Bridge falling early warning method and equipment based on OTDR (Optical Time Domain Reflectometer)

Abstract

The invention relates to the field of bridge structure safety monitoring, in particular to a bridge falling early warning method and device based on an OTDR. The distributed optical fiber sensing technology is adopted, continuous monitoring of a long-distance bridge group is achieved, specific disaster damage position coordinates are automatically inverted by monitoring link breakpoints in real time, and quantitative and accurate space information support is provided for rapid emergency rescue. Meanwhile, smooth noise reduction, tail end clutter noise area identification, peak detection based on significance difference and an OTDR distance measurement uncertainty model are introduced into OTDR curve processing, a grading early warning mechanism including tail end peak sudden change and middle section significant peak statistics is established, the positioning precision and the alarm efficiency of the fault position are greatly improved, and the fault positioning accuracy and the alarm efficiency are improved. And the requirement of carrying out large-scale and long-distance lightweight safety monitoring from a road section bridge group level is met.

Description

Technical Field

[0001] This invention relates to the field of bridge structural safety monitoring, and in particular to a bridge beam drop early warning method and device based on OTDR. Background Technology

[0002] In the event of a bridge beam collapse on a transportation route, in addition to damage to infrastructure, it can also lead to vehicles falling, or even multiple vehicles falling in succession, causing extremely serious safety accidents. However, current prevention measures for bridge beam collapses mostly focus on anti-collapse devices and their design, with less emphasis on monitoring and alarm systems for such events. Anti-collapse devices and designs primarily address the problem of main beam overturning when the substructure is safe and stable, offering limited protection against bridge collapses caused by substructure tilting or fracture. Once a bridge beam collapse occurs, conducting safety monitoring, promptly issuing alarms to following vehicles, and notifying bridge management and emergency response departments are crucial. This can minimize the risk of vehicles falling and allow for timely traffic control measures to prevent the accident from escalating.

[0003] Existing beam drop monitoring and alarm systems have several shortcomings. Monitoring methods based on mechanical triggering or circuit switching are simple in structure, but rely on the deployment of discrete components, making them difficult to deploy in long-distance or multi-bridge continuous monitoring scenarios. Point-based monitoring methods based on strain gauges, accelerometers, and displacement sensors can obtain local responses, but require a large number of devices, incur high power supply and communication costs, and lack spatial resolution, making it difficult to achieve continuous monitoring and accurate positioning of multi-span bridge structures. In addition, visual inspection methods are greatly affected by environmental factors such as lighting and weather, and are not adaptable to nighttime or inclement weather conditions.

[0004] With the development of distributed fiber optic sensing technology, optical time domain reflectometers (OTDRs) have been gradually introduced into the field of structural health monitoring due to their advantages such as long coverage distance, strong resistance to electromagnetic interference, and no need for on-site power supply. However, in bridge beam drop monitoring scenarios, traditional OTDR signal processing still faces several technical bottlenecks: fixed threshold algorithms have poor adaptability to optical path attenuation and low signal-to-noise ratio conditions, limited positioning accuracy, lack of ranging uncertainty calibration, high false alarm rate, and the system processing efficiency has not yet met the second-level response requirements in beam drop scenarios.

[0005] Therefore, there is a need for a bridge beam drop warning method and equipment that has higher alarm efficiency and more accurate beam drop positioning. Summary of the Invention

[0006] The purpose of this invention is to overcome the problems of existing beam drop monitoring and alarm systems, such as difficulty in locating the beam drop position and long alarm delay time, and to provide a bridge beam drop early warning method and device based on OTDR.

[0007] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0008] An OTDR-based bridge beam drop early warning method includes the following steps: S1: Distributed sensing optical fibers are continuously deployed on the main beam of the bridge to be warned, and the sampling sequence of the OTDR reflection signal of the bridge to be warned is acquired in real time. S2: Perform smoothing and noise reduction processing on the OTDR reflection signal sampling sequence to output a smooth signal; S3: Identify the end clutter noise region of the smoothed signal and output the effective signal range; S4: Perform adaptive saliency peak detection and terminal peak extraction processing on the effective signal interval; S5: Detect the position of the terminal peak and the significant peak respectively, and output the corresponding warning results; When the farthest possible location of the terminal peak is less than the nearest possible location detected previously, a Level 1 warning is issued; An anomaly warning is issued when the number of significant peaks exceeds a preset threshold.

