Determination device, determination method, and program

Abstract

The objective of the present invention is to provide a determining device, a determining method, and a program capable of determining a classification of various environments in which an optical cable is laid, from a vibration distribution waveform.　A determining device 301 for classifying a laid environment of an optical cable 21 comprises: a feature extracting unit 212 for extracting a feature vector at each position on the optical cable 21, from a vibration distribution in a longitudinal direction of the optical cable 21; a calculating unit 219 for calculating an identification function at each position on the optical cable 21 from the feature vector, and a weighting vector corresponding to the laid environment of the optical cable 21 to be classified; and a determining unit 220 for comparing the identification function and a teaching signal representing two states, at each position on the optical cable 21, and determining, as the state of the optical cable 21, the state of the teaching signal that is closer.

Description

Technical Field

[0001] The present disclosure relates to a determination device for classifying an environment where an optical cable is laid, a determination method thereof, and a determination program thereof.

Background Art

[0002] In the maintenance and management of an optical fiber communication network, accurate and up-to-date information management of optical facilities constituting the communication network is desired. In particular, since the skills required of workers differ depending on the laying environment of the optical cable, it is important to be able to determine whether it is underground or overhead. Identifying the positions of utility poles and cable slack portions is also useful for reducing the workload of workers.

[0003] For remote monitoring and testing of optical cables, there is distance loss measurement by the OTDR (Optical Time Domain Reflectometry) method, which is an optical evaluation method (see, for example, Patent Document 1). In the OTDR method, an optical tester is connected to one core of the optical fiber in the optical cable, pulsed light is incident on the optical fiber, and the light intensity of scattered light (backscattered light) propagating in the opposite direction to the pulsed light is detected in the longitudinal direction of the optical fiber to measure the distance loss of the optical fiber. Although the OTDR method is useful for identifying the failure location of an optical cable from the distance loss measurement, it cannot determine the laying environment of the optical cable.

[0004] In recent years, the DAS (Distributed Acoustic Sensing) method has emerged for measuring the vibration distribution from the change in the backscattered light waveform in the continuous longitudinal direction of an optical fiber due to the narrowing of the line width of a laser (see, for example, Non-Patent Document 1). By measuring the vibration distribution and using the optical fiber as a sensor, it is possible to detect vibrations added to the optical cable from the surrounding environment, which is useful material for determining the laying environment.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Non-Patent Literature

[0006]

Non-Patent Literature 1

Non-Patent Literature 2

Summary of the Invention

Problems to be Solved by the Invention

[0007] However, the vibration distribution measurement results obtained by the DAS method are the changes in the magnitude of vibration in the continuous time domain in the longitudinal direction of the optical cable. Therefore, even though it is possible to know that the vibrations are different in different local ranges of the optical cable, there is a problem that it is difficult to directly determine the factor that added the vibration only from the measurement results.

[0008] That is, the problems to be solved by the present invention are the following two. (1) Interpreting the meaning of the change in the magnitude of vibration in the longitudinal direction of the optical cable and classifying and determining the environment where the optical cable is laid. (2) Establishing a classification and determination algorithm applicable to various laying environments.

[0009] In order to solve the above problems, an object of the present invention is to provide a determination device, a determination method, and a program that can classify and determine various environments where an optical cable is laid from a vibration distribution waveform.

Means for Solving the Problems

[0010] To achieve the above object, the determination device according to the present invention performs a Fourier transform on the vibration distribution in the longitudinal direction of the optical cable, multiplies the frequency of the peak of the spectrum and its amplitude by a weighting coefficient to generate a discrimination function, and compares this with teacher data representing two environments (underground / aerial, presence / absence of utility poles, presence / absence of cable slack) where the optical cable is laid, and determines the state of the closer teacher data as the environment where the optical cable is laid.

[0011] Specifically, the determination device according to the present invention is a determination device that classifies the laying environment of an optical cable, a feature extraction unit that extracts a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable, a calculation unit that calculates a discrimination function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified, For each position of the optical cable, compare the discriminant function with a teacher signal representing two states, and a determination unit that determines the state of the optical cable to be the state of the closer teacher signal. It is characterized by comprising.

[0012] Further, the determination method according to the present invention is a determination method for classifying the laying environment of an optical cable, extracting a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable, calculating a discriminant function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified, and for each position of the optical cable, comparing the discriminant function with a teacher signal representing two states, and determining the state of the optical cable to be the state of the closer teacher signal It is characterized by performing.

[0013] This determination device (method) utilizes the fact that the optical cable has vibrations according to the environment. Specifically, when determining whether the optical cable is buried underground or laid overhead, a weight vector corresponding to the environmental determination of underground or overhead is learned in advance, and a discriminant function is formed from this and the feature vector of the vibration of the optical fiber in an unknown environment. Then, it is determined whether the environment is underground or overhead from the value of the discriminant function. Further, if the weight vector to be learned in advance is for other environments (for example, presence / absence of utility poles, presence / absence of cable slack, etc.), it can be applied to classification determination of various environments.

[0014] Therefore, the present invention can provide a determination device and a determination method capable of classifying and determining various environments in which an optical cable is laid from the vibration distribution waveform.

[0015] For example, the feature extraction unit performs a Fourier transform on the vibration distribution from a waveform in the time domain to a spectrum waveform in the frequency domain, extracts the frequency of each peak and the amplitude of the peak from the spectrum waveform, and arranges the frequency and the amplitude in the order of the peaks with large amplitudes to form the feature vector.

[0016] For example, the weight vector is composed of coefficients that multiply the frequency and the amplitude of each of the peaks, and the arithmetic unit can use the coefficients of the weight vector to multiply the frequency and the amplitude of each of the peaks and then add the resulting values to obtain the discriminant function.

