Optical cable risk monitoring device and method, controller and computer program product

Abstract

The invention discloses an optical cable risk monitoring device and method, a controller and a computer program product, and belongs to the technical field of optical cable monitoring. The device comprises a multi-dimensional feature sensing circuit and a controller, and the multi-dimensional feature sensing circuit is in communication connection with the controller; the multi-dimensional feature sensing circuit is connected with the optical fiber span section and is used for acquiring multi-dimensional feature information of the optical fiber span section; wherein the optical cable comprises a plurality of optical fiber spans which are cascaded in sequence, and each optical fiber span corresponds to one multi-dimensional feature sensing circuit; the multi-dimensional feature information comprises at least two of optical fiber vibration information, optical fiber temperature information, optical fiber stress information and optical fiber loss information; and the controller is used for acquiring the multi-dimensional feature information and determining the risk type of the optical fiber cross-section based on the multi-dimensional feature information. Through the mode, various risk types of the optical fiber span can be accurately monitored, and the monitoring efficiency of an optical cable network in a complex environment is improved.

Description

Technical Field

[0001] This application relates to the field of optical cable monitoring technology, and in particular to an optical cable risk monitoring device, method, controller and computer program product. Background Technology

[0002] As the core medium for data transmission, the physical security of optical cables directly affects the safe and stable operation of communication networks. Long-distance optical cables are typically laid in various ways, such as direct burial, ducts, and overhead installations. These cables are located in complex and diverse environments and face various risks and threats, including construction excavation, geological disasters, animal attacks, human damage, and natural aging.

[0003] Currently, the operation and maintenance of optical cables largely relies on manual inspections and reactive repairs after a fault occurs. However, manual inspections have limitations in coverage, and the types of risks associated with optical cables in complex environments are numerous, making routine monitoring of large-scale and long-distance optical cable networks difficult. Once a fault such as an optical cable break or performance degradation occurs, it usually takes several hours or even longer for manual troubleshooting, resulting in delays in service recovery. Summary of the Invention

[0004] This application provides an optical cable risk monitoring device, method, controller, and computer program product to at least solve the problem of the difficulty in monitoring large-scale, long-distance optical cable networks.

[0005] To solve the above-mentioned technical problems, this application is implemented as follows: In a first aspect, embodiments of this application provide an optical cable risk monitoring device, comprising: a multi-dimensional feature sensing circuit and a controller, wherein the multi-dimensional feature sensing circuit is communicatively connected to the controller; the multi-dimensional feature sensing circuit is connected to an optical fiber span and is used to collect multi-dimensional feature information of the optical fiber span; wherein the optical cable includes multiple cascaded optical fiber spans, each optical fiber span corresponding to a multi-dimensional feature sensing circuit; the multi-dimensional feature information includes at least two of optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information; the controller is used to acquire the multi-dimensional feature information and determine the risk type of the optical fiber span based on the multi-dimensional feature information.

[0006] Secondly, embodiments of this application provide a method for monitoring the risk of optical cables, applied to a controller, comprising: acquiring multi-dimensional feature information of an optical fiber span connected to a multi-dimensional feature sensing circuit; wherein the multi-dimensional feature information includes at least two of optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information; the optical cable includes multiple optical fiber spans, each optical fiber span corresponding to a multi-dimensional feature sensing circuit; and determining the risk type of the optical fiber span based on the multi-dimensional feature information.

[0007] Thirdly, embodiments of this application provide a controller, the controller including a processor and a memory, the memory storing programs or instructions executable on the processor, the programs or instructions being executed by the processor as described in the second aspect above.

[0008] Fourthly, embodiments of this application provide a computer-readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the second aspect above.

[0009] Fifthly, embodiments of this application provide a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, which, when executed by a computer, cause the computer to perform the steps of the method described in the second aspect above.

[0010] In this embodiment, the optical cable risk monitoring device includes a multi-dimensional feature sensing circuit and a controller, with the multi-dimensional feature sensing circuit communicatively connected to the controller. The multi-dimensional feature sensing circuit is connected to an optical fiber span and is used to collect multi-dimensional feature information of the optical fiber span. The optical cable includes multiple cascaded optical fiber spans, each span corresponding to a multi-dimensional feature sensing circuit. The multi-dimensional feature information includes at least two of the following: optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information. The controller is used to acquire the multi-dimensional feature information and determine the risk type of the optical fiber span based on it. Thus, by deploying a multi-dimensional feature sensing circuit corresponding to each optical fiber span of the optical cable, and synchronously monitoring the optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information through this circuit, the collected multi-dimensional feature information can comprehensively reflect the operating status parameters of the optical cable. Therefore, based on this multi-dimensional feature information, various risk types of the optical fiber span can be accurately monitored, thereby improving the monitoring efficiency of optical cable networks in complex environments.

[0011] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0012] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0013] Figure 1 The present application provides schematic diagrams of the structure of optical cable risk monitoring devices according to some embodiments. Figure 2The diagram shows a schematic representation of the structure of a multidimensional feature sensing circuit provided in some embodiments of this application. Figure 3 The following are schematic diagrams illustrating the structure of multidimensional feature sensing circuits provided in other embodiments of this application; Figure 4 This paper illustrates a schematic diagram of the frequency distribution of backscattered light provided in some embodiments of this application; Figure 5 This paper shows a block diagram of a multi-dimensional feature sensing circuit optical transmission system configuration provided in some embodiments of this application; Figure 6 A schematic diagram of the wavelength distribution of a multidimensional feature sensing circuit provided in some embodiments of this application is shown; Figure 7 A flowchart illustrating some embodiments of the optical cable risk monitoring method provided in this application is shown; Figure 8 A flowchart illustrating a risk type determination method provided in some embodiments of this application is shown; Figure 9 The diagram shows a structural schematic of a risk identification model provided in some embodiments of this application; Figure 10 A schematic diagram of a network structure with a self-attention layer with ROPE provided in some embodiments of this application is shown; Figure 11 A schematic diagram of the hardware structure of the controller provided in an embodiment of this application is shown. Detailed Implementation

[0014] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0015] In the field of optical cable monitoring, the operation and maintenance of optical cables mostly rely on manual inspections and reactive repairs after faults. However, manual inspections have limitations in coverage, are inefficient, and cannot achieve real-time monitoring, making routine monitoring of large-scale and long-distance optical cable networks difficult. Once a fault such as an optical cable interruption or performance degradation occurs, it usually takes several hours or even longer for manual troubleshooting, resulting in delays in service recovery. This not only leads to user complaints but can also cause significant direct and indirect economic losses. Most related optical cable risk prediction solutions are designed for monitoring and predicting fiber optic risks in single scenarios. Their perception dimensions are limited, and their false alarm rate is high, making it difficult to achieve early warning of fiber optic risks in complex environments.

[0016] To address the aforementioned problems in optical cable monitoring, this application provides an optical cable risk monitoring device. This device deploys multi-dimensional feature sensing circuits at each optical fiber span to simultaneously monitor optical fiber vibration, temperature, stress, and loss information. The collected multi-dimensional feature information comprehensively reflects the optical cable's operating status, and based on this information, it can accurately monitor various risk types in optical fiber spans, thereby improving the monitoring efficiency of optical cable networks in complex environments.

