Joint error dynamic correction method based on optical fiber and wireless sensing co-judgment event

Abstract

The application discloses a joint error dynamic correction method based on optical fiber and wireless sensing co-judgment events, relates to the technical field of distributed optical fiber sensing, and comprises the following steps: associating DAS vibration events and wireless network acoustic events through high-precision time synchronization, generating co-judgment event instances, defining to-be-estimated error parameters, establishing a joint observation model, jointly estimating the error parameters through an iterative reweighted least square algorithm, performing real-time dynamic correction on the events by using a final error parameter estimation, and obtaining a result through weighted averaging. Through joint modeling and real-time compensation on the positioning errors of a DAS host and a wireless sensing network, the application makes the fusion positioning result more approximate to the real geographical coordinates of the disturbance events, overcomes the limitations of traditional methods, improves the positioning accuracy in complex environments, and further adopts an iterative reweighted least square algorithm and a forgetting factor mechanism to enhance the robustness to abnormal events and environmental changes, so that long-term stable adaptive optimization is realized.

Description

Technical Field

[0001] This invention relates to the field of distributed optical fiber sensing technology, specifically a joint error dynamic correction method based on the joint judgment of events by optical fiber and wireless sensing. Background Technology

[0002] Fiber optic distributed acoustic sensing (DAS) technology can continuously locate and identify vibration events along sensing optical fibers, while wireless acoustic sensor networks can monitor sound sources over a region. To improve monitoring performance, existing technologies propose fusion monitoring schemes that combine the two. In existing technologies, high-precision time synchronization techniques are typically used to perform spatiotemporal correlation matching between vibration events detected by the DAS and acoustic events detected by the wireless acoustic sensor network. When the same disturbance event is identified, a typical approach is to fuse the geographical coordinates obtained by the DAS through static mapping with the coordinates estimated by the wireless acoustic sensor network through acoustic localization algorithms, for example, through direct comparison or weighted averaging.

[0003] However, this technical solution has significant drawbacks: the fusion process is essentially a shallow fusion at the post-association and data layers, failing to fundamentally address the inherent and environmentally variable system errors of the two independent sensing systems. Specifically, the positioning error of distributed acoustic sensing mainly stems from the expansion and contraction of optical fiber due to temperature and stress changes, and the inability of the preset static optical fiber-geographic coordinate mapping table to accurately reflect complex routing. The positioning error of wireless acoustic sensing networks mainly arises from the changes in the speed of sound propagation in air due to temperature, humidity, and wind speed, as well as environmental interference such as multipath effects and non-line-of-sight propagation. Because these system errors are not effectively modeled, estimated online, and compensated for, the existing technical solution offers limited improvement in fusion positioning accuracy when dealing with complex signal propagation attenuation and time delay variations in real-world environments. The system's long-term stability and environmental adaptability are insufficient, making it difficult to achieve stable and reliable high-precision positioning.

[0004] Based on this, a joint error dynamic correction method based on the joint judgment of events by optical fiber and wireless sensing is now provided, which can eliminate the drawbacks of existing technical solutions. Summary of the Invention

[0005] The purpose of this invention is to provide a joint error dynamic correction method based on the joint judgment of events by optical fiber and wireless sensing, so as to solve the problem that existing solutions in the background art have limited improvement in fusion positioning accuracy, insufficient long-term stability and environmental adaptability when dealing with complex signal propagation attenuation and time delay changes in real environment, which makes it difficult to achieve stable and reliable high-precision positioning.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The joint error dynamic correction method based on the joint event judgment of fiber optic and wireless sensors specifically includes the following steps:

[0008] Step S1: The DAS host outputs a vibration event list. Each DAS event in the vibration event list includes the fiber optic location and the DAS local timestamp. The wireless sensor network synchronously outputs an acoustic event list. Each wireless event in the acoustic event list includes the preliminary positioning coordinates and the wireless network local timestamp. The central processing unit uses the time reference provided by the high-precision synchronous clock source to uniformly convert the timestamps of the two data streams to the absolute time coordinate system.