[0017] The feature extraction unit performs a Fourier transform on the vibration distribution from the waveform in the time domain to the spectral waveform in the frequency domain, removes the fluctuation component from the components having a period in the frequency axis direction of the spectral waveform, extracts the frequency of the peak and the amplitude of the peak from the spectral waveform after removing the fluctuation component, and may construct the feature vector using the frequency of the extracted peak and the amplitude of the peak.

[0018] The feature extraction unit takes the logarithm of the absolute value of the amplitude of the spectral waveform, calculates the time waveform of the component having a period in the frequency axis direction of the logarithmically transformed logarithmic spectral waveform, extracts the time region where the value is smaller than a predetermined threshold time in the time waveform, and may remove the fluctuation component.

[0019] By removing the fluctuation component, it is possible to extract the essential feature quantities for identifying the optical cable laying environment, thereby improving the identification accuracy. Furthermore, by extracting the essentially necessary feature quantities, the dimensionality of the feature vector is reduced, so that the computing resources can be made more efficient.

[0020] Note that the determination device prepares a weight vector as follows. The determination device further includes an identification dictionary having a dictionary arithmetic unit and an update unit, and the feature extraction unit extracts known feature vectors for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable for each of the two states. The dictionary calculation unit calculates a known discrimination function for each position of the optical cable from the known feature vector and the weight vector, The update unit calculates, for each position of the optical cable, an error between the known feature vector for each of the two states and the teacher signal representing the corresponding state of the two states, and updates the weight vector so that the error becomes small. In addition, the discrimination dictionary determines the weight vector when the squared value of the error is minimized and the variation in the error before and after the update of the weight vector is equal to or less than a threshold value.

[0021] This determination method prepares a weight vector as follows. Extracting a known feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable for each of the two states, Calculating a known discrimination function for each position of the optical cable from the known feature vector and the weight vector, Calculating, for each position of the optical cable, an error between the known feature vector for each of the two states and the teacher signal representing the corresponding state of the two states, and Updating the weight vector so that the error becomes small are further performed.

[0022] The present invention is a program for causing a computer to function as the determination device. The determination device of the present invention can also be realized by a computer and a program, and the program can be recorded on a recording medium or provided through a network.

[0023] In addition, the above inventions can be combined as much as possible.

Effects of the Invention

[0024] The present invention can provide a determination device, a determination method, and a program that can classify and determine various environments in which an optical cable is laid from a vibration distribution waveform.

Brief Description of the Drawings

[0025]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Embodiments for Carrying Out the Invention

[0026] Embodiments of the present invention will be described with reference to the accompanying drawings. The embodiments described below are examples of the present invention, and the present invention is not limited to the following embodiments. In the present specification and drawings, components having the same reference numerals are assumed to be the same as each other.

[0027] (Summary of the Invention) FIG. 1 is a diagram for explaining the calculation method performed by the determination device of the present embodiment. In FIG. 1, each reference numeral refers to the following: 11: peak frequency of the Fourier spectrum, 12: peak amplitude of the Fourier spectrum, 13: linear sum, 14: output value y, 15: weight vector.

[0028] This calculation method outputs one value for a plurality of input signals. The class of the input signal is classified according to the output result. In the present invention, as a feature amount (feature vector) serving as an input signal, the peak frequency 11 (x d1 , xd2 , ···, x dn ) and peak amplitude 12(x a1 , x a2 , ···, x an ), two of which are defined, and the output value y14 of the linear sum 13 obtained by multiplying each of these variables by a coefficient (weight vector 15(ω d1 , ω d2 , ···, ω dn , ω a1 , ω a2 , ···, ω an )) is used as the discriminant function. The classification of the environment where the optical cable is laid (underground / aerial), (presence / absence of utility poles), or (presence / absence of cable slack) is determined based on the value of the discriminant function.

[0029] (Embodiment 1) Hereinafter, the method for extracting feature quantities (feature vectors), the method for calculating the linear sum, and the classification determination method will be described in detail.

[0030] FIG. 2 is a diagram for explaining the determination device 301 of the present embodiment. The determination device 301 is a determination device for classifying the laying environment of the optical cable 21, a feature extraction unit 212 that extracts a feature vector for each position of the optical cable 21 from the vibration distribution in the longitudinal direction of the optical cable 21, an arithmetic unit 219 that calculates a discriminant function for each position of the optical cable 21 from the feature vector and a weight vector corresponding to the laying environment of the optical cable 21 to be classified, a determination unit 220 that compares the discriminant function with a teacher signal representing two states for each position of the optical cable 21 and determines the state of the optical cable 21 as the state of the closer teacher signal, and includes.

[0031] In FIG. 2, each reference numeral refers to the following. 21: Optical cable, 22: Ground surface, 23: Underground, 24: Aerial, 25: Utility pole, 26: Slack, 27: Vibration distribution measuring device, 28: First vibration distribution data, 29: Second vibration distribution data, 210: Third vibration distribution data, 211: Memory unit, 212: Feature extraction unit, 213: Data reading unit, 214: Fourier transform unit, 215: Peak frequency extraction unit, 216: Peak amplitude extraction unit, 217: Identification operation unit, 218: Identification dictionary, 219: Identification function operation unit, 220: Classification determination unit, 221: Result display unit, 222: Identification function operation unit, 223: Error operation unit, 224: Teacher signal, 225: Mean squared error operation unit, 226: Weight vector update unit.