[0017] Please see Figure 1 , Figure 1 A schematic diagram of the structure of an optical cable risk monitoring device provided in some embodiments of this application is shown. The device 100 includes a multi-dimensional feature sensing circuit 110 and a controller 120, with the multi-dimensional feature sensing circuit 110 and the controller 120 communicatively connected.

[0018] The multi-dimensional feature sensing circuit 110 is connected to the optical fiber span and is used to collect multi-dimensional feature information of the optical fiber span. The optical cable 200 includes multiple cascaded optical fiber spans, and each optical fiber span corresponds to a multi-dimensional feature sensing circuit 110. The multi-dimensional feature information includes at least two of the following: optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information.

[0019] The controller 120 is used to acquire multi-dimensional feature information and determine the risk type of the fiber optic segment based on the multi-dimensional feature information.

[0020] In an exemplary embodiment, the optical cable 200 includes a plurality of sequentially cascaded optical fiber segments, for example, optical fiber segment 1, optical fiber segment 2, ..., optical fiber segment n, where n is a positive integer. Since the multidimensional feature sensing circuit 110 is a signal-to-noise ratio limited system, the multidimensional feature sensing circuit 110 is deployed segment by segment, with each optical fiber segment corresponding to one multidimensional feature sensing circuit. For example, optical fiber segment 1 corresponds to multidimensional feature sensing circuit 1101, optical fiber segment 2 corresponds to multidimensional feature sensing circuit 1102, ..., and optical fiber segment n corresponds to multidimensional feature sensing circuit 110n.

[0021] Taking fiber optic span 2 as an example, the multi-dimensional feature sensing circuit 110 synchronously monitors the fiber vibration information, fiber temperature information, fiber stress information, and fiber loss information of fiber optic span 2, forming multi-dimensional feature information. This multi-dimensional feature information can comprehensively reflect the operating status parameters of the optical cable, such as vibration, temperature, stress, and loss. The multi-dimensional feature sensing circuit 110 can send the collected multi-dimensional feature information to the controller through a preset data transmission interface (or, the controller can actively acquire the multi-dimensional feature information collected by the multi-dimensional feature sensing circuit 110 through a preset data transmission interface). Based on this multi-dimensional feature information, the controller determines the risk type of fiber optic span 2.

[0022] Multidimensional feature information may also include other feature information of optical fibers, which are not specifically limited in the embodiments of this application.

[0023] By means of the above method, multi-dimensional feature information of optical fiber spans is collected through multi-dimensional feature sensing circuit. This multi-dimensional feature information can comprehensively reflect the operating status parameters of the optical cable. Based on this multi-dimensional feature information, various risk types of optical fiber spans can be accurately monitored, thereby solving the problems of limited monitoring range and single sensing dimension of related risk monitoring methods, making it difficult to monitor the risks of optical cable networks in complex environments in real time, and improving the monitoring efficiency of optical cable networks in complex environments.

[0024] In some embodiments, external physical quantities such as vibration, stress, temperature, and loss have specific sensitivities to various types of scattered light. For example, external stress causes changes in the frequency shift of Brillouin backscattered light; temperature affects the intensity of Raman backscattered light; and vibration and fiber loss are reflected in the phase fluctuations and intensity changes of Rayleigh backscattered light, respectively. Figure 2 As shown, the multidimensional feature sensing circuit 110 mentioned above includes: an optical signal processing unit 111 and a multi-channel hardware sensing unit 112.

[0025] The optical signal processing unit 111 is connected to the optical fiber span and is used to obtain multiple backscattered optical signals of different types returned by the optical fiber span based on the probe optical signal by sending a probe optical signal to the optical fiber span. The multiple hardware sensing units 112 with different types of backscattered optical signal processing functions are connected in parallel to each other. The multiple hardware sensing units are used to obtain the corresponding type of backscattered optical signal, process the backscattered optical signal, and obtain multi-dimensional feature information of the optical fiber span.

[0026] In an exemplary embodiment, the optical signal processing unit 111 sends a probe optical signal to the optical fiber span and acquires multiple backscattered light signals of different types, such as Brillouin backscattered light, Raman backscattered light, and Rayleigh backscattered light, returned by the optical fiber span based on the probe optical signal.

[0027] Each hardware sensing unit possesses different types of backscattered light signal processing capabilities. By acquiring and processing the corresponding types of backscattered light signals, it obtains multi-dimensional characteristic information of the fiber optic segment. For example, the multi-channel hardware sensing unit includes a first hardware sensing unit, a second hardware sensing unit, a third hardware sensing unit, and a fourth hardware sensing unit. The first hardware sensing unit processes Brillouin backscattered light to obtain stress information on the fiber; the second hardware sensing unit processes Raman backscattered light to obtain fiber temperature information; the third hardware sensing unit processes Rayleigh backscattered light to obtain fiber vibration information; and the fourth hardware sensing unit processes Rayleigh backscattered light to obtain fiber loss information. These multiple hardware sensing units are connected in parallel to synchronously monitor multi-dimensional characteristic information, addressing the problem of limited sensing dimensions in related risk monitoring methods.

[0028] In some possible implementations, such as Figure 3 As shown, the aforementioned multi-channel hardware sensing unit 112 includes at least two of the following: fiber optic vibration sensing device 112a, fiber optic stress sensing device 112b, fiber optic temperature sensing device 112c, and fiber optic loss sensing device 112d.

[0029] Fiber optic vibration sensing equipment is used to sense vibration from acquired Rayleigh backscattered light signals to obtain fiber vibration information of the fiber span; fiber optic stress sensing equipment is used to sense stress from acquired Brillouin backscattered light signals to obtain stress information of the fiber span; fiber optic temperature sensing equipment is used to sense temperature from acquired Raman backscattered light signals to obtain fiber temperature information of the fiber span; and fiber optic loss sensing equipment is used to sense loss from acquired Rayleigh backscattered light signals to obtain fiber loss information of the fiber span.

[0030] Among them, the aforementioned fiber optic vibration sensing devices are sensing devices used to monitor fiber optic vibration, including optical time domain reflectometers (OTDRs) and phase optical time domain sensors; the aforementioned fiber optic stress sensing devices are sensing devices used to monitor fiber optic stress, including Brillouin OTDRs, Brillouin optical time domain analyzers, and Brillouin optical time domain sensors; the aforementioned fiber optic temperature sensing devices are sensing devices used to monitor fiber optic temperature, including Raman OTDRs and Raman sensors; and the aforementioned fiber optic loss sensing devices are sensing devices used to monitor fiber optic loss, including conventional OTDRs and optical time domain reflectometers.

[0031] In an exemplary embodiment, the fiber optic vibration sensing device 112a includes a first optical coupler (i.e., optical coupler 3), a first balanced photodetector (i.e., balanced photodetector 1), a first analog-to-digital converter (i.e., analog-to-digital converter 1), and a first digital signal processing module. The first optical coupler is used to perform optical domain coherent beat frequency conversion between the Rayleigh backscattered light signal and the local oscillator light corresponding to the probe light signal in the fiber span. Subsequently, the beat frequency signal is sequentially passed through the first balanced photodetector and the first analog-to-digital converter for photoelectric conversion, analog-to-digital conversion, and sampling. The first digital signal processing module demodulates the vibration characteristics to obtain the distribution information of external vibration along the fiber in the time and spatial dimensions, which is presented in tabular or two-dimensional matrix form, as shown in Table 1 below.