[0009] Step S2: For each DAS event, map its fiber location to estimated geographic coordinates through the fiber optic geographic information database, search for wireless events within a preset time window, calculate the spatial distance for each wireless event, and if the spatial distance is less than a preset spatial association threshold, determine that the DAS event and the wireless event are two observations of the same physical event, generate a co-judgment event instance and store it in the co-judgment event pool.

[0010] Step S3: Define the error parameters to be estimated. The error parameters include the DAS error parameter vector and the wireless network error parameter vector. Based on the error parameters, observation data pairs, and fiber optic locations, establish a joint observation model and obtain the residual equation by eliminating the real geographical location parameters.

[0011] Step S4: Using the iterative reweighted least squares algorithm, initialize the error function model, the number of iterations and the initial weights, perform a first-order Taylor expansion on the residual equation to construct a linearized incremental equation, solve the weighted normal equation to complete the parameter iterative update, and update the event weights according to the iterative residuals. Repeat the iteration until the convergence condition is met, and output the final error parameter estimate.

[0012] Step S5: Based on the final error parameter estimation, the positioning output results of DAS events and wireless events are dynamically corrected in real time to obtain the corrected coordinates. The confidence level is determined based on the historical residuals of the corrected coordinates, and the final fusion positioning result of the co-judgment events is obtained by weighted averaging.

[0013] Step S6: Maintain a fixed-size sliding window to retain the latest co-judgment event data. When the latest co-judgment event enters the window, use a recursive least squares algorithm combined with a forgetting factor to update the error parameter estimate. Adjust the forgetting factor according to the covariance trace or residual sequence of the error parameter estimate to achieve online adaptive optimization of the parameters.

[0014] Furthermore, before performing step S1, the joint error dynamic correction method also includes an initialization and optical cable identification phase:

[0015] A specially coded vibration signal is applied to the target optical cable at a known geographic coordinate point using a micro-perturbation device;

[0016] The DAS host detects and records the fiber location and vibration characteristics of each vibration event.

[0017] The wireless sensor network synchronously detects and records the preliminary location coordinates of each acoustic event;

[0018] The central processing unit spatiotemporally correlates DAS events and wireless events from the same physical stimulus, confirms the co-judgment relationship, and establishes a reference mapping library of optical cable identity ID, optical fiber location, and measured coordinates of the wireless sensor network for optical cable identification during online monitoring.

[0019] Furthermore, the DAS host in step S1 is a distributed acoustic wave sensing host. The DAS host obtains the fiber location and DAS local timestamp of the vibration event list by processing the backscattered Rayleigh signal in real time and by phase demodulation and event detection.

[0020] Furthermore, the preset time window in step S2 is a symmetrical interval before and after the absolute time of the DAS event, and the spatial distance is obtained by calculating the Euclidean distance between the estimated geographic coordinates and the preliminary location coordinates of the wireless event, expressed as: ,in, For spatial distance, To estimate geographic coordinates, These are the initial coordinates for locating the wireless event.

[0021] Furthermore, the joint observation model in step S3 satisfies:

[0022] ;

[0023] ;

[0024] in, and For parameterized error functions, This refers to random observation noise, representing the overall concept of random observation noise, specifically comprising two components from different sources: DAS observation noise and wireless network observation noise. This is the DAS error parameter vector. This is the wireless network error parameter vector. For the first event in the co-judgment event pool The actual geographical location of the event For fiber optic location, and For a co-determined event instance The observation data is relevant to information. For the first DAS observation noise for each event, For the first Wireless network observation noise for each event;

[0025] The expression for the residual equation is:

[0026] ;

[0027] in, The residual equations are represented as the DAS error parameter vector and the wireless network error parameter vector.

[0028] Furthermore, the DAS error parameter vector in step S3 is used to correct the mapping deviation from the fiber location to the geographic coordinates, and the wireless network error parameter vector is used to correct the positioning deviation caused by the deviation in sound wave propagation speed.

[0029] Furthermore, the error function model in step S4 includes a linear drift model on the fiber optic sensing side and a regional translation model on the wireless sensing side.