[0032] The optical cable 21 is divided into a laying environment of underground 23 and aerial 24 with the ground surface 22 as a boundary, and is laid by utility poles 25 in the case of aerial. Also, in an abnormal state of laying, slack 26 may locally occur, and repair work may be required. It is an object of the present invention to classify and determine the above-mentioned laying environment of the optical cable by an optical test. The optical test installs a vibration distribution measuring device 27 at one end of the optical cable 21, and uses a DAS (Distributed Acoustic Sensing) method (for example, refer to Non-Patent Document 1 and Non-Patent Document 2.) using C-OFDR (Coherent Optical Frequency Domain Reflectometry) or C-OTDR (Coherent Optical Time Domain Reflectometry) described later with reference to FIG. 3 to detect the distribution of vibrations in the longitudinal direction of the optical cable. By continuously detecting the vibration distribution, three-dimensional data of the optical cable distance-time-vibration magnitude can be obtained. Hereinafter, the three-dimensional data will be referred to as vibration distribution data. The vibration distribution data includes vibration components in various laying environments such as underground, aerial, utility pole positions, and cable slack.

[0033] In the present embodiment, an example of classifying and determining whether the laying environment of the cable 21 is underground or aerial will be described. The classification determination of the presence or absence of the utility pole position and the presence or absence of cable slack may be replaced with the classification determination of underground or aerial in the following description.

[0034] For classifying and determining the laying environment, the vibration distribution measuring device 27 acquires three vibration distribution data. The first vibration distribution data 28 is the data for which the laying environment is to be grasped, and is the vibration distribution data in which the underground and overhead laying environments are mixed. The second vibration distribution data 29 is the vibration distribution data measured in advance only in the underground laying environment. The third vibration distribution data 210 is, similarly to the second vibration distribution data 29, the vibration distribution data measured in advance only in the overhead laying environment. Since the second vibration distribution data 29 and the third vibration distribution data 210 are used as learning data, it is preferable that the environment is different from the laying environment of the first vibration distribution data 28. The three vibration distribution data are stored in the storage unit 211.

[0035] The vibration distribution data 28, 29, 210 are sent to the feature extraction unit 212. The data reading unit 213 reads the vibration distribution data 28, 29, 210 independently and passes the data to the subsequent process. The vibration distribution data 28, 29, 210 passed from the data reading unit 213 are converted by the Fourier transform unit 214 from the magnitude of the vibration continuously observed at each point in the longitudinal direction of the optical cable in the time domain to the frequency domain. Therefore, the vibration distribution data 28, 29, 210 are converted from the three-dimensional data of the optical cable distance - time - vibration magnitude to the three-dimensional data of the optical cable distance - frequency - amplitude (Fourier spectrum data in the longitudinal direction of the optical cable) by the Fourier transform unit 214.

[0036] The Fourier spectrum data is passed to the peak frequency extraction unit 215 and the peak amplitude extraction unit 216 respectively. In the peak frequency extraction unit 215, the peak frequency 11(x d1 , x d2 , ···, x dn ) of the Fourier spectrum is extracted from the Fourier spectrum data at each point in the longitudinal direction of the optical cable. Also, in the peak amplitude extraction unit 216, the peak amplitude 12(x a1 , x a2 , ···, x an ) of the Fourier spectrum is extracted. The extracted peak frequency 11(x d1 , x d2 , ···, x dn) and peak amplitude 12(x a1 , x a2 , ···, x an ) are passed as feature quantities (feature vectors) to subsequent processes. The feature quantity (feature vector) of the first vibration distribution data 28 is passed to the discrimination operation unit 217. The feature quantities (known feature vectors) of the second vibration distribution data 29 and the third vibration distribution data 210 are passed to the discrimination dictionary 218. The Fourier transform performed by the feature extraction unit 212 and the definition of the feature vector will be described later with reference to FIG. 4.

[0037] The feature quantity (feature vector) of the first vibration distribution data 28 passed to the discrimination operation unit 217 is passed to the discrimination function operation unit 219 as an input signal, and one output value y14 is calculated as shown in FIG. 1. The output value y14 becomes the discrimination function. The calculation method of the discrimination function operation unit 219 will be described. The feature quantity (feature vector) of the first vibration distribution data 28 (x d1 , x d2 , ···, x dn and x a1 , x a2 , ···, x an ) becomes 2n + 1 input signals (x d1 , x d2 , ···, x dn , x a1 , x a2 , ···, x an , 1). Also, 2n + 1 coefficients (weight vectors) (ω d1 , ω d2 , ···, ω dn , ω a1 , ω a2 , ···, ω an , ω 0 ) to be multiplied are received from the discrimination dictionary 218, and the discrimination function operation unit 219 calculates the following linear sum.

Equation

[0038] If the longitudinal distance of the optical cable 21 is composed of L points, the discrimination function calculation unit 219 calculates L (y 0 , y 1 , ···, y L-1 ) from the feature amounts (feature vectors) of all the passed first vibration distribution data 28. The calculated discrimination function (y 0 , y 1 , ···, y L-1 ) is passed to the classification determination unit 220.

[0039] The classification determination unit 220 classifies into the category of the teacher signal that is closest by comparing with the teacher signal set in advance for each category for each discrimination function (y 0 , y 1 , ···, y L-1 ). For example, the teacher signal “-1” is set for the category of the underground laying environment, the teacher signal “1” is set for the overhead category, and 0 is set as the threshold value. In the case of the discrimination function (y 0 (negative), y 1 (positive), ···, y L-1 (positive)), if a negative value is determined as underground and a positive value is determined as overhead, it is classified as (underground, overhead, ···, overhead). The classification determination performed by the classification determination unit 220 will be described later with reference to FIG. 5.

[0040] The classified result is passed to the result display unit 221, and the laying environment in the longitudinal direction of the optical cable is displayed as underground or overhead.

