimension interval of the Rayleigh scattering spectrum of the signal in the fiber channel, establishes a corresponding dictionary learning model for each interval, and learns and observes the original dictionary to form a polymorphic equivalent dictionary;

多回波散射谱的联合稀疏表示模块，基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示；The joint sparse representation module of multiple echo scattering spectra, based on the polymorphic equivalent dictionary, establishes a joint sparse representation model under the cascade of polymorphic equivalent dictionaries to form a joint sparse representation of multiple echo scattering spectra;

扰动信号特征提取模块，构建基于多态级联字典的联合优化重构算法，提取远端扰动信号特征。The disturbance signal feature extraction module constructs a joint optimization and reconstruction algorithm based on polymorphic cascade dictionary to extract the remote disturbance signal features.

具体的，分布式光纤传感系统的远端扰动特征提取系统的工作流程如下：Specifically, the workflow of the remote disturbance feature extraction system of the distributed optical fiber sensing system is as follows:

一、首先，建立分布式光纤传感系统回波瑞利谱曲线的特征字典，对于已完成布线的光缆，其在未收扰动情况下的瑞利散射谱具有固定的分量组合，本发明将上述固定的分量组合称为光缆瑞利散射谱的基态谱，若考虑光缆中分布式传感器的间隔为Δ，同时探测总长为LΔ，为方便后续内容中公式的简洁性，本发明中假设Δ＝1，上述假设不会影响Δ在其他取值情况下本发明内容的一般性。令ψ_l表示在第l个探测距离上的单位响应，则可以针对所有的探测距离构造一个本征特征字典：1. First, a feature dictionary of the echo Rayleigh spectrum curve of the distributed optical fiber sensing system is established. For the optical cable that has been wired, its Rayleigh scattering spectrum under the condition of no disturbance has a fixed combination of components. The fixed combination of components is called the ground state spectrum of the Rayleigh scattering spectrum of the optical cable. If the interval of the distributed sensors in the optical cable is considered to be Δ, and the total detection length is LΔ, in order to facilitate the simplicity of the formula in the subsequent content, it is assumed in the present invention that Δ=1 , the above assumptions will not affect the generality of the content of the present invention under other values of Δ. Let ψ_l denote the unit response at the l-th detection distance, then an intrinsic feature dictionary can be constructed for all detection distances:

Ψ＝[ψ₁,ψ₂,…,ψ_L] (1)Ψ=[ψ₁ ,ψ₂ ,...,ψ_L ] (1)

此时基态谱

可以表示为：ground state spectrum

It can be expressed as:

其中η为基态特征系数。由于基态谱

是一个连续谱，因此上式所描述的不是一个理想的稀疏问题，因此可以采用模态分解等手段，从光缆的基态谱

中，分解得到基态特征系数向量η。where η is the ground state characteristic coefficient. Since the ground state spectrum

is a continuous spectrum, so the above formula is not an ideal sparse problem, so means such as mode decomposition can be used to obtain the ground state spectrum of the optical cable from the

, decompose to get the ground state characteristic coefficient vector η.

本发明基于基态特征系数η，设计了一种多态字典学习观测方法，具体为：The present invention designs a polymorphic dictionary learning and observation method based on the ground state characteristic coefficient η, specifically:

令分布式光纤传感系统所关心的扰动事件类型数量共为N，则第n个扰动事件记为ζ_n。基于前期的扰动事件样本训练，可以刻画各类扰动事件的扰动特征分布情况，即扰动特征空间。针对扰动事件ζ_n，令

表示其扰动特征空间，则进一步定义其扰动特征互补空间

基于

得到对应的特征系数向量

具体为：Let the number of disturbance event types concerned by the distributed optical fiber sensing system be N in total, then the nth disturbance event is recorded as ζ_n . Based on the previous sample training of disturbance events, the disturbance feature distribution of various disturbance events can be described, that is, the disturbance feature space. For the disturbance event ζ_n , let

represents its perturbed feature space, then further defines its perturbed feature complementary space

based on

Get the corresponding eigencoefficient vector

Specifically:

其中，η_l与

分别表示η和

的第l个元素。where η_l and

denote η and

the lth element of .