[0009] As a preferred embodiment of the present invention, the smoothing and noise reduction process in S2 uses a one-dimensional Gaussian convolution kernel to smooth the OTDR reflection signal sampling sequence, and the expression of the convolution kernel is: , in, σ is a one-dimensional Gaussian convolution kernel, σ is the standard deviation, and i is the discrete offset of the convolution window.

[0010] As a preferred embodiment of the present invention, step S3 includes the following steps: S31: Calculate the local standard deviation sequence of the smoothed signal using a sliding window of a set length; S32: Find the global median standard deviation in the local standard deviation sequence; S33: Traverse each window. When L consecutive windows satisfy the condition that the local standard deviation is greater than k times the global median standard deviation, determine the starting coordinates of the first window that satisfies the condition as the starting point of the end clutter noise region, and determine the region after the starting point in the smoothed signal as the end clutter noise region. S34: Filter out the end noise region of the smoothed signal and output the effective signal range; Where L and k are preset values.

[0011] As a preferred embodiment of the present invention, the expression for the local standard deviation is: , in, Starting coordinates are x i The local standard deviation of the window. Starting coordinates are x j The smoothing signal of the window, The length of the sliding window. Indicates the starting coordinates as The local average value of the signal within the window.

[0012] As a preferred embodiment of the present invention, step S4 includes the following steps: S41: Traverse local maximum values ​​within the effective signal range; S42: Calculate the significance of each local maximum; When the significance of a local maximum is greater than a preset significance threshold, the local maximum is determined to be a significant peak. Wherein, the significance threshold = the overall standard deviation of the effective signal range × a set scaling factor; S43: Output each significant peak value, and output the last significant peak value in the effective signal interval as the terminal peak value.

[0013] As a preferred embodiment of the present invention, the expression for the significance of the local maximum value is: , Where P is the current local maximum value. The significance value, It is the minimum value to the left of the current local maximum value. It is the minimum value to the right of the current local maximum value.

[0014] As a preferred embodiment of the present invention, step S5 includes the following steps: The detection is performed based on the position of the terminal peak. When the farthest possible location detected this time is less than the nearest possible location detected in the previous time, a Level 1 warning is issued; Significant peak values ​​are detected, and when the number of significant peak values ​​exceeds a preset threshold, a secondary judgment is initiated. If the average significance amplitude of each significance peak is less than c times the significance threshold and occurs within a number of consecutive periods less than the set number, it is judged as a mild anomaly. If the average significance magnitude of each significance peak is less than c times the significance threshold and occurs within a set number of consecutive periods, it is judged as a mild to moderate anomaly. If the average significance magnitude of each significance peak is greater than c times the significance threshold and occurs within a number of consecutive periods less than the set number, it is judged as a mild to moderate anomaly. If the average significance magnitude of each significance peak is greater than c times the significance threshold and occurs within a set number of consecutive periods, it is determined to be a moderate anomaly.

[0015] As a preferred embodiment of the present invention, the expression for detecting the position of the terminal peak is: , i=1,2 in, The location of the terminal peak. This is the location of the previous detection. To account for the ranging uncertainty of the previous detection position, For the distance measurement uncertainty of the terminal peak position, Let be the measured distance of the i-th detection. The system fixed error is represented by k; k is the proportional error coefficient. This represents the sampling resolution error.

[0016] As a preferred embodiment of the present invention, the expression for the average significance magnitude is: , in, The number of significant peaks, It is the local maximum value of the j-th significance peak.