[0041] The discrimination accuracy in the classification determination varies depending on the weight vectors (ω d1 , ω d2 , ···, ω dn , ω a1 , ω a2 , ···, ω an , ω 0 ) received from the discrimination dictionary 218. The method for updating the weight vectors will be described below.

[0042] The determination device 301 further includes a discrimination dictionary 218 having a dictionary calculation unit (222, 223, 225) and an update unit (226). The feature extraction unit 212 extracts known feature vectors (29, 210) for each position of the optical cable from the longitudinal vibration distribution of the optical cable for each of two states (for example, underground and overhead). The dictionary operation unit 222 calculates a known discrimination function for each position of the optical cable from the known feature vector and the weight vector. The update unit 226 calculates, for each position of the optical cable, the error between the known feature vector for each of the two states and the teacher signal 224 representing the corresponding state among the two states, and updates the weight vector so that the error is reduced.

[0043] The feature amounts (known feature vectors) of the second vibration distribution data 29 and the third vibration distribution data 210 passed to the identification dictionary 218 are passed to the discrimination function calculation unit 222 as input signals. At this time, similar to the discrimination operation unit 217, the feature amounts (known feature vectors) of the second or third vibration distribution data (x d1 , x d2 , ···, x dn and x a1 , x a2 , ···, x an ) become 2n + 1 input signals (x d1 , x d2 , ···, x dn , x a1 , x a2 , ···, x an , 1). The discrimination function calculation unit 222 calculates a known discrimination function in the same manner as Equation 1 based on a weight vector (ω d1 , ω d2 , ···, ω dn , ω a1 , ω a2 , ···, ω an , ω 0 ) with 2n + 1 elements appropriately set in advance as an initial value. Here, a known discrimination function y for one point of the distance in the longitudinal direction of the optical cable composed of L points is calculated. The calculated known discrimination function y is passed to the error calculation unit 223, and the error (y - b i ) from the teacher signal 224 set in advance for each category is calculated. However, b iis the teacher signal (for example, the category of underground is "-1" and the category of overhead is "1").

[0044] Also, the error is passed to the mean squared error calculation unit 225, and the following mean squared error is calculated. (y - b i ) 2

[0045] The calculated mean squared error is passed to the weight vector update unit 226. On the other hand, the error is passed to the weight vector update unit 226, and the weight vector is updated by the following formula.

Equation

[0046] The updated weight vector (ω’ d1 , ω’ d2 , ···, ω’ dn , ω’ a1 , ω’ a2 , ···, ω’ an , ω’ 0 ) is passed to the discriminant function calculation unit 222, and the discriminant function is calculated as shown in Equation 1 using the weight vector (ω d1 , ω d2 , ···, ω dn , ω a1 , ω a2 , ···, ω an , ω 0 )

[0047] The discrimination dictionary 218 repeatedly calculates Equations 1 to 4. If the feature amounts (known feature vectors) of the second vibration distribution data 29 and the third vibration distribution data 210 are each composed of L points in the longitudinal direction of the optical cable, it is repeated more than 2L times. The repetition ends when the mean squared error calculated by the mean squared error calculation unit 225 reaches the minimum value and stabilizes. At this time, the calculation of the mean squared error may use the following. Σ((y 0 , y 1 , ···, y L-1 ) - b i ) 2

[0048] The weight vectors (ω d1 , ω d2 , ···, ω dn , ω a1 , ω a2 , ···, ω an , ω 0 ) are used by the discriminant function calculation unit 219.

[0049] [Supplementary Explanation 1] FIG. 3 is a diagram for explaining the detection of the vibration distribution by the optical test (C-OFDR) performed by the vibration distribution measuring device 27. In FIG. 3, each reference numeral denotes the following. 31: optical intensity distribution, 32: local section, 33: waveform pattern, 34: waveform pattern after Δt seconds, 35: change Δν. FIG. 3(A) is the optical intensity distribution of the entire longitudinal direction of the optical fiber, and FIG. 3(B) is a waveform in which the optical intensity of the local section 32 is enlarged and displayed.

[0050] In the conventional C-OFDR, the optical intensity distribution 31 in the longitudinal direction of the optical fiber is measured. By injecting laser light into the optical fiber in the optical cable and receiving the backward Rayleigh scattered light propagating in the direction opposite to the incident direction, the change in the optical intensity is observed. Focusing on the optical intensity waveform of the local section 32, a waveform corresponding to the characteristics unique to the optical fiber can be observed, and when the states of the optical fiber, laser light, etc. are unchanged, the same waveform pattern 33 is shown. The central wavelength of the laser light with a wide linewidth always changes, and the waveform pattern also changes.

[0051] When a narrow linewidth laser appears due to the progress of laser technology, the influence of the change in the central wavelength becomes a non-problem, and when the optical fiber state is the same, the waveform is also the same pattern. Here, when vibrations unique to the optical fiber are added, the waveform pattern 34 after Δt seconds is different due to the influence of the vibrations. The vibration is detected from the change Δν35 between the waveform pattern 33 and the waveform pattern 34.

[0052] By moving the local section 32, the vibration distribution in the longitudinal direction of the optical fiber is detected. Also, by continuously measuring the light intensity distribution and detecting vibrations each time, the temporal change of the vibration in the longitudinal direction of the optical fiber can be observed. In this way, three-dimensional vibration distribution data of the optical cable distance-time-vibration magnitude is obtained. In the case of C-OTDR, attention is paid to the change in the phase information obtained from the waveform pattern, and the vibration distribution is detected.