[0032] Table 1 Vibration monitoring results of fiber optic vibration sensing equipment

[0033] Its corresponding matrix form is shown below: ; in, This represents the fiber vibration information at time t and position z. , .

[0034] The matrix elements in the formula are consistent with those in Table 1 above. Each row represents the distribution of vibration characteristics along the optical fiber at a specific moment; each column represents the evolution of vibration characteristics over time at a fixed position. Since tables and matrices are completely equivalent in terms of information representation, and the matrix form is more convenient for subsequent algorithm processing, the monitoring results of each hardware sensing unit will be uniformly represented in matrix form below.

[0035] Because there is a frequency shift of approximately 10-11 GHz between the Brillouin backscattered light and the incident light, the local oscillator light needs to be frequency modulated to shift its center frequency to match that of the Brillouin backscattered light. The fiber optic stress sensing device 112b includes an optical coupler comprising a second optical coupler (i.e., optical coupler 4), a second balanced photodetector (i.e., balanced photodetector 2), a second analog-to-digital converter (i.e., analog-to-digital converter 2), and a second digital signal processing module. The second optical coupler is used to perform optical domain coherent beat frequency conversion between the Brillouin backscattered light signal and the modulated local oscillator light. The beat frequency signal is then sequentially processed by the second balanced photodetector and the second analog-to-digital converter through photoelectric conversion, analog-to-digital conversion, and sampling before being sent to the second digital signal processing unit. The second digital signal processing module extracts the Brillouin frequency shift to inversely determine the stress change in the fiber. The stress sensing result is expressed as an M×N matrix, as shown in the following equation.

[0036] .

[0037] in, This represents the stress information of the optical fiber at time t and position z. , .

[0038] Raman backscattered light signals consist of Stokes light and anti-Stokes light, with the intensity of the anti-Stokes light being highly sensitive to temperature changes. There are three temperature demodulation methods: the first uses only a single anti-Stokes light for demodulation; the second uses the intensity ratio of the anti-Stokes light to the Rayleigh backscattered light for demodulation; and the third uses the intensity ratio of the anti-Stokes light to the Stokes light for demodulation. The first two methods are susceptible to fluctuations in light source power, uneven fiber loss, and system noise, resulting in relatively limited stability and accuracy in temperature detection. In contrast, the third method, through comparison operations of co-source Raman scattering, can effectively suppress the influence of interference; therefore, this embodiment employs the third method for temperature sensing.

[0039] In an exemplary embodiment, the fiber optic temperature sensing device 112c includes an optical demultiplexer, a first photodetector (i.e., photodetector 1), a second photodetector (i.e., photodetector 2), a third analog-to-digital converter (i.e., analog-to-digital converter 3), a fourth analog-to-digital converter (i.e., analog-to-digital converter 4), and a third digital signal processing module. The optical demultiplexer demultiplexes the Raman backscattered light signal into Stokes light and anti-Stokes light. The Stokes light undergoes photoelectric conversion, analog-to-digital conversion, and sampling processing by the first photodetector and the third analog-to-digital converter before being sent to the third digital signal processing module. The anti-Stokes light undergoes photoelectric conversion, analog-to-digital conversion, and sampling processing by the second photodetector and the fourth analog-to-digital converter before being sent to the third digital signal processing module. The third digital signal processing module inverts the change in fiber optic temperature by the intensity ratio of the anti-Stokes light to the Stokes light. The temperature sensing result is expressed in M×N matrix form, as shown in the following equation: .

[0040] in, This represents the fiber optic temperature information at time t and location z. , .

[0041] The fiber optic loss sensing device 112d is used to detect fiber optic loss. Its working principle is based on the measurement of the intensity of Rayleigh backscattered light. Unlike fiber optic vibration sensing devices, this method only needs to acquire the intensity information of the Rayleigh scattering signal, without needing to analyze its phase. Since the intensity of Rayleigh backscattered light is usually much higher than that of Brillouin scattering and Raman scattering signals, it can be directly received and photoelectrically converted by a high-performance third photodetector (i.e., photodetector 3) and a fifth analog-to-digital converter (i.e., analog-to-digital converter 5). After the obtained signal is analyzed by the fourth digital signal processing unit, the distribution of fiber loss along the line with distance can be reconstructed, thereby realizing effective monitoring of the attenuation characteristics of the fiber optic link. The loss sensing result is expressed in M×N matrix form, as shown in the following equation: .

[0042] in, This represents the fiber optic temperature information at time t and location z. , .

[0043] The vibration, stress, temperature and loss sensing results generated above are combined to form a 4×M×N three-dimensional matrix, called multi-dimensional feature information, and then input to the controller 120 through the data transmission interface to determine the risk type of the fiber optic segment.

[0044] In practical applications, the aforementioned analog-to-digital converter 1, first digital signal processing module, analog-to-digital converter 2, second digital signal processing module, analog-to-digital converter 3, analog-to-digital converter 4, third digital signal processing module, analog-to-digital converter 5, and fourth digital signal processing module use a synchronized clock signal to ensure the synchronous sampling and processing of multiple backscattered light signals of different types. This avoids problems such as sampling errors and data misalignment caused by clock deviation, thereby improving the detection accuracy of multi-dimensional characteristic information such as fiber vibration information, fiber temperature information, fiber stress information, and fiber loss information.

[0045] In this way, by collecting fiber optic vibration information through fiber optic vibration sensors, stress information through fiber optic stress sensors, temperature information through fiber optic temperature sensors, and loss information through fiber optic loss sensors, it is possible to simultaneously collect at least two of the physical quantities of vibration, stress, temperature, and loss. This solves the problem of the single sensing dimension of related risk monitoring methods, improves the comprehensiveness of monitoring the operating status parameters of optical cables, and provides a comprehensive and accurate data foundation for subsequent controllers to determine the risk type of fiber optic segments.

[0046] In some embodiments, the optical signal processing unit 111 described above includes a signal transceiver unit 1111 and a signal distribution unit 1112.

[0047] The signal transceiver unit 1111 is connected to the optical fiber span and is used to obtain the composite backscattered light signal returned by the optical fiber span based on the probe light signal by sending a probe light signal to the optical fiber span.

[0048] The signal distribution unit 1112 is used to acquire the composite backscattered light signal and separate the composite backscattered light signal into multiple backscattered light signals of different types.

[0049] Continuing with the above embodiment, a probe light signal is transmitted to the optical fiber span via the signal transceiver unit 1111, and the composite backscattered light signal returned by the optical fiber span based on the probe light signal is acquired. The composite backscattered light signal is separated into multiple backscattered light signals of different types by the signal distribution unit 1112. Specifically, the signal distribution unit 1112 can separate the composite backscattered light signal into multiple backscattered light signals of different types according to the characteristics of the backscattered light signal.

[0050] Taking Rayleigh backscattered light signals, Brillouin backscattered light signals, and Raman backscattered light signals as examples, the characteristics of each type of backscattered light are explained. Figure 4 As shown, in terms of intensity characteristics, the intensity of the three types of backscattered light decreases in the following order: Rayleigh backscattered light is the strongest, followed by Brillouin backscattered light, and Raman backscattered light is the weakest. In terms of frequency characteristics, compared to the center frequency of the probe pulse, the frequency shift of Brillouin backscattered light is approximately 10-11 GHz, the frequency shift of Raman backscattered light is approximately 13 THz, while the frequency of Rayleigh backscattered light is basically the same as that of the probe pulse.