[0017] An OTDR-based bridge beam falling warning device includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform any of the aforementioned OTDR-based bridge beam falling warning methods.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention employs distributed fiber optic sensing technology to achieve continuous monitoring of long-distance bridge clusters. By monitoring link breakpoints in real time, it automatically derives the specific location coordinates of disaster damage, providing quantitative and precise spatial information support for rapid emergency rescue. Simultaneously, by introducing smoothing and noise reduction, end-point clutter noise region identification, peak detection based on significant difference, and an OTDR ranging uncertainty model into OTDR curve processing, this invention establishes a hierarchical early warning mechanism including end-point peak mutations and mid-section significant peak statistics. This significantly improves the location accuracy and alarm efficiency of faults, meeting the need for large-scale, long-distance lightweight safety monitoring at the road segment and bridge cluster level. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a bridge beam drop warning method based on OTDR as described in Embodiment 1 of the present invention. Figure 2 This is a schematic diagram of the arrangement of sensing optical fibers in a bridge beam falling early warning method based on OTDR as described in Embodiment 2 of the present invention. Figure 3 This is a simulation diagram of the peak value identification at the end of the OTDR reflection curve in the bridge beam falling early warning method based on OTDR described in Embodiment 2 of the present invention. Figure 4 This is a peak detection simulation diagram in the bridge beam drop early warning method based on OTDR described in Embodiment 2 of the present invention; Figure 5 This is a simulation diagram of significant peak identification in the middle segment of an optical fiber in an OTDR-based bridge beam drop early warning method according to Embodiment 2 of the present invention. Figure 6 This is a schematic diagram of a bridge beam falling warning device based on OTDR, which utilizes the bridge beam falling warning method based on OTDR described in the foregoing embodiments, as described in Embodiment 3 of the present invention. Detailed Implementation

[0020] The present invention will be further described in detail below with reference to experimental examples and specific embodiments. However, this should not be construed as limiting the scope of the above-mentioned subject matter of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0021] Example 1 like Figure 1 As shown, a bridge beam drop warning method based on OTDR includes the following steps: S1: Distributed sensing optical fibers are continuously deployed on the main beam of the bridge to be warned, and the OTDR reflection signal sampling sequence of the bridge to be warned is acquired in real time.

[0022] S2: Perform smoothing and noise reduction processing on the OTDR reflection signal sampling sequence to output a smooth signal.

[0023] S3: Identify the end clutter noise region of the smoothed signal and output the effective signal range.

[0024] S4: Perform adaptive saliency peak detection and terminal peak extraction processing on the effective signal interval.

[0025] S5: Detect the position of the terminal peak and the significant peak respectively, and output the corresponding warning results; When the farthest possible location of the terminal peak is less than the nearest possible location detected previously, a Level 1 warning is issued; An anomaly warning is issued when the number of significant peaks exceeds a preset threshold.

[0026] Example 2 This embodiment is a specific implementation of the bridge beam drop warning method based on OTDR described in Embodiment 1, including the following steps: S1: Distributed sensing optical fibers are continuously deployed on the main beam of the bridge to be warned, and the OTDR reflection signal sampling sequence of the bridge to be warned is acquired in real time.

[0027] In this embodiment, the OTDR reflection signal sampling sequence is represented by (x[n], y[n]), where x[n] represents the distance from the sampling point and y[n] represents the corresponding reflection power value.

[0028] Furthermore, this embodiment also performs a deep copy when inputting data to prevent memory reuse pollution and ensure data independence.

[0029] Fiber optic sensors, with their advantages of low cost, strong anti-interference capability, and high sensitivity, have been widely used in structural monitoring and safety assurance. In such applications, the breakage or damage of the fiber optic link is often related to the potential damage to the monitored structure and can serve as an important signal to trigger emergency alarms. Using fiber optics as the sensing device for beam drop monitoring can achieve meter-level positioning and second-level alarm for beam drop location. Therefore, such as Figure 2 As shown, in this embodiment, distributed sensing optical fibers are continuously deployed on the main beam of the bridge to be warned. A special device is used to fix them to the web of the main beam. When the main beam experiences a large lateral deviation or even falls, the optical signal on the optical fiber deployed on the main beam changes. The optical time domain reflectometer receives the real-time optical signal, analyzes and judges the event, and transmits the result to the central control edge. The central control edge converts the optical signal event into an engineering safety event through analysis and processing. Based on the processing result, it reports the alarm information to the bridge management and emergency departments. At the same time, the control information is transmitted to the audible and visual alarm device base station, and the base station controls the audible and visual alarm device to issue an alarm signal.