[0030] The expression for the linear drift model is: ,in, , This is a fixed offset vector in the DAS error. This is the vector of linear drift coefficients in the DAS error. The fiber location for the DAS event;

[0031] The expression for the regional translation model is: ,in, , This is the regional translation vector in the wireless network error.

[0032] Furthermore, the iterative reweighted least squares algorithm in step S4 uses the Huber weight function and the Tukey weight function to update the event weights, thereby reducing the impact of abnormal co-judgment events on the error parameter estimation results.

[0033] Furthermore, the formula for calculating the corrected coordinates in step S5 is as follows:

[0034] For newly arrived DAS events, the corrected coordinates are: ;

[0035] For newly arrived wireless events, the corrected coordinates are: ;

[0036] in, The coordinates after DAS event correction. The coordinates are corrected for the wireless event. For the final error parameter estimation of DAS, This is for estimating the final error parameters of the wireless network.

[0037] Furthermore, the update formula for the error parameter estimation in step S6 includes:

[0038] ;

[0039] ;

[0040] in, It is the regression vector corresponding to the new event. It is the observation difference of the new event. For the updated error parameter estimation, For the error parameter estimation before the update, It is a forgetting factor and satisfies , For the updated covariance, The covariance before the update. This represents the number of iterations.

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0042] This invention constructs a joint observation equation that simultaneously reflects the positioning errors of DAS and wireless sensor networks for each successfully spatiotemporally correlated co-judgment event. By accumulating multiple such events, a unified parameterized error model is established and solved online to jointly estimate the mapping deviation parameter of the former and the acoustic propagation deviation parameter of the latter. The original observation coordinates of new events are dynamically compensated using the real-time estimated parameters, and then high-precision fusion positioning is performed. This principle suppresses systematic bias, making the fusion positioning result closer to the true geographical coordinates of the disturbed event, thus improving positioning accuracy. This overcomes the limitations of traditional fusion methods that only perform shallow data fusion, fundamentally improving positioning accuracy in complex environments. Furthermore, this invention employs an iterative reweighted least squares algorithm and a forgetting factor mechanism to enhance the system's robustness to abnormal events and environmental changes, achieving long-term stable adaptive optimization. Attached Figure Description

[0043] Figure 1 This is a schematic diagram illustrating the steps of the joint error dynamic correction method of the present invention.

[0044] Figure 2 This is a flowchart illustrating the joint error dynamic correction method of the present invention.

[0045] Figure 3 This is a schematic diagram of the overall process of the present invention.

[0046] Figure 4 This is a schematic diagram of the joint error dynamic correction system of the present invention.

[0047] Figure label annotations: DAS host 10, wireless sensor network 20, high-precision synchronous clock source 30, central processing unit 40, monitoring terminal 50. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0049] To address the shortcomings of existing fusion monitoring schemes, this invention provides a joint error dynamic correction method based on co-judgment events of fiber optic and wireless sensors. Its core objective is to fundamentally improve upon the simple post-data fusion mode in existing technologies. By constructing an error function model and utilizing the co-judgment events themselves as observation data, it jointly estimates and dynamically compensates for the positioning errors of both types of sensing systems online. This overcomes the problems of positioning drift in distributed acoustic wave sensing systems caused by changes in the physical state of optical fibers and inaccurate static mapping, and positioning errors in wireless acoustic wave sensing networks caused by changes in the acoustic wave propagation environment. By using co-judgment event information for dynamic estimation of error parameters, it achieves real-time correction of DAS positioning errors and wireless network positioning errors, improving the fusion positioning accuracy, long-term stability, and overall reliability of asset disturbance events such as those involving optical cables in complex environments.

[0050] The implementation of this method is based on a hardware system comprising a DAS host 10, a wireless sensor network 20, a high-precision synchronous clock source 30, a central processing unit 40, and a monitoring terminal 50, wherein, for example... Figure 4 As shown, the central processing unit 40 is configured to execute instructions stored in its memory to run the following method flow.