[0051] Based on the characteristics of the three types of backscattered light signals mentioned above, the composite backscattered light signal is separated into Rayleigh backscattered light signal, Brillouin backscattered light signal, and Raman backscattered light signal. Furthermore, multi-dimensional characteristic information of the fiber optic span is determined based on these three types of backscattered light.

[0052] In this way, by acquiring the composite backscattered light signal returned by the probe light signal across the optical fiber segment, and separating the composite backscattered light signal into multiple backscattered light signals of different types, it is possible to achieve parallel sensing and accurate differentiation of multiple physical quantities such as optical cable vibration, temperature, stress, and loss, thereby improving the accuracy and anti-interference capability of optical fiber risk monitoring in complex environments and reducing the false alarm rate of risk warning.

[0053] In some possible implementations, the signal transceiver unit 1111 described above includes a laser, a signal modulation unit, and an optical circulator.

[0054] A laser is used to output continuous light.

[0055] The signal modulation unit is used to modulate continuous light into a probe light signal.

[0056] The optical circulator includes a first port p1, a second port p2, and a third port p3. The optical circulator 1111c is used to acquire a probe light signal through the first port p1, send the probe light signal to the fiber optic span through the second port p2, receive a composite backscattered light signal, and send the composite backscattered light signal to the signal distribution unit through the third port p3.

[0057] For example, the signal modulation unit includes Figure 3 The modulator 1, optical amplifier and optical filter are included.

[0058] Continuing with the above embodiment, a continuous light is output by a laser, the frequency of which can be set according to actual needs; a signal modulation unit modulates the continuous light into a probe light signal, wherein the modulator 1 in the signal modulation unit is used to modulate the continuous light into pulse light, and the optical amplifier is used to compensate for the modulation loss of the modulator 1 and the loss introduced by subsequent filters, circulators and other devices, thereby ensuring that the optical power of the probe light signal input to the fiber optic span reaches the target value; the optical filter is used to filter out the broadband amplifier spontaneous emission noise (ASE) generated by the optical amplifier, so as to avoid the system signal-to-noise ratio from decreasing due to excessive noise.

[0059] like Figure 3 As shown, the optical circulator adopts a three-port structure, following the directional transmission rule of "input at port p1, output at port p2, input at port p2, and output at port p3". The filtered probe light signal enters from the first port p1 of the circulator and exits from the second port p2, propagating as probe light into the fiber optic span. During propagation within the fiber optic span, this probe light signal generates a composite backscattered light signal, which includes one or more of Rayleigh backscattered light, Brillouin backscattered light, and Raman backscattered light. This backscattered light returns along its original path to the second port p2 of the optical circulator, and is then output via the third port p3, entering the signal distribution unit 1112 for further processing.

[0060] The above structure enables unidirectional emission of the probe light signal and directional reception and separation of the composite backscattered light signal, ensuring that the emission path of the probe light signal and the reception path of the composite backscattered light signal are isolated from each other, thus improving the stability of signal detection.

[0061] In some possible implementations, in response to the inclusion of an optical fiber vibration sensing device 112a in the multi-channel hardware sensing unit 112, the signal transceiver unit 1111 further includes an optical coupler ( Figure 3 In the optical coupler 1), the laser is connected to the signal modulation unit through the optical coupler.

[0062] An optical coupler is used to split the acquired continuous light into signal light and local oscillator light. A signal modulation unit is used to modulate the acquired signal light into a probe light signal. An optical coupler is connected to an optical fiber vibration sensing device 112a and is used to send local oscillator light to the optical fiber vibration sensing device 112a; the optical fiber vibration sensing device 112a obtains optical fiber vibration information of the optical fiber span based on the corresponding type of backscattered light signal and local oscillator light.

[0063] Continuing with the above embodiments, the fiber optic vibration sensing device 112a achieves vibration monitoring by demodulating the phase information that changes with the external vibration of the optical cable, thus employing a coherent reception method. Specifically, the optical coupler 1 divides the acquired continuous light into signal light and local oscillator light; the signal modulation unit is used to modulate the acquired signal light into a probe light signal; the local oscillator light output from the optical coupler 1 undergoes optical domain coherent beat frequency interaction with the Rayleigh backscattered light signal in the optical coupler 3, and the fiber vibration information of the fiber span is determined based on this beat frequency signal.

[0064] This method can provide local oscillator light for fiber optic vibration sensing devices, enabling coherent detection of Rayleigh backscattered light signals and improving the accuracy of fiber optic vibration information detection.

[0065] In some other possible implementations, in response to the inclusion of an optical fiber stress sensing device 112b in the multiplex hardware sensing unit 112, the optical coupler transmits signals through a modulator (…). Figure 3 The modulator 2) is connected to the fiber optic stress sensing device 112b.

[0066] The modulator is used to modulate the local oscillator light and sends the modulated local oscillator light to the fiber optic stress sensing device 112b. The fiber optic stress sensing device 112b obtains the stress information of the fiber span based on the backscattered light signal of the corresponding type and the modulated local oscillator light.

[0067] For example, the modulator can shift the center frequency of the local oscillator light to the same frequency as the Brillouin backscattered light signal using a preset radio frequency signal, and send the modulated local oscillator light to the fiber optic stress sensing device 112b; the preset radio frequency signal can be generated by a radio frequency signal generator.

[0068] Continuing with the above embodiment, since there is a frequency offset of approximately 10-11 GHz between the Brillouin backscattered light and the incident light, the local oscillator light can be frequency modulated by a modulator to shift its center frequency to match that of the Brillouin backscattered light. An RF signal generator produces a 10-11 GHz RF signal, which, along with the local oscillator light, is input into modulator 2. After the center frequency of the modulated local oscillator light is shifted to match the frequency of the Brillouin backscattered light, the fiber optic stress sensing device 112b obtains the stress information of the fiber span based on the corresponding type of backscattered light signal and the modulated local oscillator light.

[0069] This method can provide fiber optic stress sensing devices with modulated local oscillator light of the same frequency as the corresponding type of backscattered light signal, enabling coherent detection of Brillouin backscattered light signals and improving the detection accuracy of fiber optic stress information.

[0070] In some other possible implementations, the aforementioned optical coupler uses an optical switching component ( Figure 3 The optical switch 1) is connected to the fiber optic vibration sensing device 112a and the modulator 2 respectively.

[0071] In this way, by setting up an optical switch component, the optical fiber vibration sensor 112a or the optical fiber stress sensor 112b can be selected according to actual needs. This enables flexible switching and optical path multiplexing of multiple hardware sensing units, simplifies the system structure, and reduces hardware costs.

[0072] In some embodiments, the signal distribution unit 1112 described above includes at least one of an optical coupler and an optical switch.

[0073] Continuing with the above embodiments, the optical coupler is used to split the acquired composite backscattered light signal into multiple channels, resulting in multiple backscattered light signals of different types. Each backscattered light signal can be detected simultaneously. If the optical coupler is replaced with an optical switch, time-division multiplexing of multiple different types of backscattered light signals can be achieved. The multiple different types of backscattered light signals correspond to fiber optic vibration sensing device 112a, fiber optic stress sensing device 112b, fiber optic temperature sensing device 112c, and fiber optic loss sensing device 112d, respectively.