[0030] Optical Time Domain Reflectometry (OTDR), a common method for fiber optic detection and localization, works by emitting test laser pulses into the fiber optic link and analyzing Rayleigh scattering and Fresnel reflection signals generated along the way to obtain curve data reflecting the attenuation distribution of the fiber optic link. Common characteristic points in OTDR curves include reflection events, non-reflection events, saturated reflection events, and fiber ends. In fiber optic sensing applications, reflection events caused by fiber breaks are of particular interest. These events are usually accompanied by strong Fresnel reflections, with the curve showing a distinct peak and a signal interruption or sharp drop after the break point. Numerous research and patents have been developed for the automatic identification and localization of these characteristic events. Traditional methods often use a two-point method combined with least squares fitting to locate event points, but their accuracy is limited under noisy conditions. Chinese patent application CN107664571A proposes a multi-scale processing method based on wavelet analysis, which uses thresholding to extract high-frequency coefficients for event identification. International patent PCT / CN2014 / 081105 describes an online method for detecting events at the end of optical fibers, which can distinguish between reflective and non-reflective end events, thereby improving the detection accuracy of end breakages. However, this method is mainly aimed at the end of the fiber and still lacks an effective identification and location mechanism for breakage events in the middle of the link.

[0031] In summary, existing methods for OTDR event detection have the following problems: First, they lack an adaptive mechanism for signal fluctuation characteristics, threshold selection relies on experience, and they are highly dependent on parameters; second, they focus on detecting events at the end of the link, making it difficult to handle complex real-world scenarios such as multiple spikes in the middle of the link, leading to positioning errors; third, deep learning-based methods have high computational overhead in real-time monitoring and require labeled data or robust training sets to avoid overfitting.

[0032] Therefore, this embodiment, when using OTDR for positioning, still needs to address the problems of insufficient positioning accuracy, threshold parameter sensitivity, and poor adaptability to link differences inherent in traditional methods under low signal-to-noise ratio conditions. To address the noise sensitivity, numerous false peaks, and unstable positioning issues of traditional OTDR signal end detection algorithms, this embodiment proposes the following processing steps: Through a triple constraint mechanism of standard deviation dynamic threshold, neighborhood extreme value comparison, and end noise identification, it achieves automatic identification of true breakpoint peaks and maintains high-precision detection performance even in low signal-to-noise ratio environments.

[0033] S2: Perform smoothing and noise reduction processing on the OTDR reflection signal sampling sequence to output a smooth signal.

[0034] Furthermore, in this embodiment, while preserving the overall trend and event characteristics of the signal, a one-dimensional Gaussian convolution kernel is used to smooth the OTDR reflection signal sampling sequence in order to effectively suppress local small-amplitude noise and random fluctuations. The expression for the convolution kernel is: , in, σ is a one-dimensional Gaussian convolution kernel, σ is the standard deviation, and i is the discrete offset of the convolution window.

[0035] Smoothed signal for:

[0036] in, Let be the value of the signal at the nth sampling point after Gaussian convolution smoothing, and let i represent the offset index of the convolution window, ranging from [ , ] represents the effective convolution window; G(i) represents the weight of the Gaussian convolution kernel at offset i. This represents the normalization factor for the convolution kernel weights.

[0037] S3: Identify the end clutter noise region of the smoothed signal and output the effective signal range.

[0038] When the test light pulse from the OTDR reaches the end of the fiber, it encounters a highly reflective interface (usually an air interface or connector end face). This reflected light is extremely strong, causing the detector to momentarily "saturate," resulting in optical tailing and electronic overshoot. This manifests as a dense, high-frequency oscillating wave signal appearing after the end of the fiber. These signals do not correspond to actual optical reflections but are the response of the electronic system itself. Therefore, it is necessary to locate and filter these noise signals. The specific steps include: S31: Calculate the local standard deviation sequence of the smoothed signal using a sliding window of a set length; Furthermore, the expression for the local standard deviation is: , in, Starting coordinates are x i The local standard deviation of the window. Starting coordinates are x j The smoothing signal of the window, The length of the sliding window. Indicates the starting coordinates as The local average value of the signal within the window.