进一步的，针对扰动事件ζ_n，构建扰动模态观测矩阵：Further, for the disturbance event ζ_n , the disturbance modal observation matrix is constructed:

其中，diag(z)表示以向量z为对角元素构成矩阵，β⁽ⁿ⁾为权系数向量。基于上述扰动模态观测矩阵，即可对本征特征字典进行学习观测，形成扰动事件ζ_n所对应的模态字典：Among them, diag(z) indicates that the matrix is formed with the vector z as the diagonal elements, and β⁽ⁿ⁾ is the weight coefficient vector. Based on the above disturbance modal observation matrix, the intrinsic feature dictionary can be learned and observed to form the modal dictionary corresponding to the disturbance event ζ_n :

Ψ⁽ⁿ⁾＝Φ⁽ⁿ⁾Ψ (5)Ψ⁽ⁿ⁾ = Φ⁽ⁿ⁾ Ψ (5)

将所有关注的扰动事件所对应的模态字典与本征特征字典进行级联组合，即可得到多态等效字典：A polymorphic equivalent dictionary can be obtained by cascading the modal dictionaries corresponding to all concerned disturbance events with the eigenfeature dictionaries:

二、本发明随后在上述基础上对分布式光纤传感系统的多回波光脉冲进行联合优化表示。若P_i表示第i个回波光脉冲对应的采样瑞利曲线，则由M个回波光脉冲组成的联合回波矩阵可以表示为：2. The present invention then performs joint optimization and representation of the multi-echo optical pulses of the distributed optical fiber sensing system on the basis of the above. If P_i represents the sampled Rayleigh curve corresponding to the ith echo light pulse, the joint echo matrix composed of M echo light pulses can be expressed as:

其中，[·]^T表示向量与矩阵的转置。where [ ]^T represents the transpose of a vector and a matrix.

基于式(6)所示的多态级联字典，可以得到联合回波矩阵的联合稀疏表示形式：Based on the polymorphic cascade dictionary shown in Eq. (6), the joint sparse representation of the joint echo matrix can be obtained:

其中，Θ为联合特征矩阵，同时也是一个列稀疏矩阵，其非零列表征了相应多态字典所对应的扰动事件。Among them, Θ is the joint feature matrix, which is also a column sparse matrix, and its non-zero column represents the perturbation events corresponding to the corresponding polymorphic dictionary.

三、基于上述联合稀疏表示模型，从联合回波矩阵中重构得到联合特征矩阵的过程可以概括为以下步骤：3. Based on the above joint sparse representation model, the process of reconstructing the joint feature matrix from the joint echo matrix can be summarized as the following steps:

步骤1：对联合回波矩阵进行向量化得到r＝vec(R)，进而得到扩展联合稀疏表示：Step 1: Vectorize the joint echo matrix to obtain r=vec(R), and then obtain the extended joint sparse representation:

r＝Γθ (9)r=Γθ (9)

其中，

I_(N+1)×(N+1)表示维度为(N+1)×(N+1)的单位阵，

表示克罗内克积。in,

I_(N+1)×(N+1) represents a unit matrix of dimension (N+1)×(N+1),

represents the Kronecker product.

步骤2：初始化重构特征向量θ⁽⁰⁾＝0，残差ε⁽⁰⁾＝r；索引集

设置终止迭代门限ξ；t＝1，Θ₀为空矩阵；Step 2: Initialize reconstructed feature vector θ⁽⁰⁾ = 0, residual ε⁽⁰⁾ = r; index set

Set the termination iteration threshold ξ; t=1, Θ₀ is an empty matrix;