[0051] Example 1

[0052] In this embodiment, wherein, as Figures 1 to 3 As shown, the joint error dynamic correction method based on the joint judgment of events by optical fiber and wireless sensing specifically includes the following steps:

[0053] Step S1: The DAS host 10 outputs a vibration event list. Each DAS event in the vibration event list includes the fiber optic location and the DAS local timestamp. The wireless sensor network 20 synchronously outputs an acoustic event list. Each wireless event in the acoustic event list includes the preliminary positioning coordinates and the wireless network local timestamp. The central processing unit 40 uses the time reference provided by the high-precision synchronous clock source 30 to uniformly convert the timestamps of the two data streams to the absolute time coordinate system. The output of this step is two asynchronous event streams with synchronized time.

[0054] Specifically, the fiber location is represented as The DAS local timestamp is represented as The initial location coordinates of the wireless event are represented as follows: The local timestamp of the wireless network is represented as The absolute time coordinate system is represented as ;

[0055] Step S2: For each DAS event, map its fiber location to estimated geographic coordinates through the fiber optic geographic information database (GIS database), search for wireless events within a preset time window, calculate the spatial distance for each wireless event, and if the spatial distance is less than a preset spatial association threshold, determine that the DAS event and the wireless event are two observations of the same physical event, generate a co-judgment event instance and store it in the co-judgment event pool. A co-judgment event refers to the same physical event that is detected and spatiotemporally associated by both DAS and wireless network.

[0056] Specifically, a DAS event is represented as The estimated geographic coordinates are expressed as The preset time window is Wireless events are represented as The preset spatial correlation threshold is expressed as Determine DAS events and wireless events ( and A common-determined event instance is generated after two observations of the same physical event. and its observation data to and the corresponding fiber optic locations Store in a pool of jointly judged events;

[0057] Step S3: Define the error parameters to be estimated. The error parameters include the DAS error parameter vector and the wireless network error parameter vector. Based on the error parameters, observation data pairs, and fiber optic locations, establish a joint observation model and obtain the residual equation by eliminating the real geographical location parameters.

[0058] Step S4: Using the iterative reweighted least squares algorithm, initialize the error function model, the number of iterations and the initial weights, perform a first-order Taylor expansion on the residual equation to construct a linearized incremental equation, solve the weighted normal equation to complete the parameter iterative update, and update the event weights according to the iterative residuals. Repeat the iteration until the convergence condition is met, and output the final error parameter estimate.

[0059] Step S5: Based on the final error parameter estimation, the positioning output results of DAS events and wireless events are dynamically corrected in real time to obtain the corrected coordinates. The confidence level is determined based on the historical residuals of the corrected coordinates, and the final fusion positioning result of the co-judgment events is obtained by weighted averaging.

[0060] Step S6: Maintain a fixed-size sliding window to retain the latest co-judgment event data. When the latest co-judgment event enters the window, use a recursive least squares algorithm combined with a forgetting factor to update the error parameter estimate. Adjust the forgetting factor according to the covariance trace or residual sequence of the error parameter estimate to achieve online adaptive optimization of the parameters. The sliding window is used to store the data structure of the latest co-judgment event.

[0061] Specifically, the principle of this method is as follows: the DAS positioning error (mainly caused by fiber-to-geographic coordinate mapping distortion) and the wireless network positioning error (mainly caused by environmental sound speed deviation) are parameterized, and each co-judgment event is regarded as a joint observation equation for these two sets of error parameters. Through the continuously accumulated co-judgment event data, a set of equations is constructed, and an algorithm is used to solve them jointly, thereby realizing online estimation and dynamic compensation of positioning error.

[0062] Specifically, in step S1, the DAS host 10 is a distributed acoustic wave sensing host. The DAS host 10 obtains the fiber location and DAS local timestamp of the vibration event list by processing the back Rayleigh scattering signal in real time and by phase demodulation and event detection.