[0053] [Supplementary Explanation 2] FIG. 4 is a diagram for explaining the Fourier transform and the definition of the feature vector performed by the feature extraction unit 212. In FIG. 4, each reference numeral refers to the following. 41: waveform g(t), 42: Fourier spectrum. FIG. 4(A) is a diagram for explaining the waveform 41g(t), and FIG. 4(B) is a diagram for explaining the Fourier spectrum obtained by Fourier-transforming the waveform 41g(t).

[0054] The waveform 41g(t) showing the change in the continuous vibration in the local section 32 indicates the magnitude of the vibration in the time direction at the plot interval Δt. An example of the conversion formula from the time domain to the frequency domain is shown below. [Equation] However, F(ω) is the waveform obtained by converting the waveform g(t) into the frequency domain, and f is the frequency [Hz].

[0055] The waveform g(t) is converted into the frequency domain and becomes the Fourier spectrum 42. In the Fourier spectrum 42, peaks are observed at specific frequencies according to the vibration components. The frequencies of the spectral peaks are assigned as x d1 , x d2 , ···, x dn , and the amplitudes are assigned as x a1 , x a2 , ···, x an . These processes are performed for all of the longitudinal direction of the optical fiber by moving the local section 32. The extracted peak frequencies (x d1 , x d2 , ···, x dn ) and peak amplitudes (x a1 , xa2 , ···, x an is passed to the discrimination operation unit 217 or the discrimination dictionary 218 as a feature amount (feature vector or known feature vector).

[0056] [Supplementary Explanation 3] FIG. 5 is a diagram for explaining the classification determination performed by the classification determination unit 220. Here, an example is described in which a teacher signal “-1” is set for the category of underground laying environment and a teacher signal “1” is set for the category of overhead laying environment. Here, 51: positive discrimination function, 52: negative discrimination function, and 53: threshold value.

[0057] When the discrimination functions calculated by the discrimination function operation unit 219 are arranged for each distance in the longitudinal direction of the optical cable, a graph in which the positive discrimination function 51 and the negative discrimination function 52 are mixed is obtained. Using the intermediate value “0” of the teacher signal as the threshold value 53, the negative discrimination function 52 is classified and determined as “underground” and the positive discrimination function 51 is classified and determined as “overhead”. The closer the value is to the teacher signal of “-1” or “1”, the closer it is to the vibration distribution data of underground or overhead learned in the discrimination dictionary. By using the algorithm of the present invention, it is possible to classify the optical cable laying environment from the vibration distribution waveform into two categories.

[0058] (Embodiment 2) The frequency components may include signal components that are not essential as features for identifying the surrounding environment, which is considered a factor in reducing accuracy. In addition, in order to obtain high discrimination accuracy, a high-dimensional feature vector is required, so the computational resources required for learning are large. Therefore, in the present embodiment, the following problems are solved. First, it is to extract essential feature amounts for identifying the optical cable laying environment and improve the discrimination accuracy. Second, it is to reduce the dimensionality of the feature vector by selecting and constituting the feature vector by selecting the essentially necessary feature amounts.

[0059] In the present embodiment, the object is to select and extract an essentially necessary feature vector from the vibration distribution waveform and provide an identification technique for the optical cable laying environment using a machine learning method.

[0060] Examples of embodiments related to the present disclosure will be described below. FIG. 7 is a diagram for explaining the selection and extraction of feature amounts from the envelope of the Fourier spectrum, which is the key point of the present embodiment.

[0061] In Embodiment 1, the peak frequency and the peak amplitude are obtained from a Fourier spectrum 61 obtained by converting the change in vibration on the time axis added in the longitudinal direction of the optical fiber into the frequency axis by Fourier transform, and these are used as feature amounts. When components having a frequency higher than the target frequency component are superimposed on the Fourier spectrum, the spectrum fluctuates and is displayed as a waveform. Therefore, many peaks are detected by the feature extraction unit 212 due to the fluctuation.

[0062] For comparison with the effects of the present embodiment, an explanation will be given using a logarithmic Fourier spectrum waveform 62 obtained by taking the logarithm of the absolute value of the amplitude of the Fourier spectrum 61. By taking the logarithm, it becomes possible to perform arithmetic processing on the fluctuation as a wave having a period in the frequency axis direction (in order to retain the periodicity information, the absolute value of the amplitude is calculated before taking the logarithm).

[0063] Here, when a threshold value 63 is determined and peak detection is performed on the waveform having a value equal to or higher than the threshold value, the peak frequency and the absolute value of the logarithmically transformed peak amplitude (log|amplitude|) become the feature vectors of Embodiment 1. As an example, the number of peaks exceeding the threshold value in the logarithmic Fourier spectrum waveform 62 is 10. Therefore, the dimensionality of the feature vector of Embodiment 1 is 20 in total, with 10 each for the peak frequency and the absolute value of the logarithmically transformed peak amplitude.

[0064] In order to remove the fluctuation component in the Fourier spectrum 61, when the logarithmic Fourier spectrum waveform 62 is set as g(f), the following formula is calculated.

Equation

[0065] The value (F(s))64 of the Fourier transform of the logarithmic Fourier spectrum obtained by the operation of the formula is a time waveform in which components having a period in the frequency axis direction of the logarithmic Fourier spectrum waveform 62 are plotted on the time axis. This time, the Fourier transform formula used FFT, but DCT (Discrete Cosine Transform) may be used to calculate Σ f g(f)cos{2πsf}.

[0066] Next, a waveform 65 in a time region having a value smaller than a predetermined threshold time is extracted from F(s)64. This threshold time can adopt any value that can remove fluctuations. At this time, other than the extracted waveform 66 may be replaced with the average value of the waveform 65 in the time region having a small value. Subsequently, the waveform 65 in the time region having a small value (or the concatenated waveform of the waveform obtained by the above replacement and the waveform 65 in the time region having a small value) is set as F’(s), and the following formula is calculated.