步骤3：将Γ中各模态字典Ψ⁽ⁿ⁾与第l列抽取组成字矩阵Γ^l，将每个Γ^l,l＝1,2…L与残差ε相乘，找出积为最大值所对应的索引为λ_t，即

Step 3: Extract each modal dictionary Ψ⁽ⁿ⁾ and the lth column in Γ to form a word matrix Γ^l , multiply each Γ^l , l=1, 2...L with the residual ε to find the maximum product The index corresponding to the value is λ_t , that is

步骤4：更新索引集Λ_t＝Λ_t-1∪{λ_t}，记录找到的多态级联字典中与残差相关度最高的列组合，并重建原子集合为γ_t＝[γ_t-1,Γ^l]；Step 4: Update the index set Λ_t =Λ_t-1 ∪{λ_t }, record the column combination with the highest degree of correlation with the residual in the found polymorphic cascade dictionary, and reconstruct the atomic set as γ_t =[γ_{t- 1} , Γ^l ];

步骤5：求解得到θ^(t)＝argmin||r-γ_tθ^(t)||₂，更新残差ε^(t)＝r-γ_tθ^(t)，t＝t+1；Step 5: Solve to obtain θ^(t) =argmin||r-γ_t θ^(t) ||₂ , update the residual ε^(t) =r-γ_t θ^(t) , t=t+1;

步骤6：判断残差ε^(t)是否小于ξ，如是，则迭代停止，输出重构特征向量θ；反之，则跳转至步骤3继续上述步骤。Step 6: Determine whether the residual ε^(t) is less than ξ, if so, stop the iteration and output the reconstructed feature vector θ; otherwise, jump to step 3 to continue the above steps.

步骤7：由重构特征向量θ恢复得到联合特征矩阵Θ。Step 7: Restore the joint feature matrix Θ from the reconstructed feature vector Θ.

另一实施例中，在得到联合特征矩阵Θ后，对Θ中数值进行分析，将Θ的前L列置零，随后确定剩余的非零列，获得非零列索引向量υ，由索引向量υ指向扰动事件类型。至此，本发明实现了包括远端扰动事件在内的多类型扰动事件的同步识别。In another embodiment, after the joint feature matrix Θ is obtained, the values in Θ are analyzed, the first L columns of Θ are set to zero, and then the remaining non-zero columns are determined to obtain a non-zero column index vector υ, which is determined by the index vector υ. Points to the disturbance event type. So far, the present invention realizes synchronous identification of multiple types of disturbance events including remote disturbance events.

上述实施例为本发明优选地实施方式，但本发明的实施方式并不受上述实施例的限制，其他的任何未背离本发明的精神实质与原理下所作的改变、修饰、替代、组合、简化，均应为等效的置换方式，都包含在本发明的保护范围之内。The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited by the above embodiments, and any other changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principle of the present invention , all should be equivalent replacement modes, and all are included in the protection scope of the present invention.

Claims (10)

Translated fromChinese

1.一种分布式光纤传感系统的远端扰动特征提取方法，其特征在于，包括以下步骤：1. a far-end disturbance feature extraction method of a distributed optical fiber sensing system, is characterized in that, comprises the following steps:S01：对光纤信道内信号瑞利散射谱的距离维区间进行划分，并针对各区间建立相应的字典学习模型，对原有的字典进行学习观测形成多态等效字典；S01: Divide the distance dimension interval of the Rayleigh scattering spectrum of the signal in the fiber channel, establish a corresponding dictionary learning model for each interval, and learn and observe the original dictionary to form a polymorphic equivalent dictionary;S02：基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示；S02: Based on the polymorphic equivalent dictionary, a joint sparse representation model under the cascade of polymorphic equivalent dictionaries is established to form a joint sparse representation of the multi-echo scattering spectrum;S03：构建基于多态级联字典的联合优化重构算法，提取远端扰动信号特征。S03: Construct a joint optimization and reconstruction algorithm based on a polymorphic cascade dictionary to extract features of remote disturbance signals.2.根据权利要求1所述的分布式光纤传感系统的远端扰动特征提取方法，其特征在于，所述步骤S01中形成多态等效字典的方法包括：2. The method for extracting remote disturbance features of a distributed optical fiber sensing system according to claim 1, wherein the method for forming a polymorphic equivalent dictionary in the step S01 comprises:S11：针对所有的探测距离构造一个本征特征字典：S11: Construct an intrinsic feature dictionary for all detection distances:Ψ＝[ψ₁,ψ₂,…,ψ_L]；Ψ=[ψ₁ ,ψ₂ ,...,ψ_L ];其中，ψ_l表示在第l个探测距离上的单位响应；Among them, ψ_l represents the unit response at the l-th detection distance;得到基态谱