[0063] Among them, such as Figure 2 As shown, the preset time window in step S2 is a symmetrical interval before and after the absolute time of the DAS event. The spatial distance is obtained by calculating the Euclidean distance between the estimated geographic coordinates and the preliminary location coordinates of the wireless event, and the expression is: ,in, For spatial distance, To estimate geographic coordinates, These are the initial coordinates for locating the wireless event.

[0064] Specifically, step S3 defines the error parameters to be estimated by mathematically modeling the two types of systematic errors:

[0065] DAS error parameter vector It is used to correct the mapping deviation from fiber optic location to geographic coordinates;

[0066] Wireless Network Error Parameter Vector It is used to correct positioning errors caused by deviations in the speed of sound propagation;

[0067] Establish a joint observation model for the first event in the co-judgment event pool. An event, assuming its real geographical location is... Then we have:

[0068] ;

[0069] ;

[0070] in, and For parameterized error functions, This refers to random observation noise, representing the overall concept of random observation noise, specifically comprising two components from different sources: DAS observation noise and wireless network observation noise. This is the DAS error parameter vector. This is the wireless network error parameter vector. For the first event in the co-judgment event pool The actual geographical location of the event For fiber optic location, and For a co-determined event instance The observation data is relevant to information. For the first DAS observation noise for each event, For the first Wireless network observation noise for each event;

[0071] Subtract the two equations above to eliminate the unknown true geographical location. The residual equation is obtained, and its expression is:

[0072] ;

[0073] in, The residual equations are represented as the DAS error parameter vector and the wireless network error parameter vector.

[0074] Among them, such as Figure 2 As shown, step S4 is used to solve the problem constructed in step S3 and find the error parameters. To minimize the sum of squared residuals for all co-judgment events, the Iterative Reweighted Least Squares (IRLS) algorithm is employed. This algorithm is an optimization of the classical least squares (LS) algorithm, used to enhance its robustness to anomalous co-judgment events. The specific steps are as follows:

[0075] An error function model is defined, which includes a linear drift model on the fiber optic sensing side and a regional translation model on the wireless sensing side. For example, assuming the DAS error is mainly linear drift along the fiber optic path, then it is a linear drift model, and its expression is: ,in, , This is a fixed offset vector in the DAS error. This is the vector of linear drift coefficients in the DAS error. Let the fiber location of the DAS event be denoted by . Assuming the wireless error is a regional translation, then the regional translation model is given by: ,in, , Let be the regional translation vector in the wireless network error, and let the iteration number be... Initial weights ;

[0076] In the In the next iteration, for the co-judgment event pool The event will affect the residual equation at the current error parameter estimate. Perform a first-order Taylor expansion to construct a linearized incremental equation. ,in, For Jacobian matrices, It is the residual vector;

[0077] Perform a weighted solution operation: solve the weighted normal equation. ,in, It is a diagonal weight matrix, whose diagonal elements The parameter update for this iteration is represented as ;

[0078] Perform a weight update operation: based on the residual after this iteration. Update the weight of each event. The iterative reweighted least squares algorithm uses Huber and Tukey weight functions to update event weights, reducing the impact of common anomalies on the error parameter estimation results. For example, the Huber weight function can be used.

[0079] ;

[0080] in, It is an estimate of the standard deviation of the residuals. As an adjustment constant, it is usually taken as 1.345;

[0081] Alternatively, the Tukey weight function can be used:

[0082] ;

[0083] in, As an adjustment constant, it is usually taken as 4.685;

[0084] This operation can automatically reduce the weight of outlier observations with large residuals, thus overcoming the drawback of standard least squares being sensitive to outliers.

[0085] like If the number of iterations is less than the set threshold or the maximum number of iterations is reached, the iteration stops, and the final error parameter estimate is output. Otherwise, let Continue performing the above operations;

[0086] Among them, such as Figure 2As shown, step S5 is used to estimate the final error parameters estimated in step S4. This will correct all subsequent positioning outputs.