Equation

[0067] By the operation of this formula, the spectrum envelope 67 of the logarithmic Fourier spectrum waveform from which fluctuations have been removed can be obtained by re-converting from the time domain to the frequency domain. Thereby, an essential feature vector for identifying the optical cable laying environment can be extracted. At this time, the inverse Fourier transform (IFFT) may be used for this formula.

[0068] The ranges of the magnitudes (log|amplitude|) of the logarithmic Fourier spectrum waveform 62 and the spectrum envelope 67 are made the same, and peak detection is performed on waveforms above the aforementioned threshold 63. The peak frequency and the absolute value of the logarithmically converted peak amplitude (log|amplitude|) are used as a new feature vector 68. The number of peaks exceeding the threshold in the spectrum envelope 67 is 5. Therefore, the dimensionality of the feature vector is 10 in total, which is 5 for each of the peak frequency and the absolute value of the logarithmically converted peak amplitude. Therefore, by this process, it is possible to reduce the dimensionality of the feature vector from 20 to 10 without losing essential features. This process is performed at each point in the longitudinal direction of the optical fiber.

[0069] Classification determination is performed using an identification dictionary that has been learned based on the feature vectors selected and extracted from the spectral envelope 67. Examples of the learning method and the classification determination method include the learning rules and the discrimination function method described in Embodiment 1. Hereinafter, the learning method and the classification determination method using the feature vectors will be described in detail.

[0070] FIG. 8 is a diagram for explaining the determination device 301 of the present embodiment. In FIG. 8, each reference numeral indicates the following: 21: optical cable, 22: ground, 23: underground, 24: overhead, 25: utility pole, 27: vibration distribution measuring device, 28: first vibration distribution data, 29: second vibration distribution data, 210: third vibration distribution data, 211: storage unit, 212: feature extraction unit, 213: data reading unit, 214: Fourier transform unit, 212E: spectral envelope extraction unit, 215: peak frequency extraction unit, 216: peak amplitude extraction unit, 217: discrimination operation unit, 218: identification dictionary, 219: discrimination function operation unit, 220: classification determination unit, 221: result display unit, 222: discrimination function operation unit, 223: error operation unit, 224: teacher signal, 225: mean squared error operation unit, 226: weight vector update unit.

[0071] In the present embodiment, the feature extraction unit 212 includes the spectral envelope extraction unit 212E, and the three-dimensional data of the optical cable distance-frequency-amplitude (Fourier spectrum data in the longitudinal direction of the optical cable) converted by the Fourier transform unit 214 is passed to the spectral envelope extraction unit 212E. In the spectral envelope extraction unit 212E, as described above, the amplitude (magnitude) of the Fourier spectrum at each point in the longitudinal direction of the optical cable is logarithmized, converted to the time domain by Fourier transform, then the components in the time domain with small values are extracted, and the spectral envelope of the logarithmic Fourier spectrum is extracted by re-converting to the frequency domain.

[0072] The spectral envelope is passed to the peak frequency extraction unit 215 and the peak amplitude extraction unit 216, respectively. In the peak frequency extraction unit 215, the peak frequency (x d1 , x d2 , ···, x dn)Extract 11. Also, in the peak amplitude extraction unit 216, the peak amplitude (x a1 , x a2 , ···, x an )12 of the spectral envelope is extracted. The extracted peak frequencies (x d1 , x d2 , ···, x dn )11 and the peak amplitudes (x a1 , x a2 , ···, x an )12 are passed as feature quantities (feature vectors) to subsequent processes.

[0073] The feature quantity (feature vector) of the first vibration distribution data 28 is passed to the discrimination operation unit 217. The feature quantities (feature vectors) of the second vibration distribution data 29 and the third vibration distribution data 210 are passed to the discrimination dictionary 218.

[0074] The feature quantity (feature vector) of the first vibration distribution data 28 passed to the discrimination operation unit 217 is passed as an input signal to the discrimination function operation unit 219, and one output value y14 is calculated. This output value becomes the discrimination function. The operations of the discrimination operation unit 217 and the discrimination dictionary 218 are the same as those in Embodiment 1.

[0075] (Effect of this embodiment) The optical communication cable laying environment determination device, determination method, and program according to the present disclosure are considered to have the following advantages. First, by detecting the peak frequency from the spectral envelope 67 of the logarithmic Fourier spectrum, it is possible to extract essential feature quantities for identifying the optical cable laying environment, and thereby an improvement in identification accuracy is expected. Second, by extracting essentially necessary feature quantities, the dimensionality of the feature vector is reduced, so that the calculation resources can be made more efficient.

[0076] (Embodiment 3) The determination device 301 can also be realized by a computer and a program, and it is also possible to record the program on a recording medium or provide it through a network. Figure 6 shows a block diagram of system 100. System 100 includes a computer 105 connected to a network 135.

[0077] Network 135 is a data communication network. Network 135 can be a private network or a public network, and can include any or all of (a) a personal area network covering, for example, a room, (b) a local area network covering, for example, a building, (c) a campus area network covering, for example, a campus, (d) a metropolitan area network covering, for example, a city, (e) a wide area network covering, for example, an area spanning city, regional, or national boundaries, or (f) the Internet. Communication is performed via network 135 by electronic signals and optical signals.

[0078] Computer 105 includes a processor 110 and a memory 115 connected to processor 110. Although computer 105 is represented herein as a stand-alone device, it is not so limited and may rather be connected to other devices not shown in a distributed processing system.

[0079] Processor 110 is an electronic device composed of logic circuits that respond to and execute instructions.