[0039] S32: Find the global median standard deviation in the local standard deviation sequence. ; S33: Traverse each window, and when L consecutive windows satisfy the condition that the local standard deviation is greater than k times the global median standard deviation (i.e., ... When the condition is met, the starting coordinates corresponding to the first window that meets the condition (i.e., The region after the starting point in the smoothed signal is determined as the end noise region; where L and k are preset values.

[0040] S34: Filter out the end noise region of the smoothed signal and output the effective signal range.

[0041] This step filters out abnormal fluctuations in the later section based on the clutter initiation position, retaining only the effective signal range in the earlier section, thereby achieving adaptive identification of the clutter region at the end of the OTDR curve and avoiding interference from false peaks.

[0042] S4: Perform adaptive saliency peak detection and terminal peak extraction processing on the effective signal interval.

[0043] S41: Traverse the local maximum value y[p] within the effective signal interval; the determination adjustment is as follows: y[p]>y[p-1]&y[p]>y[p+1].

[0044] S42: Calculate the significance of each local maximum; Wherein, the significance threshold = the overall standard deviation of the effective signal interval × a set scaling factor; specifically, the mathematical expression for the significance threshold T is: , in, The overall standard deviation is calculated based on the effective signal range. This is a preset scaling factor (generally 0.2-0.3). This step achieves adaptive adjustment of the threshold by linking the significance threshold to the signal's own fluctuation level, avoiding the problem of traditional fixed thresholds easily failing under different conditions.

[0045] The significance of the local maximum is expressed as follows: , Where P is the current local maximum value. The significance value, It is the minimum value to the left of the current local maximum value. It is the minimum value to the right of the current local maximum value.

[0046] When the significance P of a local maximum is greater than a preset significance threshold T, the local maximum is determined to be a significant peak.

[0047] S43: Output each significant peak value, and output the last significant peak value in the effective signal interval as the terminal peak value.

[0048] All detected peaks are arranged in order, and the last significant peak point (x) is selected. end ,y end () as the terminal event peak.

[0049] S5: Detect the position of the terminal peak and the significant peak respectively, and output the corresponding warning results.

[0050] This step, based on the OTDR ranging uncertainty model and the results of multi-peak event identification, constructs a graded early warning mechanism to distinguish different degrees of structural anomalies.

[0051] ① Detection is performed based on the position of the terminal peak; When the furthest possible location detected this time is less than the closest possible location detected in the previous time, a Level 1 warning is issued; this result indicates that the optical path length has shortened, possibly due to fiber breakage, main beam slippage, or beam collapse. At this point, the alarm device is triggered, the system enters a Level 1 alarm state, and the emergency management center is notified.

[0052] Furthermore, the expression for detecting the position of the terminal peak is: , in, The location of the terminal peak. This is the location of the previous detection. To account for the ranging uncertainty of the previous detection position, The uncertainty in the distance measurement of the position of the terminal peak. Since there is a distance measurement error in the OTDR system, it typically consists of three parts: (1) system fixed error; (2) cumulative error proportional to the distance; and (3) sampling resolution error. Therefore, the measurement distance... The total uncertainty is: i=1,2 in, Let be the measured distance of the i-th detection. The system fixed error is 0.75m in this embodiment; k is the proportional error coefficient (5×10 in this embodiment). 5 ); This represents the sampling resolution error, which is determined by the OTDR sampling accuracy.

[0053] ② Detect significant peaks; when the number of significant peaks... If the value is greater than the preset threshold, it indicates that there are multiple protruding peaks or reflection events in the middle section of the optical fiber, which may correspond to local damage, partial breakage or abnormal reflection. Therefore, it is necessary to calculate the average significance amplitude and proceed to a secondary judgment. When the average significance magnitude of each significance peak is less than c times the significance threshold (i.e.) Where c is the amplification factor, which can be adjusted offline or online using an adaptive algorithm according to the fiber type, installation conditions and on-site noise level in practical applications. If it occurs within a continuous period less than the set number, it is judged as a minor anomaly (prompting maintenance checks and log recording). If the average significance magnitude of each significance peak is less than c times the significance threshold and occurs within a set number of consecutive periods, it is determined to be a mild to moderate anomaly (the anomaly lasts for a long time and may have a cumulative impact on the monitored object).