[0087] The formula for calculating the corrected coordinates is:

[0088] For newly arrived DAS events, the corrected coordinates are: ;

[0089] For newly arrived wireless events, the corrected coordinates are: ;

[0090] in, The coordinates after DAS event correction. The coordinates are corrected for the wireless event. For the final error parameter estimation of DAS, For the estimation of final error parameters of wireless network;

[0091] For newly generated co-judgment events, their final fusion location can be and It is obtained by weighting the confidence level determined by its historical residuals.

[0092] Among them, such as Figure 2 As shown, in order to enable the system to adapt to long-term environmental changes, step S6 introduces a sliding window recursive least squares with an adaptive forgetting factor to perform online optimization of the batch processing IRLS in step S4.

[0093] Maintain a fixed size The sliding window always contains the latest... For each co-decidable event, whenever a new co-decidable event enters the window (while simultaneously removing the oldest event), a recursive least squares approach is used, combined with a forgetting factor. To update the error parameter estimate;

[0094] The update formulas for error parameter estimation include:

[0095] ;

[0096] ;

[0097] in, It is the regression vector corresponding to the new event. It is the observation difference of the new event. For the updated error parameter estimation, For the error parameter estimation before the update, It is a forgetting factor and satisfies , For the updated covariance, The covariance before the update. This represents the number of iterations.

[0098] Specifically, the aforementioned forgetting factors It is not fixed, but rather adaptively adjusted based on the covariance trace or residual sequence estimated from the error parameters. When the residuals suddenly increase, the adjustment can be temporarily reduced. To quickly track changes, increase when the system stabilizes. To improve estimation accuracy, and thus optimize the balance between tracking performance and steady-state accuracy under dynamic changing scenarios with a fixed forgetting factor;

[0099] Specifically, this step designs a closed-loop model update mechanism for long-term reliable operation, which can ensure the system's persistent accuracy in complex time-varying environments. It incorporates online high-confidence co-judgment events and their correction results into a dynamic training dataset managed by a sliding window, and then continuously updates the model parameters based on this dataset, enabling the system to automatically track and compensate for performance drift caused by slow environmental changes.

[0100] In summary, through steps S1 to S6, this invention completes a full closed loop from raw data acquisition, event correlation, error modeling, parameter estimation, real-time correction to adaptive update. By introducing weighted iteration (IRLS) to standard least squares, robustness is enhanced, and efficient online learning is achieved through recursive updates of the adaptive forgetting factor. Thus, targeted optimizations are made on the basis of existing algorithms, and effective dynamic correction of system positioning errors in complex environments is realized.

[0101] Example 2

[0102] Unlike Embodiment 1, this invention addresses the limitation of existing distributed acoustic sensing (DAS) technology in distinguishing the specific vibration sources in multiple cables laid in the same trench. Before executing step S1, it further includes an initialization and optical cable identification stage. This stage is used to establish an initial optical cable resource mapping library and calibrate the initial state of the system, wherein... Figure 3 As shown, this invention uses a dedicated micro-perturbation device to actively excite the target optical cable at known geographical coordinates, simultaneously recording the resulting precise fiber location, preliminary positioning coordinates of the wireless sensor network, and known optical cable identity. This constructs a reference mapping library. During online monitoring, the system matches the fiber location of the real-time sensed vibration event with this mapping library and verifies it using the joint judgment information from the wireless sensor network, thereby deterministically outputting the physical optical cable identity to which the event belongs. This stage specifically includes:

[0103] A specially coded vibration signal is applied to the target optical cable at a known geographic coordinate point using a micro-perturbation device;

[0104] The DAS host 10 detects and records the fiber optic location and vibration characteristics of each vibration event.

[0105] The wireless sensor network 20 synchronously detects and records the preliminary location coordinates of each acoustic event;

[0106] The central processing unit 40 spatiotemporally correlates DAS events and wireless events from the same physical stimulus, confirms the co-judgment relationship, and establishes a reference mapping library of optical cable identity ID, optical fiber location, and measured coordinates of wireless sensor network for optical cable identification during online monitoring.