[0074] The above structure allows for the separation of composite backscattered light signals into multiple backscattered light signals of different types, enabling parallel detection of different types of backscattered light signals.

[0075] In some embodiments, in response to the above-described multidimensional feature sensing circuit 110 and the fiber optic span being deployed in an in-band manner, the wavelength of the detected optical signal is located in a first band, which is the band used by the service signal of the fiber optic span; or, in response to the above-described multidimensional feature sensing circuit 110 and the fiber optic span being deployed in an out-of-band manner, the wavelength of the detected optical signal is located in a second band, which is a band other than the band used by the service signal of the fiber optic span.

[0076] In one exemplary embodiment, such as Figure 5 As shown, the optical transmission system includes multiple fiber optic spans, each consisting of service optical modules 1 to N, wavelength selective switches, optical amplifiers, multiplexers / demultiplexers, and optical fibers. When the multidimensional feature sensing circuit 110 is connected to this optical transmission system, an in-band or out-of-band configuration can be selected according to actual deployment requirements.

[0077] When the multi-dimensional feature sensing circuit 110 is deployed in an in-band configuration to connect to the fiber optic span, the multi-dimensional feature sensing circuit 110 can be connected to the fiber optic span via a multiplexer / demultiplexer. The multiplexer / demultiplexer is used to perform wavelength multiplexing and demultiplexing of the probe optical signal and the service signal of the fiber optic span, allowing them to share the same fiber core and transmit at different wavelengths. In this case, the wavelength of the probe optical signal is located in the first band, which is the band used by the service signal. In specific applications, the band used by the service signal is usually the C-band or C+L band, hereinafter collectively referred to as the "operating band," such as... Figure 6 As shown, the detection optical signal and the service signal share the optical fiber spectrum resources, which will occupy part of the service bandwidth and have a certain impact on the system communication capacity. It is suitable for scenarios with low requirements for communication capacity.

[0078] When the multi-dimensional feature sensing circuit 110 is deployed out-of-band to the optical fiber span, the multi-dimensional feature sensing circuit 110 can independently transmit the detection optical signal using a spare fiber core in the optical cable, physically isolated from the service signal. Alternatively, the operating wavelength of the multi-dimensional feature sensing circuit 110 can be configured outside the operating band, i.e., the wavelength of the detection optical signal is located in a second band, which is a band other than the band used by the service signal of the optical fiber span, such as the S-band, the guard band between the C and L bands, or the U-band. In this way, the monitoring channel can be independent of the service channel, without occupying service bandwidth, and without affecting communication capacity. This is suitable for application scenarios that are sensitive to bandwidth or have high requirements for monitoring reliability.

[0079] In some possible implementations, in response to the in-band deployment of the connection between the multidimensional feature sensing circuit 110 and the optical fiber span, an optical filter is connected to the end of the optical fiber span to which the multidimensional feature sensing circuit 110 is connected. This optical filter is used to block the probe optical signal from entering the next level optical fiber span.

[0080] Continuing with the above embodiments, when the connection between the multi-dimensional feature sensing circuit 110 and the optical fiber span is deployed in an in-band manner, since the operating wavelength of the multi-dimensional feature sensing circuit 110 is within the gain range of the optical amplifier, if the probe light signal is not filtered out and directly enters the next span, even after amplification by the optical amplifier, the signal-to-noise ratio will be significantly degraded due to accumulated noise, thus failing to achieve effective sensing of the optical fiber status. Therefore, when using the in-band deployment method, an optical filter is added at the end of each optical fiber span to block the probe light signal of the multi-dimensional feature sensing circuit 110 from continuing to transmit to the next optical fiber span.

[0081] After the multi-dimensional feature sensing circuit 110 is configured, it can support multiple measurement and control modes such as manual, periodic polling or custom strategies to monitor the status of optical cables in real time.

[0082] Figure 7 A flowchart illustrating a method for monitoring optical cable risks according to some embodiments of this application is shown. This method can be applied to a controller, such as the controller 120 described above. As shown in the figure, the method 700 specifically includes the following steps: Step 701: Obtain multi-dimensional feature information of the optical fiber segment connected to the multi-dimensional feature sensing circuit.

[0083] The multi-dimensional feature information includes at least two of the following: fiber vibration information, fiber temperature information, fiber stress information, and fiber loss information; the optical cable includes multiple fiber spans, and each fiber span corresponds to a multi-dimensional feature sensing circuit.

[0084] In an exemplary embodiment, the controller can acquire multi-dimensional feature information of the optical fiber span connected to the multi-dimensional feature sensing circuit in the following ways: 1) The multi-dimensional feature sensing circuit can send the acquired multi-dimensional feature information of the optical fiber span to the controller through a preset data transmission interface, and the controller receives the multi-dimensional feature sensing information sent by the multi-dimensional feature sensing circuit; 2) The multi-dimensional feature sensing circuit can also store the acquired multi-dimensional feature information of the optical fiber span to a preset storage area, and the controller reads the multi-dimensional feature information stored in the storage area.

[0085] Step 702: Determine the risk type of fiber optic segment based on multi-dimensional feature information.

[0086] Continuing with the above embodiments, the controller can input the multi-dimensional feature information obtained in step 701 into the pre-trained risk identification model, and determine the risk type of the fiber optic segment through the risk identification model.

[0087] The risk and hazard identification model can be trained in the following way: A dataset is constructed based on multiple multi-dimensional feature information samples and corresponding risk type labels for each sample. Risk type labels include "normal," "construction and excavation," "icing," and "fire." This dataset is then used to iteratively train a pre-defined neural network. During each iteration, feature information from each multi-dimensional feature information sample in the dataset is extracted. Based on this feature information, the risk type labels, and the loss function of the neural network model, the network parameters are adjusted until the neural network model meets the pre-defined convergence condition. The trained neural network model is then identified as the risk and hazard identification model.

[0088] During the model training phase, multi-dimensional feature information samples cover typical features of various target events, such as continuous temperature rise and stress mutation in fire scenarios, stress accumulation during icing processes, and periodic mechanical vibrations caused by municipal construction. The training process aims to optimize the model structure and parameters of the neural network so that it can accurately distinguish the feature patterns corresponding to different events.

[0089] After model training is complete, it can be deployed to a hardware board or integrated into a network management system. During actual operation, the neural network analyzes and calculates the multi-dimensional feature information input in real time, outputting the probability of occurrence of various potential events. Finally, the potential event with the highest probability is selected as the risk type for the fiber optic segment, thereby achieving automatic identification and early warning of external risks and hazards to the optical cable.

[0090] Through the above steps, multi-dimensional feature information of the optical fiber span connected to the multi-dimensional feature sensing circuit is obtained; based on the multi-dimensional feature information, the risk type of the optical fiber span is determined. This allows for simultaneous analysis of multi-dimensional feature information change patterns caused by external events. For example, in a composite scenario of "fire accompanied by structural damage," the concurrent multi-feature pattern of "high temperature - sudden stress change - sharp increase in attenuation - impact vibration" at a specific location on the optical fiber at a particular moment can be captured; while in a "optical cable icing" scenario, the typical progressive feature combination of "continuously increasing stress and a stable temperature below freezing point" can be identified, avoiding the problems of single sensing dimensions and high false alarm rates in related risk monitoring methods, and improving the reliability of optical cable risk early warning in complex environments.