[0054] When the average significance magnitude of each significance peak is greater than c times the significance threshold (i.e.) If it occurs within a continuous period less than the set number, it may be an instantaneous impact anomaly and is judged as a mild to moderate anomaly (it can be used as a warning target, but it will not be directly classified as a moderate anomaly).

[0055] If the average significance amplitude of each significant peak is greater than c times the aforementioned significance threshold, and occurs within a set number of consecutive periods, it is judged as a moderate anomaly. Staff should be alerted that the structure may be experiencing slippage or fatigue propagation, and an on-site verification should be conducted.

[0056] Furthermore, the expression for the average significance magnitude is: , in, The number of significant peaks, It is the local maximum value of the j-th significance peak.

[0057] The following are the test results of this embodiment using a sampling frequency of 1 Hz: like Figure 3 As shown, in the 35th test, the peak position of the OTDR curve at the end was 526.081 m; in the 36th test, the peak position suddenly shifted forward to 524.050 m. According to the distance measurement uncertainty model proposed in this invention, the system automatically determines that the two measurement results meet the uncertainty threshold condition, that is, the farthest possible position detected in the latest test is less than the nearest possible position detected in the previous test, and determines it as a real fracture event, triggering the alarm mechanism.

[0058] like Figure 4As shown, the changes in the terminal peak value and signal processing flow are shown in the 35th and 36th detections. The total time for the entire event detection and alarm linkage is only 53 ms.

[0059] like Figure 5 As shown, it demonstrates the identification of significant peaks in the middle section of the optical fiber, with red circles representing the locations of the identified peaks.

[0060] The experimental results above demonstrate that the method of this invention can achieve rapid, stable, and accurate breakpoint identification and alarm within one detection cycle after an optical fiber breakage, verifying the algorithm's superior performance in terms of real-time performance and reliability. These results also indicate that the ranging uncertainty model can effectively eliminate false alarms caused by inherent OTDR losses or signal fluctuations, ensuring the reliability of alarm judgments and the engineering stability of the system.

[0061] In summary, the bridge beam drop warning method based on OTDR described in this invention has the following advantages: 1. Wide monitoring range and high positioning accuracy, overcoming the limitations of mechanical methods: This invention employs distributed optical fiber sensing technology, enabling continuous monitoring coverage of 2 km to 5 km with a single optical fiber and a spatial positioning resolution of 1 m ± 0.125 m. Compared to traditional mechanical trigger sensors or segmented displacement gauges, this invention not only achieves continuous monitoring of long-distance bridge groups but also automatically retrieves the specific coordinates of disaster damage after detecting link breakpoints, providing quantitative and precise spatial information support for rapid disaster relief and rescue.

[0062] 2. Adaptable to long-distance mountainous bridge scenarios, reducing construction and maintenance costs: For scenarios where traditional monitoring methods are difficult to implement, such as mountainous areas or long, multi-span bridges, this invention requires only a single optical fiber to achieve continuous monitoring of multiple bridges. Compared to an equivalent mechanical monitoring system, it reduces the number of sensor units and some electrical wiring by more than 50% at the same monitoring distance, significantly lowering the overall system cost and reducing subsequent maintenance workload.

[0063] 3. Lightweight deployment, simple construction, and strong environmental adaptability: The fiber optic deployment structure is compact, requiring no large machinery or complex support structures during installation, making it suitable for complex terrains such as mountainous areas, high embankments, or river valleys. The system exhibits good weather resistance and strong electromagnetic interference resistance, supporting long-term maintenance-free operation.