表示为：get ground state spectrum

Expressed as:

其中，η为基态特征系数；Among them, η is the ground state characteristic coefficient;S12：针对第n个扰动事件ζ_n，其扰动特征空间为

其扰动特征互补空间为

基于

得到对应的特征系数向量

为：S12: For the nth disturbance event ζ_n , its disturbance feature space is

Its perturbation feature complementary space is

based on

Get the corresponding eigencoefficient vector

for:

其中，η_l与

分别表示η和

的第l个元素；where η_l and

denote η and

the lth element of ;S13：针对扰动事件ζ_n，构建扰动模态观测矩阵：S13: For the disturbance event ζ_n , construct the disturbance modal observation matrix:

其中，diag(z)表示以向量z为对角元素构成矩阵，β⁽ⁿ⁾为权系数向量；Among them, diag(z) represents a matrix with the vector z as the diagonal element, and β⁽ⁿ⁾ is the weight coefficient vector;S14：基于上述扰动模态观测矩阵，对本征特征字典进行学习观测，形成扰动事件ζ_n所对应的模态字典：S14: Based on the above disturbance modal observation matrix, learn and observe the eigenfeature dictionary to form a modal dictionary corresponding to the disturbance event_ζn :Ψ⁽ⁿ⁾＝Φ⁽ⁿ⁾Ψ；Ψ⁽ⁿ⁾ = Φ⁽ⁿ⁾ Ψ;将所有扰动事件N所对应的模态字典与本征特征字典进行级联组合，得到多态等效字典：The modal dictionaries corresponding to all disturbance events N are combined with the eigenfeature dictionaries in cascade to obtain a polymorphic equivalent dictionary:

3.根据权利要求1或2所述的分布式光纤传感系统的远端扰动特征提取方法，其特征在于，所述步骤S02中形成多回波散射谱的联合稀疏表示的方法包括：3. The method for extracting remote disturbance features of a distributed optical fiber sensing system according to claim 1 or 2, wherein the method for forming a joint sparse representation of multiple echo scattering spectra in the step S02 comprises:将M个回波光脉冲组成的联合回波矩阵表示为：The joint echo matrix composed of M echo light pulses is expressed as:

其中，P_i表示第i个回波光脉冲对应的采样瑞利曲线，[·]^T表示向量与矩阵的转置；Among them, P_i represents the sampled Rayleigh curve corresponding to the ith echo light pulse, [ ]^T represents the transpose of the vector and the matrix;基于多态等效字典，得到联合回波矩阵的联合稀疏表示模型为：Based on the polymorphic equivalent dictionary, the joint sparse representation model of the joint echo matrix is obtained as:

其中，Θ为联合特征矩阵，同时也是一个列稀疏矩阵，其非零列表征了相应多态字典所对应的扰动事件。Among them, Θ is the joint feature matrix, which is also a column sparse matrix, and its non-zero column represents the perturbation events corresponding to the corresponding polymorphic dictionary.4.根据权利要求3所述的分布式光纤传感系统的远端扰动特征提取方法，其特征在于，所述步骤S03中提取远端扰动信号特征的方法包括：4. The remote disturbance feature extraction method of the distributed optical fiber sensing system according to claim 3, wherein the method for extracting the remote disturbance signal feature in the step S03 comprises:S31：对联合回波矩阵进行向量化得到r＝vec(R)，进而得到扩展联合稀疏表示：S31: Vectorize the joint echo matrix to obtain r=vec(R), and then obtain the extended joint sparse representation:r＝Γθr=Γθ其中，