[0091] Furthermore, the above method is also applicable to scenarios dominated by a single feature. Taking municipal construction as an example, mechanical vibration only generates periodic impact signals in the vibration channel, while channel data such as temperature, stress, and loss remain stable (non-contact vibration of construction machinery generally does not cause changes in other physical quantities). Therefore, the method provided in this application embodiment can accurately distinguish between single external disturbances and complex risk events based on the distribution law of multi-channel features, improving the accuracy and interpretability of event identification.

[0092] In some embodiments, such as Figure 8 As shown, in step 702 above, the risk type of fiber optic segment is determined based on multi-dimensional feature information, including: Step 7021: Extract local joint feature sequences from multi-dimensional feature information. The local joint feature sequences include multiple local spatiotemporal joint features. The local spatiotemporal joint features are coupled and correlated features of multi-dimensional feature information within a preset time and space range.

[0093] In an exemplary embodiment, a risk identification model can be constructed based on a multi-channel joint analysis architecture. This model determines the risk type of an optical fiber segment based on multi-dimensional feature information. Specifically, the risk identification model includes three cascaded processing modules that sequentially perform local spatiotemporal feature extraction, long-range temporal dependency modeling, and comprehensive risk classification.

[0094] In the local spatiotemporal feature extraction stage, a multi-channel deep three-dimensional convolutional neural network (3D-CNN) module can be used to extract local joint feature sequences from multi-dimensional feature information. This module, acting as the front-end feature extractor of the entire risk and hazard identification model, employs a three-dimensional convolutional kernel to process the 4×M×N four-channel input data (i.e., multi-dimensional feature information of the fiber optic span) composed of vibration, temperature, stress, and loss in parallel. The three-dimensional convolutional kernel performs synchronous sliding operations in the time, space, and channel dimensions, enabling joint analysis and in-depth mining of the coupling correlation of multi-source physical features within a local spatiotemporal range, yielding multiple local spatiotemporal joint features. This is particularly relevant considering the large span of fiber optic sensing systems (typically tens to hundreds of kilometers) and the localized nature of the impact range of external events.

[0095] In this way, by using the local receptive field mechanism of the three-dimensional convolution kernel, we can focus on areas along the optical cable where anomalies may occur, and capture composite patterns such as sudden local temperature rises, stress changes and vibration shocks occurring in a spatiotemporal manner, providing highly discriminative local joint feature sequences for subsequent processing.

[0096] In some possible implementations, step 7021 above, which involves extracting local joint feature sequences from multi-dimensional feature information, includes: inputting multi-dimensional feature information into a spatiotemporal feature extraction network to obtain local joint feature sequences.

[0097] The spatiotemporal feature extraction network includes at least one feature extraction layer. Each feature extraction layer is used to sequentially perform convolution, pooling, and concatenation on the input data through a target convolution kernel to obtain the output data.

[0098] Continuing with the above embodiments, as follows: Figure 9 As shown, the multi-channel 3D-CNN adopts a four-layer structure, with each layer following a processing flow of "convolution-pooling-consolidation". First, local features are extracted from the input data through 3D convolution operations. Taking the first layer as an example, the input data dimension is 4×M×N, where 4 represents the four physical channels of vibration, temperature, stress, and loss, and M and N correspond to the time and spatial dimensions, respectively. This layer is configured with S independent 3D convolution kernels, each performing parallel analysis of multi-channel spatiotemporal information from different feature response perspectives.

[0099] Since the input is a three-dimensional structure, the size of each convolutional kernel is also designed to be three-dimensional, 4×k×k. The value of k can be determined by the sampling data points corresponding to the spatial resolution of the multi-dimensional feature perception circuit.

[0100] In some possible implementations, the size of the target convolutional kernel can be determined based on the number of dimensions of the multi-dimensional feature information and the number of sampling data points corresponding to the spatial resolution of the multi-dimensional feature sensing circuit.

[0101] Continuing with the above embodiments, in order to ensure that the perception range of the convolution kernel in the spatial dimension matches the actual physical sampling range of the system, feature initialization extraction that conforms to the sensing mechanism is achieved in the first layer.

[0102] 1) When the probe light signal is a rectangular pulse, k can be determined by the following formula:

[0103] in, The sampling rate of the multi-dimensional feature sensing circuit. The width of the rectangular pulse.

[0104] 2) When the detection optical signal is a rectangular pulse combined with linear frequency modulation technology, k can be determined by the following formula:

[0105] in, This refers to the sweep bandwidth using linear frequency modulation technology.

[0106] For example, , , The two situations mentioned above correspond to k They are 500 and 50 respectively.

[0107] The convolution operation involves summing the element-wise multiplication of the convolution kernel with the local input region and then sliding the kernel with a stride of 1. To ensure that the data dimensions are the same before and after convolution, zeros are padded at the input edges during the calculation, so that the input dimension becomes 4×(M+1)×(N+1).

[0108] In convolution operations, the convolution kernel slides across the input data with a stride of 1, performing element-wise multiplication and summation of local regions at each step. To maintain the data dimensionality before and after convolution, zero-padding is performed at the input edges. Specifically, for an input with dimensions of 4×4×M×N, its dimension becomes 4×(M+1)×(N+1) after padding. Since the multiplication and addition operations of convolution are still linear transformations, simply stacking multiple convolutional layers is still equivalent to a single linear system, limiting its expressive power. Therefore, a Rectified Linear Unit (RELU) activation function is introduced after each convolutional layer. Through its non-linear mapping ability f(x)=max(0,x), the multi-layer network can learn and express complex feature patterns.

[0109] To further compress feature size, reduce the number of parameters, and enhance the spatial invariance of features, pooling (downsampling) is performed after convolution and activation. In this embodiment, a pooling window of size 2×2 can be used, and max pooling (extracting the most salient features) or average pooling (preserving the average response) can be performed within each window, thereby achieving dimensionality reduction and abstraction of the data while retaining key information.

[0110] Since multiple convolutional kernels are used for feature extraction, the pooled feature maps need to be concatenated along the channel dimension and merged into a three-dimensional output matrix. This matrix retains the abstract features extracted by all convolutional kernels, and its dimensions are S×M / 2×N / 2, where S is the number of convolutional kernels in the current layer, and M and N are the temporal and spatial dimensions of the input, respectively.

[0111] After the first layer of convolution and pooling, the data has been effectively compressed in both spatial and temporal dimensions. Starting from the second convolutional layer, to adapt to the reduced output scale of the previous layer and to further refine the extraction of multi-level features, all subsequent convolutional layers uniformly use small convolutional kernels with a size of h=3. Small-sized convolutional kernels not only significantly reduce model parameters and improve computational efficiency, but also capture local patterns more precisely on the compressed feature maps. As the network depth increases, to extract more diverse and abstract high-dimensional features, the number of convolutional kernels in each layer increases progressively. This allows the algorithm to analyze the input from different "perspectives," enhancing the richness and discriminative power of the features, thereby better supporting subsequent classification tasks. The input / output dimensions, number of convolutional kernels, and size of each convolutional layer in the multi-channel 3D-CNN are shown in Table 2 below.

[0112] Table 2 Algorithm parameters of convolutional neural networks

[0113] Step 7022: Encode the time intervals of multiple local spatiotemporal joint features by rotational position encoding to generate a position-enhanced feature sequence.