[0064] 4. Algorithm Innovation: Combining Noise Adaptation with Significance Peak Detection This invention introduces Gaussian smoothing, adaptive threshold setting, and a peak detection mechanism based on significant difference into OTDR curve processing. This allows the threshold to dynamically adjust with the signal noise level, achieving an integrated signal processing framework of "threshold self-adjustment—noise suppression—real event enhancement." By significantly differentiating the difference between the peak value and the valleys on both sides, a peak-valley difference judgment model is established. This effectively distinguishes real reflection spikes from small fluctuations or background noise, significantly improving the event recognition accuracy in low signal-to-noise ratio environments.

[0065] 5. Significant improvement in real-time performance and computational efficiency: The core computation of this invention includes only one Gaussian convolution (linear complexity O(N)), standard deviation calculation, and local extremum search, with an overall time complexity of O(N). Compared with traditional wavelet multi-scale decomposition methods (O(N log N)) and deep learning methods, the computational load is reduced by approximately 3 to 5 times. Processing 50,000 points of OTDR data on a typical PC platform (Intel® Core™ i5-13500) requires only about 80 ms, far less than the 1-second monitoring sampling cycle, which can meet the real-time response requirements for bridge online monitoring and beam drop warning.

[0066] 6. The hierarchical early warning mechanism based on the uncertainty model improves system reliability: This invention introduces a ranging uncertainty correction model for the first time in OTDR beam drop monitoring. This model is used to correct for changes in peak values ​​at the end of the beam during continuous monitoring, enabling the system to robustly distinguish between actual structural changes and measurement errors even under conditions of signal attenuation, reflection loss, or environmental disturbances. This eliminates false alarms and ensures the reliability and accuracy of alarm judgments. Simultaneously, the system can automatically distinguish different abnormal states based on event location, peak intensity, and trend. This improvement addresses the engineering pain points of high false alarm rates and numerous duplicate alarms in current fiber optic monitoring systems.

[0067] Example 3 like Figure 6 As shown, an OTDR-based bridge beam falling warning device includes at least one processor, a memory communicatively connected to the at least one processor, and at least one input / output interface communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed, enables the at least one processor to perform the OTDR-based bridge beam falling warning method described in the foregoing embodiments. The input / output interface may include a display, keyboard, mouse, and USB interface for inputting and outputting data.

[0068] Furthermore, the OTDR-based bridge beam falling warning device can be a desktop computer, mobile phone, tablet computer, wearable OTDR-based bridge beam falling warning device, or any other OTDR-based bridge beam falling warning device capable of deep information recognition.

[0069] Furthermore, the processor may include one or more processing cores. The processor connects to various parts of the OTDR-based bridge beam fall warning device using various interfaces and lines. It executes various functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory, and by calling data stored in memory. Optionally, the processor may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor and may be implemented separately through a communication chip.

[0070] The memory may include random access memory (RAM) or read-only memory (ROM). The memory can be used to store instructions, programs, code, code sets, or instruction sets, such as instructions or code sets used to implement the OTDR-based bridge beam drop warning method provided in this application embodiment. The memory may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described above, etc. The data storage area may also store data created during the use of the OTDR-based bridge beam drop warning device (such as a modulation sequence-depth mapping table, image data, spectrogram data, etc.).

[0071] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0072] When the integrated units of the present invention are implemented as software functional units and sold or used as independent products, they can also be stored in a computer-readable storage medium. The computer-readable storage medium stores program code, which can be called by a processor to execute the methods described in the above method embodiments. Based on this understanding, the technical solution of the embodiments of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes electronic memories such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Optionally, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium has storage space for program code that executes any of the method steps described above. This program code can be read from or written to one or more computer program products. The program code can be compressed, for example, in an appropriate form.

[0073] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A bridge beam drop early warning method based on OTDR, characterized in that, Includes the following steps: S1: Distributed sensing optical fibers are continuously deployed on the main beam of the bridge to be warned, and the sampling sequence of the OTDR reflection signal of the bridge to be warned is acquired in real time. S2: Perform smoothing and noise reduction processing on the OTDR reflection signal sampling sequence to output a smooth signal; S3: Identify the end clutter noise region of the smoothed signal and output the effective signal range; S4: Perform adaptive saliency peak detection and terminal peak extraction processing on the effective signal interval; S5: Detect the position of the terminal peak and the significant peak respectively, and output the corresponding warning results; When the farthest possible location of the terminal peak is less than the nearest possible location detected previously, a Level 1 warning is issued; An anomaly warning is issued when the number of significant peaks exceeds a preset threshold.