表示维度为(N+1)×(N+1)的单位阵，

表示克罗内克积，θ为重构特征向量；in,

represents a unit matrix of dimension (N+1)×(N+1),

represents the Kronecker product, and θ is the reconstructed eigenvector;S32：初始化重构特征向量θ⁽⁰⁾＝0，残差ε⁽⁰⁾＝r；索引集

设置终止迭代门限ξ；t＝1，Θ₀为空矩阵；S32: Initialize reconstruction feature vector θ⁽⁰⁾ = 0, residual ε⁽⁰⁾ = r; index set

Set the termination iteration threshold ξ; t=1, Θ₀ is an empty matrix;S33：将Γ中各模态字典Ψ⁽ⁿ⁾与第l列抽取组成字矩阵Γ^l，将每个Γ^l,l＝1,2…L与残差ε相乘，找出积为最大值所对应的索引为λ_t，即

S33: Extract each modal dictionary Ψ⁽ⁿ⁾ and the l-th column in Γ to form a word matrix Γ^l , multiply each Γ^l , l=1, 2...L with the residual ε, and find the product as the maximum value The corresponding index is λ_t , that is,

S34：更新索引集Λ_t＝Λ_t-1∪{λ_t}，记录找到的多态等效字典中与残差相关度最高的列组合，并重建原子集合为Υ_t＝[Υ_t-1,Γ^l]；S34: Update the index set Λ_t = Λ_t-1 ∪{λ_t }, record the column combination with the highest degree of correlation with the residual in the found polymorphic equivalent dictionary, and reconstruct the atomic set as Υ_t = [Υ_t-1 , Γ^l ];S35：求解得到θ^(t)＝argmin||r-Υ_tθ^(t)||₂，更新残差ε^(t)＝r-Υ_tθ^(t)，t＝t+1；S35: Solve to obtain θ^(t) = argmin||r-Υ_t θ^(t) ||₂ , update the residual ε^(t) =r-Υ_t θ^(t) , t=t+1;S36：判断残差ε^(t)是否小于ξ，如是，则迭代停止，输出重构特征向量θ；反之，则跳转至步骤S33继续执行；S36: determine whether the residual ε^(t) is less than ξ, if so, stop the iteration, and output the reconstructed feature vector θ; otherwise, jump to step S33 to continue execution;S37：由重构特征向量θ重构得到联合特征矩阵Θ。S37: Reconstruction from the reconstructed feature vector θ to obtain a joint feature matrix Θ.5.根据权利要求4所述的分布式光纤传感系统的远端扰动特征提取方法，其特征在于，所述步骤S03之后还包括，将重构得到的联合特征矩阵Θ的前L列置零，随后确定剩余的非零列，获得非零列索引向量υ，由索引向量υ指向扰动事件类型。5. the far-end disturbance feature extraction method of distributed optical fiber sensing system according to claim 4, is characterized in that, also comprises after described step S03, the first L column of the joint feature matrix Θ that reconstruction obtains is set to zero , then determine the remaining non-zero columns, and obtain a non-zero column index vector υ, which points to the disturbance event type by the index vector υ.6.一种分布式光纤传感系统的远端扰动特征提取系统，其特征在于，包括：6. A remote disturbance feature extraction system of a distributed optical fiber sensing system, characterized in that, comprising:多态等效字典构建模块，对光纤信道内信号瑞利散射谱的距离维区间进行划分，并针对各区间建立相应的字典学习模型，对原有的字典进行学习观测形成多态等效字典；The polymorphic equivalent dictionary building module divides the distance dimension interval of the Rayleigh scattering spectrum of the signal in the fiber channel, establishes a corresponding dictionary learning model for each interval, and learns and observes the original dictionary to form a polymorphic equivalent dictionary;多回波散射谱的联合稀疏表示模块，基于多态等效字典，建立在多态等效字典级联下的联合稀疏表示模型，形成多回波散射谱的联合稀疏表示；The joint sparse representation module of multiple echo scattering spectra, based on the polymorphic equivalent dictionary, establishes a joint sparse representation model under the cascade of polymorphic equivalent dictionaries to form a joint sparse representation of multiple echo scattering spectra;扰动信号特征提取模块，构建基于多态级联字典的联合优化重构算法，提取远端扰动信号特征。The disturbance signal feature extraction module constructs a joint optimization and reconstruction algorithm based on polymorphic cascade dictionary to extract the remote disturbance signal features.7.根据权利要求6所述的分布式光纤传感系统的远端扰动特征提取系统，其特征在于，所述形成多态等效字典的方法包括：7. The remote disturbance feature extraction system of the distributed optical fiber sensing system according to claim 6, wherein the method for forming a polymorphic equivalent dictionary comprises:S11：针对所有的探测距离构造一个本征特征字典：S11: Construct an intrinsic feature dictionary for all detection distances:Ψ＝[ψ₁,ψ₂,…,ψ_L]；Ψ=[ψ₁ ,ψ₂ ,...,ψ_L ];其中，ψ_l表示在第l个探测距离上的单位响应；Among them, ψ_l represents the unit response at the l-th detection distance;得到基态谱