[0114] Continuing with the above embodiments, in order to establish the dependencies between multiple local spatiotemporal joint features over a long time series, a self-attention layer with Rotary Position Embedding (ROPE) can be used. After obtaining the 128×M / 16×N / 16 local joint feature sequence extracted by 3D-CNN, this layer encodes the time intervals of multiple local spatiotemporal joint features through ROPE encoding to generate a position-enhanced feature sequence. This ROPE encoding explicitly marks the relative position information between multiple local spatiotemporal joint features in the attention calculation, enabling the model to distinguish and effectively utilize the order and interval of events, thereby enhancing the modeling ability for events with temporal evolution characteristics (such as the continuous temperature rise and stress mutation of a fire, the stress accumulation process of icing, the periodic mechanical vibration of construction, etc.), and improving the system's recognition accuracy and robustness for complex temporal patterns.

[0115] Step 7023: Input the location enhancement feature sequence into the multi-head self-attention module to obtain attention features.

[0116] Among them, attention features are used to indicate the temporal dependencies between multiple local spatiotemporal joint features.

[0117] Continuing with the above embodiments, as follows: Figure 10 As shown, the position-enhanced feature sequence encoded by ROPE is input into the multi-head self-attention modules Head1, Head2, ... Head. a By using self-attention operations, global information is integrated across the entire temporal range to obtain attention features; these attention features are used to indicate the temporal dependencies between multiple local spatiotemporal joint features.

[0118] Step 7024: Determine the risk type of fiber optic segment based on attention characteristics.

[0119] Continuing with the above embodiments, by inputting the aforementioned attention features into a fully connected layer, a multi-layer fully connected neural network is used to perform nonlinear transformation and dimensionality reduction on the integrated attention feature representation, ultimately mapping it to probability distributions for different risk types.

[0120] This step enables classification decisions from a multi-dimensional feature space to a discrete risk label space, ultimately outputting event category judgments such as "normal," "construction excavation," "icing," and "fire," thus completing the risk and hazard identification results for the entire optical cable monitoring process.

[0121] This application provides a method for determining risk types. It extracts local joint feature sequences from multi-dimensional feature information; encodes the time intervals of multiple local spatiotemporal joint features using rotational position encoding to generate a position-enhanced feature sequence; inputs the position-enhanced feature sequence into a multi-head self-attention module to obtain attention features; and determines the risk type of the optical fiber segment based on the attention features. In this way, by extracting the coupling and correlation features of multiple physical quantities within a local spatiotemporal range from multi-dimensional feature information and identifying the evolution law of features over time, the accuracy of optical cable risk type identification in complex environments can be improved, thereby enhancing the reliability of optical cable risk early warning in complex environments.

[0122] The aforementioned fiber optic cable risk monitoring methods can be applied to the commissioning, expansion, and maintenance phases of optical networks. During the commissioning phase, baseline health assessments can be performed on the initial state of newly laid fiber optic cables to detect hidden damage such as micro-bending or stress concentration caused by improper construction, ensuring the reliability of the physical layer before network operation and reducing the probability of later failures. During network expansion, real-time security monitoring of newly added or adjusted fiber optic cable routes can be performed, promptly identifying external threats in complex construction environments, avoiding service interruptions due to expansion projects, and ensuring smooth network expansion and upgrades. In routine maintenance, a shift from "manual periodic inspections + emergency repairs after failures" to a "24 / 7 real-time awareness + proactive risk warning" maintenance model can be achieved. Automatic identification of various typical risks such as construction damage, fire, and icing, with high positioning accuracy and fast response speed, significantly shortens fault diagnosis time, improves service availability and maintenance efficiency, and reduces direct and indirect economic losses caused by fiber optic cable interruptions.

[0123] Furthermore, the optical cable risk monitoring method provided in this application has good universality. It can be applied not only to various types of solid optical fibers such as single-mode optical fiber, few-mode optical fiber, multi-mode optical fiber and multi-core optical fiber, but also to new hollow optical fibers such as hollow-core photonic bandgap optical fiber and hollow-core anti-resonant optical fiber. It can provide a unified and reliable security monitoring solution for various optical fiber communication scenarios.

[0124] Figure 11The diagram illustrates the hardware structure of the controller provided in this application embodiment. Referring to the diagram, at the hardware level, the controller 1100 includes a processor 1110, and optionally includes an internal bus 1120, a network interface 1130, and memory. The memory may include main memory 1141, such as high-speed random-access memory (RAM), and may also include non-volatile memory 1142, such as at least one disk storage device. Of course, the controller 1100 may also include other hardware required for other services.

[0125] The processor 1110, network interface 1130, and memory can be interconnected via an internal bus 1120. This internal bus 1120 can be an Advanced Microcontroller Bus Architecture (AMIC) bus, a Wishbone bus, an Open Core Protocol (OCP) bus, an Avalon bus, etc. The bus can be categorized as an address bus, data bus, control bus, etc. For ease of illustration, only a single bidirectional arrow is used in this diagram, but this does not imply that there is only one bus or one type of bus.

[0126] The memory stores programs. Specifically, the program may include program code, which includes computer operation instructions. The memory may include main memory 1141 and non-volatile memory 1142, and provides instructions and data to the processor 1110.

[0127] Processor 1110 reads the corresponding computer program from non-volatile memory 1142 into memory and then runs it, forming a device for locating the target user at the logical level. Processor 1110 executes the program stored in memory and specifically performs the following: Figure 7 or Figure 8 The methods disclosed in the embodiments shown achieve the functions and beneficial effects of the methods described in the preceding method embodiments, and will not be repeated here.

[0128] The above is as stated in this application. Figure 7 or Figure 8The methods disclosed in the illustrated embodiments can be applied to or implemented by processor 1110. Processor 1110 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by integrated logic circuits in the hardware of processor 1110 or by instructions in software form. The processor 1110 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0129] The computer device can also execute the methods described in the preceding method embodiments and achieve the functions and beneficial effects of the methods described in the preceding method embodiments, which will not be repeated here.

[0130] Of course, in addition to the software implementation, the controller 1100 of this application does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. In other words, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0131] This application also proposes a computer-readable storage medium that stores one or more programs, which, when executed by a controller comprising multiple applications, cause the controller to perform... Figure 7 or Figure 8 The methods disclosed in the embodiments shown achieve the functions and beneficial effects of the methods described in the preceding method embodiments, and will not be repeated here.

[0132] The computer-readable storage medium mentioned above includes read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, etc.

[0133] Furthermore, embodiments of this application also provide a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, which, when executed by a computer, implement the following process: Figure 7 or Figure 8 The methods disclosed in the embodiments shown achieve the functions and beneficial effects of the methods described in the preceding method embodiments, and will not be repeated here.

[0134] The embodiments of this application can be applied to various controller collaboration or interconnection scenarios, including: collaboration and interconnection between mobile phones and laptops / tablets; collaboration and interconnection between mobile terminals and smart TVs / monitors; collaboration and interconnection between mobile phones or tablets and in-vehicle entertainment systems; collaboration and interconnection between mobile terminals and smart conferencing systems, etc. This satisfies users' diverse needs in smart home, smart office, and smart travel scenarios.