2. The bridge beam drop early warning method based on OTDR according to claim 1, characterized in that, The smoothing and noise reduction process in S2 uses a one-dimensional Gaussian convolution kernel to smooth the OTDR reflection signal sampling sequence. The expression for the convolution kernel is: ， in, σ is a one-dimensional Gaussian convolution kernel, σ is the standard deviation, and i is the discrete offset of the convolution window.

3. The bridge beam drop early warning method based on OTDR according to claim 1, characterized in that, S3 includes the following steps: S31: Calculate the local standard deviation sequence of the smoothed signal using a sliding window of a set length; S32: Find the global median standard deviation in the local standard deviation sequence; S33: Traverse each window. When L consecutive windows satisfy the condition that the local standard deviation is greater than k times the global median standard deviation, determine the starting coordinates of the first window that satisfies the condition as the starting point of the end clutter noise region, and determine the region after the starting point in the smoothed signal as the end clutter noise region. S34: Filter out the end noise region of the smoothed signal and output the effective signal range; Where L and k are preset values.

4. The bridge beam drop early warning method based on OTDR according to claim 3, characterized in that, The expression for the local standard deviation is: ， in, Starting coordinates are x i The local standard deviation of the window. Starting coordinates are x j The smooth signal of the window, The length of the sliding window. Indicates the starting coordinates as The local average value of the signal within the window.

5. The bridge beam drop early warning method based on OTDR according to claim 1, characterized in that, S4 includes the following steps: S41: Traverse local maximum values ​​within the effective signal range; S42: Calculate the significance of each local maximum; When the significance of a local maximum is greater than a preset significance threshold, the local maximum is determined to be a significant peak. Wherein, the significance threshold = the overall standard deviation of the effective signal range × a set scaling factor; S43: Output each significant peak value, and output the last significant peak value in the effective signal interval as the terminal peak value.

6. The bridge beam drop early warning method based on OTDR according to claim 5, characterized in that, The expression for the significance of the local maximum is: ， Where P is the current local maximum value. The significance value, It is the minimum value to the left of the current local maximum value. It is the minimum value to the right of the current local maximum value.

7. The bridge beam drop early warning method based on OTDR according to claim 6, characterized in that, S5 includes the following steps: The detection is performed based on the position of the terminal peak. When the farthest possible location detected this time is less than the nearest possible location detected in the previous time, a Level 1 warning is issued; Significant peak values ​​are detected, and when the number of significant peak values ​​exceeds a preset threshold, a secondary judgment is initiated. If the average significance amplitude of each significance peak is less than c times the significance threshold and occurs within a number of consecutive periods less than the set number, it is judged as a mild anomaly. If the average significance magnitude of each significance peak is less than c times the significance threshold and occurs within a set number of consecutive periods, it is judged as a mild to moderate anomaly. If the average significance magnitude of each significance peak is greater than c times the significance threshold and occurs within a number of consecutive periods less than the set number, it is judged as a mild to moderate anomaly. If the average significance magnitude of each significance peak is greater than c times the significance threshold and occurs within a set number of consecutive periods, it is determined to be a moderate anomaly.

8. The bridge beam drop early warning method based on OTDR according to claim 7, characterized in that, The expression for detecting the position of the terminal peak is: ， ，i=1,2， in, The location of the terminal peak. This is the location of the previous detection. To account for the ranging uncertainty of the previous detection position, The ranging uncertainty is due to the location of the terminal peak. Let be the measured distance of the i-th detection. The system fixed error is represented by k; k is the proportional error coefficient. This represents the sampling resolution error.

9. A bridge beam drop early warning method based on OTDR according to claim 7, characterized in that, The expression for the average significance magnitude is: ， in, The number of significant peaks, It is the local maximum value of the j-th significance peak.

10. A bridge beam drop warning device based on OTDR, characterized in that, The method includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform an OTDR-based bridge beam drop warning method according to any one of claims 1 to 9.

Images