表示为：get ground state spectrum

Expressed as:

其中，η为基态特征系数；Among them, η is the ground state characteristic coefficient;S12：针对第n个扰动事件ζ_n，其扰动特征空间为Ω_ζn，其扰动特征互补空间为

基于

得到对应的特征系数向量

为：S12: For the nth disturbance event ζ_n , its disturbance feature space is Ω_ζn , and its disturbance feature complementary space is

based on

Get the corresponding eigencoefficient vector

for:

其中，η_l与

分别表示η和

的第l个元素；where η_l and

denote η and

the lth element of ;S13：针对扰动事件ζ_n，构建扰动模态观测矩阵：S13: For the disturbance event ζ_n , construct the disturbance modal observation matrix:

其中，diag(z)表示以向量z为对角元素构成矩阵，β⁽ⁿ⁾为权系数向量；Among them, diag(z) represents a matrix with the vector z as the diagonal element, and β⁽ⁿ⁾ is the weight coefficient vector;S14：基于上述扰动模态观测矩阵，对本征特征字典进行学习观测，形成扰动事件ζ_n所对应的模态字典：S14: Based on the above disturbance modal observation matrix, learn and observe the eigenfeature dictionary to form a modal dictionary corresponding to the disturbance event_ζn :Ψ⁽ⁿ⁾＝Φ⁽ⁿ⁾Ψ；Ψ⁽ⁿ⁾ = Φ⁽ⁿ⁾ Ψ;将所有扰动事件N所对应的模态字典与本征特征字典进行级联组合，得到多态等效字典：The modal dictionaries corresponding to all disturbance events N are combined with the eigenfeature dictionaries in cascade to obtain a polymorphic equivalent dictionary:

8.根据权利要求6或7所述的分布式光纤传感系统的远端扰动特征提取系统，其特征在于，所述多回波散射谱的联合稀疏表示的方法包括：8. The remote disturbance feature extraction system of the distributed optical fiber sensing system according to claim 6 or 7, wherein the method for the joint sparse representation of the multiple echo scattering spectra comprises:将M个回波光脉冲组成的联合回波矩阵表示为：The joint echo matrix composed of M echo light pulses is expressed as:

其中，P_i表示第i个回波光脉冲对应的采样瑞利曲线，[·]^T表示向量与矩阵的转置；Among them, P_i represents the sampled Rayleigh curve corresponding to the ith echo light pulse, [ ]^T represents the transpose of the vector and the matrix;基于多态等效字典，得到联合回波矩阵的联合稀疏表示模型为：Based on the polymorphic equivalent dictionary, the joint sparse representation model of the joint echo matrix is obtained as:

其中，Θ为联合特征矩阵，同时也是一个列稀疏矩阵，其非零列表征了相应多态字典所对应的扰动事件。Among them, Θ is the joint feature matrix, which is also a column sparse matrix, and its non-zero column represents the perturbation events corresponding to the corresponding polymorphic dictionary.9.根据权利要求8所述的分布式光纤传感系统的远端扰动特征提取系统，其特征在于，所述提取远端扰动信号特征的方法包括：9. The remote disturbance feature extraction system of the distributed optical fiber sensing system according to claim 8, wherein the method for extracting the remote disturbance signal feature comprises:S31：对联合回波矩阵进行向量化得到r＝vec(R)，进而得到扩展联合稀疏表示：S31: Vectorize the joint echo matrix to obtain r=vec(R), and then obtain the extended joint sparse representation:r＝Γθr=Γθ其中，

表示维度为(N+1)×(N+1)的单位阵，

表示克罗内克积，θ为重构特征向量；in,

represents a unit matrix of dimension (N+1)×(N+1),

represents the Kronecker product, and θ is the reconstructed eigenvector;S32：初始化重构特征向量θ⁽⁰⁾＝0，残差ε⁽⁰⁾＝r；索引集

设置终止迭代门限ξ；t＝1，Θ₀为空矩阵；S32: Initialize the reconstructed feature vector θ⁽⁰⁾ = 0, residual ε⁽⁰⁾ = r; index set

Set the termination iteration threshold ξ; t=1, Θ₀ is an empty matrix;S33：将Γ中各模态字典Ψ⁽ⁿ⁾与第l列抽取组成字矩阵Γ^l，将每个Γ^l,l＝1,2...L与残差ε相乘，找出积为最大值所对应的索引为λ_t，即

S33: Extract each modal dictionary Ψ⁽ⁿ⁾ and the lth column in Γ to form a word matrix Γ^l , multiply each Γ^l , l=1, 2...L with the residual ε, and find the product as The index corresponding to the maximum value is λ_t , that is,

S34：更新索引集Λ_t＝Λ_t-1∪{λ_t}，记录找到的多态等效字典中与残差相关度最高的列组合，并重建原子集合为Υ_t＝[Υ_t-1,Γ^l]；S34: Update the index set Λ_t = Λ_t-1 ∪{λ_t }, record the column combination with the highest degree of correlation with the residual in the found polymorphic equivalent dictionary, and reconstruct the atomic set as Υ_t = [Υ_t-1 , Γ^l ];S35：求解得到θ^(t)＝argmin||r-Υ_tθ^(t)||₂，更新残差ε^(t)＝r-Υ_tθ^(t)，t＝t+1；S35: Solve to obtain θ^(t) = argmin||r-Υ_t θ^(t) ||₂ , update the residual ε^(t) =r-Υ_t θ^(t) , t=t+1;S36：判断残差ε^(t)是否小于ξ，如是，则迭代停止，输出重构特征向量θ；反之，则跳转至步骤S33继续执行；S36: determine whether the residual ε^(t) is less than ξ, if so, stop the iteration, and output the reconstructed feature vector θ; otherwise, jump to step S33 to continue execution;S37：由重构特征向量θ重构得到联合特征矩阵Θ。S37: Reconstruction from the reconstructed feature vector θ to obtain a joint feature matrix Θ.10.根据权利要求9所述的分布式光纤传感系统的远端扰动特征提取系统，其特征在于，还包括扰动事件类型识别模块，将重构得到的联合特征矩阵Θ的前L列置零，随后确定剩余的非零列，获得非零列索引向量υ，由索引向量υ指向扰动事件类型。10. the far-end disturbance feature extraction system of distributed optical fiber sensing system according to claim 9, is characterized in that, also comprises disturbance event type identification module, and the first L column of the joint feature matrix Θ that reconstruction obtains is set to zero , then determine the remaining non-zero columns, and obtain a non-zero column index vector υ, which points to the disturbance event type by the index vector υ.

Images