[0135] In summary, the above description is merely a preferred embodiment of this application and does not limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

[0136] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0137] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0138] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0139] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

Claims

1. A fiber optic cable risk monitoring device, characterized in that, include: A multi-dimensional feature sensing circuit and a controller, wherein the multi-dimensional feature sensing circuit is communicatively connected to the controller; The multidimensional feature sensing circuit is connected to the optical fiber span and is used to collect multidimensional feature information of the optical fiber span; wherein, the optical cable includes multiple cascaded optical fiber spans, and each optical fiber span corresponds to a multidimensional feature sensing circuit; the multidimensional feature information includes at least two of the following: optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information. The controller is used to acquire the multi-dimensional feature information and determine the risk type of the optical fiber segment based on the multi-dimensional feature information.

2. The apparatus according to claim 1, characterized in that, The multi-dimensional feature sensing circuit includes: an optical signal processing unit and a multi-channel hardware sensing unit; The optical signal processing unit is connected to the optical fiber span and is used to obtain multiple backscattered optical signals of different types returned by the optical fiber span based on the probe optical signal by sending a probe optical signal to the optical fiber span. The multiple hardware sensing units, each equipped with different types of backscattered light signal processing functions, are connected in parallel. Each hardware sensing unit is used to acquire the corresponding type of backscattered light signal, process the backscattered light signal, and obtain multi-dimensional feature information of the optical fiber segment.

3. The apparatus according to claim 2, characterized in that, The multi-channel hardware sensing unit includes at least two of the following: fiber optic vibration sensing device, fiber optic stress sensing device, fiber optic temperature sensing device, and fiber optic loss sensing device. The fiber optic vibration sensing device is used to sense the vibration of the acquired Rayleigh backscattered light signal and obtain the fiber vibration information of the fiber span. The fiber optic stress sensing device is used to sense the stress of the acquired Brillouin backscattered light signal and obtain the stress information of the fiber optic segment. The fiber optic temperature sensing device is used to sense the temperature of the acquired Raman backscattered light signal to obtain the fiber temperature information of the fiber span. The fiber loss sensing device is used to sense the loss of the acquired Rayleigh backscattered light signal and obtain the fiber loss information of the fiber span.

4. The apparatus according to claim 2, characterized in that, The optical signal processing unit includes: a signal transceiver unit and a signal distribution unit; The signal transceiver unit is connected to the optical fiber span and is used to obtain the composite backscattered light signal returned by the optical fiber span based on the probe light signal by sending a probe light signal to the optical fiber span. The signal distribution unit is used to acquire the composite backscattered light signal and separate the composite backscattered light signal into multiple backscattered light signals of different types.

5. The apparatus according to claim 4, characterized in that, The signal transceiver unit includes a laser, a signal modulation unit, and an optical circulator; The laser is used to output continuous light; The signal modulation unit is used to modulate the continuous light into a probe light signal; The optical circulator includes a first port, a second port, and a third port; the optical circulator is used to acquire the probe light signal through the first port, send the probe light signal to the optical fiber span through the second port, receive the composite backscattered light signal, and send the composite backscattered light signal to the signal distribution unit through the third port.

6. The apparatus according to claim 5, characterized in that, In response to the inclusion of an optical fiber vibration sensing device in the multi-channel hardware sensing unit, the signal transceiver unit further includes an optical coupler, and the laser is connected to the signal modulation unit through the optical coupler; The optical coupler is used to split the acquired continuous light into signal light and local oscillator light; the signal modulation unit is used to modulate the acquired signal light into the probe light signal. The optical coupler is connected to the fiber optic vibration sensing device and is used to send the local oscillator light to the fiber optic vibration sensing device. The fiber optic vibration sensing device obtains fiber optic vibration information of the fiber optic segment based on the corresponding type of backscattered light signal and the local oscillator light.

7. The apparatus according to claim 6, characterized in that, In response to the inclusion of an optical fiber stress sensing device in the multi-channel hardware sensing unit, the optical coupler is connected to the optical fiber stress sensing device via a modulator. The modulator is used to modulate the local oscillator light and send the modulated local oscillator light to the fiber stress sensing device. The fiber stress sensing device obtains the stress information of the fiber span based on the backscattered light signal of the corresponding type and the modulated local oscillator light.

8. The apparatus according to claim 7, characterized in that, The optical coupler is connected to the fiber optic vibration sensing device and the modulator respectively via an optical switch component.

9. The apparatus according to claim 4, characterized in that, The signal distribution unit includes at least one of an optical coupler and an optical switch.

10. The apparatus according to any one of claims 2 to 9, characterized in that, In response to the in-band deployment of the connection between the multi-dimensional feature sensing circuit and the optical fiber span, the wavelength of the probe optical signal is located in a first band, which is the band used by the service signal of the optical fiber span; or... In response to the out-of-band deployment of the connection between the multi-dimensional feature sensing circuit and the optical fiber span, the wavelength of the detection optical signal is located in the second band, which is a band other than the band used by the service signal of the optical fiber span.

11. The apparatus according to claim 10, characterized in that, In response to the in-band deployment of the connection between the multi-dimensional feature sensing circuit and the optical fiber span, an optical filter is connected to the end of the optical fiber span to which the multi-dimensional feature sensing circuit is connected. The optical filter is used to block the detection light signal from entering the next level optical fiber span.

12. A method for monitoring the risk of optical cables, characterized in that, Applied to controllers, including: The multi-dimensional feature information of the optical fiber span connected to the multi-dimensional feature sensing circuit is obtained; wherein, the multi-dimensional feature information includes at least two of the following: optical fiber vibration information, optical fiber temperature information, optical fiber stress information, and optical fiber loss information; the optical cable includes multiple optical fiber spans, and each optical fiber span corresponds to a multi-dimensional feature sensing circuit. Based on the multi-dimensional feature information, the risk type of the optical fiber segment is determined.

13. The method according to claim 12, characterized in that, The determination of the risk type of the optical fiber segment based on the multi-dimensional feature information includes: Local joint feature sequences are extracted from the multi-dimensional feature information. The local joint feature sequences include multiple local spatiotemporal joint features. The local spatiotemporal joint features are coupled and correlated features of multi-dimensional feature information within a preset time and space range. The time intervals of the multiple local spatiotemporal joint features are encoded by rotational position encoding to generate a position-enhanced feature sequence; The location enhancement feature sequence is input into a multi-head self-attention module to obtain attention features; wherein, the attention features are used to indicate the temporal dependencies between multiple local spatiotemporal joint features; Based on the attention characteristics, the risk type of the fiber optic segment is determined.

14. The method according to claim 13, characterized in that, The step of extracting local joint feature sequences from the multi-dimensional feature information includes: The multi-dimensional feature information is input into the spatiotemporal feature extraction network to obtain the local joint feature sequence; wherein, the spatiotemporal feature extraction network includes at least one feature extraction layer, and each feature extraction layer is used to sequentially perform convolution, pooling and concatenation on the input data through the target convolution kernel to obtain the output data.

15. The method according to claim 14, characterized in that, The size of the target convolutional kernel is determined based on the number of dimensions of the multi-dimensional feature information and the number of sampling data points corresponding to the spatial resolution of the multi-dimensional feature perception circuit.

16. A controller, characterized in that, The controller includes a processor and a memory, the memory storing programs or instructions that can run on the processor, the programs or instructions being executed by the processor as steps of the method as described in any one of claims 12 to 15.

17. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions that, when executed by a computer, cause the computer to perform the steps of the method as described in any one of claims 12 to 15.

Images