[0080] Memory 115 is a tangible computer-readable storage medium encoded with a computer program. In this regard, memory 115 stores data and instructions, i.e., program code, that are readable and executable by processor 110 to control the operation of processor 110. Memory 115 can be implemented with a random access memory (RAM), a hard drive, a read-only memory (ROM), or a combination thereof. One of the components of memory 115 is program module 120.

[0081] Program module 120 includes instructions for controlling processor 110 to execute the processes described herein. Although operations are described herein as being performed by computer 105 or a method or process or a sub-process thereof, those operations are in fact performed by processor 110.

[0082] The term "module" is used herein to refer to a functional operation that can be embodied as either a stand-alone component or an integrated configuration consisting of a plurality of sub-components. Thus, program module 120 can be implemented as a single module or as a plurality of modules operating in cooperation with each other. Further, although program module 120 is described herein as being installed in memory 115 and thus implemented in software, it can be implemented in any of hardware (e.g., electronic circuitry), firmware, software, or combinations thereof.

[0083] Program module 120 is shown as already loaded into memory 115, but it may be configured to be located on storage device 140 for later loading into memory 115. Storage device 140 is a tangible computer-readable storage medium that stores program module 120. Examples of storage device 140 include compact disks, magnetic tapes, read-only memories, optical storage media, hard drives or a memory unit composed of a plurality of parallel hard drives, and universal serial bus (USB) flash drives. Alternatively, storage device 140 can be a random access memory or another type of electronic storage device located in a remote storage system (not shown) and connected to computer 105 via network 135.

[0084] System 100 further includes data sources 150A and 150B, which are collectively referred to herein as data source 150 and are communicatively connected to network 135. In fact, data source 150 can include any number of data sources, i.e., one or more data sources. Data source 150 includes unstructured data and can include social media.

[0085] System 100 further includes user device 130, which is operated by user 101 and connected to computer 105 via network 135. Examples of user device 130 include input devices such as a keyboard or a voice recognition subsystem that enable user 101 to convey selections of information and commands to processor 110. User device 130 further includes an output device such as a display device, a printer, or a voice synthesizer. A cursor control unit such as a mouse, a trackball, or a touch-sensitive screen enables user 101 to manipulate a cursor on the display device to convey further selections of information and commands to processor 110.

[0086] Processor 110 outputs the result 122 of the execution of program module 120 to user device 130. Alternatively, processor 110 can direct the output to a storage device 125 such as a database or a memory, or to a remote device (not shown) via network 135.

[0087] For example, a program that performs the operations of FIG. 1 may be used as program module 120. System 100 can be operated as determination device 301.

[0088] The terms "comprising" or "including" are to be construed as specifying the presence of the features, integers, steps, or components recited therein, but not as precluding the presence of one or more other features, integers, steps, or components, or groups thereof. The terms "a" and "an" are indefinite articles and, therefore, do not preclude embodiments having a plurality of the same.

[0089] (Other embodiments) It should be noted that the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the gist of the present invention. In short, the present invention is not limited to the upper embodiments as they are, and at the implementation stage, the components can be modified without departing from the gist of the invention to be embodied.

[0090] In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiments. Furthermore, components from different embodiments may be appropriately combined.

[0091] (Advantages of the invention) The determination device, determination method, and program disclosed in this specification have the following advantages. First, by using the peak frequency and peak amplitude of vibrations in the longitudinal direction of the optical cable as characteristic quantities and learning the vibration distribution waveform, it is possible to classify and determine the laying environment into two learned categories. Second, by using this classification determination algorithm, it can be applied to the classification determination of various laying environments such as underground or overhead, as well as pole positions and cable slack.

Explanation of reference numerals

[0092] 11: Peak frequency of Fourier spectrum 12: Peak amplitude of Fourier spectrum 13: Linear sum 14: Output value y 15: Weight vector 21: Optical cable 22: Ground 23: Underground 24: Aerial 25: Utility pole 26: Slack 27: Vibration distribution measuring instrument 28: First vibration distribution data 29: Second vibration distribution data 31: Light intensity distribution 32: Local section 33: Waveform pattern 34: Waveform pattern after Δt seconds 35: Change Δν 41: Waveform g(t) 42: Fourier spectrum 100: System 101: User 105: Computer 110: Processor 115: Memory 120: Program module 122: Result 125: Storage device 130: User device 135: Network 140: Storage device 150: Data source 210: Third vibration distribution data 211: Storage unit 212: Feature extraction unit 212E: Spectrum envelope extraction unit 213: Data reading unit 214: Fourier transform unit 215: Peak frequency extraction unit 216: Peak amplitude extraction unit 217: Discrimination operation unit 218: Discrimination dictionary 219: Discrimination function operation unit 220: Classification determination unit 221: Result display unit 222: Discrimination function operation unit 223: Error operation unit 224: Teacher signal 225: Mean squared error operation unit 226: Weight vector update unit 301: Determination device

Claims

1. A determination device for classifying the laying environment of an optical cable, comprising: a feature extraction unit that extracts a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable; a calculation unit that calculates an identification function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified; a determination unit that compares the identification function with a teacher signal representing two states for each position of the optical cable and determines the state of the optical cable as the state of the closer teacher signal; and the feature extraction unit converts the vibration distribution from a waveform in the time domain to a spectral waveform in the frequency domain, removes the fluctuation components included in the spectral waveform, and extracts the feature vector using the envelope of the spectral waveform from which the fluctuations have been removed. The determination device is characterized by this.

2. A determination device for classifying the laying environment of an optical cable, comprising: a feature extraction unit that extracts a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable; a calculation unit that calculates an identification function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified; a determination unit that compares the identification function with a teacher signal representing two states for each position of the optical cable and determines the state of the optical cable as the state of the closer teacher signal; and the feature extraction unit performs a Fourier transform on the vibration distribution from a waveform in the time domain to a spectral waveform in the frequency domain, extracts the frequency of each peak and the amplitude of the peak from the spectral waveform, arranges the frequency and the amplitude in the order of the peaks with larger amplitudes as the feature vector, the weight vector is composed of coefficients that multiply the frequency and the amplitude of each peak, and the calculation unit adds the values obtained by multiplying the coefficients of the weight vector by the frequency and the amplitude of each peak to obtain the identification function. The determination device is characterized by this.

3. A determination device for classifying the laying environment of an optical cable, comprising: a feature extraction unit that extracts a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable; a calculation unit that calculates an identification function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified; For each position of the optical cable, compare the discriminant function with a teacher signal representing two states, and a determination unit that determines the state of the optical cable to be the state of the closer teacher signal; comprising; The feature extraction unit: Performs a Fourier transform on the vibration distribution from the time-domain waveform to the frequency-domain spectral waveform; Removes the fluctuation component from among the components having a period in the frequency-axis direction of the spectral waveform; Extracts the frequency of the peak and the amplitude of the peak from the spectral waveform after removing the fluctuation component; Constructs the feature vector using the extracted frequency of the peak and the amplitude of the peak; Determination device.

4. The feature extraction unit: Takes the logarithm of the absolute value of the amplitude of the spectral waveform; Calculates the time waveform of the component having a period in the frequency-axis direction of the logarithmic spectral waveform from the logarithmized logarithmic spectral waveform; Extracts a time region having a value smaller than a predetermined threshold time in the time waveform to Remove the fluctuation component; The determination device according to claim 3.

5. A determination device for classifying the laying environment of an optical cable, comprising: A feature extraction unit that extracts a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable; An arithmetic unit that calculates a discriminant function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified; A determination unit that compares the discriminant function with a teacher signal representing two states for each position of the optical cable, and determines the state of the optical cable to be the state of the closer teacher signal; An identification dictionary having a dictionary arithmetic unit and an update unit; comprising; The feature extraction unit extracts a known feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable for each of the two states; The dictionary arithmetic unit calculates a known discriminant function for each position of the optical cable from the known feature vector and the weight vector; The update unit calculates, for each position of the optical cable, the error between the known feature vector for each of the two states and the teacher signal representing the corresponding state among the two states, and updates the weight vector so that the error becomes smaller. The determination device characterized by the above.

6. A determination method executed by a determination device for classifying the laying environment of an optical cable, comprising: Extracting a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable; Calculating an identification function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified, and Comparing, for each position of the optical cable, the identification function with a teacher signal representing two states, and determining the state of the closer teacher signal as the state of the optical cable, comprising Converting the vibration distribution from a waveform in the time domain to a spectral waveform in the frequency domain, Removing the fluctuation components included in the spectral waveform, Extracting the feature vector using the envelope of the spectral waveform from which the fluctuations have been removed characterized by the above.

7. A determination method executed by a determination device for classifying the laying environment of an optical cable, Extracting a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable, Calculating an identification function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified, and Comparing, for each position of the optical cable, the identification function with a teacher signal representing two states, and determining the state of the closer teacher signal as the state of the optical cable, comprising Performing a Fourier transform on the vibration distribution from a waveform in the time domain to a spectral waveform in the frequency domain, extracting the frequency of each peak and the amplitude of the peak from the spectral waveform, arranging the frequency and the amplitude in the order of the peaks with larger amplitudes as the feature vector, The weight vector is composed of coefficients that multiply the frequency and the amplitude of each peak, Adding the values obtained by multiplying the coefficients of the weight vector by the frequency and the amplitude of each peak as the identification function, characterized by the above.

8. A determination method executed by a determination device for classifying the laying environment of an optical cable, Extracting a feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable, Calculating an identification function for each position of the optical cable from the feature vector and a weight vector corresponding to the laying environment of the optical cable to be classified, and Comparing, for each position of the optical cable, the identification function with a teacher signal representing two states, and determining the state of the closer teacher signal as the state of the optical cable, comprising Performing a Fourier transform on the vibration distribution from a waveform in the time domain to a spectral waveform in the frequency domain, Remove the fluctuation component from the components having a period in the frequency axis direction of the spectrum waveform, Extract the frequency of the peak and the amplitude of the peak from the spectrum waveform after removing the fluctuation component, Construct the feature vector using the frequency of the extracted peak and the amplitude of the peak, Determination method.

9. Logarithmize the absolute value of the amplitude of the spectrum waveform, Calculate the time waveform of the component having a period in the frequency axis direction of the logarithm spectrum waveform from the logarithmized logarithm spectrum waveform, Extract the time region having a value smaller than a predetermined threshold time in the time waveform, Remove the fluctuation component, The determination method according to claim 8.

10. A determination method executed by a determination device for classifying the laying environment of an optical cable, Extract the feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable, Calculate the discriminant function for each position of the optical cable from the feature vector and the weight vector corresponding to the laying environment of the optical cable to be classified, For each position of the optical cable, compare the discriminant function with the teacher signal representing two states, and determine the state of the closer teacher signal as the state of the optical cable, Extract the known feature vector for each position of the optical cable from the vibration distribution in the longitudinal direction of the optical cable for each of the two states, Calculate the known discriminant function for each position of the optical cable from the known feature vector and the weight vector, For each position of the optical cable, calculate the error between the known feature vectors for each of the two states and the teacher signal representing the corresponding state among the two states, and Update the weight vector so that the error becomes small A determination method characterized by performing the above.

11. A program for causing a computer to function as the determination device according to any one of claims 1 to 5.

Images