[0107] Specifically, through multi-point calibration, the system constructs an initial optical cable resource mapping library for a region and collects the first batch of high-quality co-judgment event data for the initial learning of subsequent algorithms. After initialization, the system enters the real-time monitoring state shown in Example 1. This example takes the handling of an unknown external touch event as an example, and the specific process is as follows:

[0108] Step 1: When the optical cable is touched (such as during construction excavation), the DAS host 10 senses the vibration in real time by demodulating the phase change of the backscattered Rayleigh light in the optical fiber and outputs the optical fiber location of the vibration event. The wireless sensor network synchronously senses sound waves and outputs the preliminary location coordinates of the acoustic event, along with the intensity. All data carries a unified high-precision timestamp;

[0109] Step 2: The central processing unit 40 performs rapid correlation matching between DAS events and wireless events based on timestamps and spatial proximity. Upon successful matching, a co-judgment event is generated, containing key information such as the fiber optic location provided by the DAS. And intensity, wirelessly provided sound source region coordinates Harmonic characteristics;

[0110] Step 3: Based on the fiber optic location The system queries the optical cable resource mapping library established during the initialization phase to determine which specific optical cable the current event occurred on. The built-in algorithm treats each co-judgment event as a joint observation of DAS positioning error and wireless network positioning error. Through continuously accumulated co-judgment event data, the algorithm estimates and dynamically corrects both types of errors online (such as DAS fiber-to-ground mapping deviation and wireless sound velocity model deviation). Using the corrected parameters, the system determines the fiber location of the current event. and Compensation is performed, and the location information from both is fused to ultimately output a high-precision geographic coordinate system with fiber optic cable identification tags. ;

[0111] Step 4: Push all the above processing results to the monitoring terminal 50 for optical cable identification and positioning in real time. The processing results include event time and precise location. The information such as the fiber optic cable's identification ID, event intensity, and confidence level is visualized in various forms, including electronic maps, lists, and curves.

[0112] Step 5: The system continuously feeds back the processed high-quality co-judgment events (including the corrected location information) to the central processing unit 40 to update the model, enabling the system to have long-term adaptive capabilities to cope with the impact of environmental changes.

[0113] In summary, compared with existing technologies, the joint error dynamic correction method provided by this invention is a shallow fusion scheme that commonly uses simple spatiotemporal correlation and weighted averaging of fiber optic and wireless sensor data. In contrast, the joint error dynamic correction method based on the joint judgment of events of fiber optic and wireless sensors provided by this invention brings about a fundamental paradigm shift. It moves from passively fusing noisy observation results to actively constructing models and dynamically estimating and compensating for the core error sources of the sensing system itself. This achieves a fundamental and systematic improvement in positioning accuracy, enhances the long-term autonomous stability and adaptability of the system to complex environments, and has good application prospects.

[0114] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A joint error dynamic correction method based on co-judgment of events by optical fiber and wireless sensing, characterized in that, Specifically, the following steps are included: Step S1: The DAS host outputs a vibration event list. Each DAS event in the vibration event list includes the fiber optic location and the DAS local timestamp. The wireless sensor network synchronously outputs an acoustic event list. Each wireless event in the acoustic event list includes the preliminary positioning coordinates and the wireless network local timestamp. The central processing unit uses the time reference provided by the high-precision synchronous clock source to uniformly convert the timestamps of the two data streams to the absolute time coordinate system. Step S2: For each DAS event, map its fiber location to estimated geographic coordinates through the fiber optic geographic information database, search for wireless events within a preset time window, calculate the spatial distance for each wireless event, and if the spatial distance is less than a preset spatial association threshold, determine that the DAS event and the wireless event are two observations of the same physical event, generate a co-judgment event instance and store it in the co-judgment event pool. Step S3: Define the error parameters to be estimated. The error parameters include the DAS error parameter vector and the wireless network error parameter vector. Based on the error parameters, observation data pairs, and fiber optic locations, establish a joint observation model and obtain the residual equation by eliminating the real geographical location parameters. Step S4: Using the iterative reweighted least squares algorithm, initialize the error function model, the number of iterations and the initial weights, perform a first-order Taylor expansion on the residual equation to construct a linearized incremental equation, solve the weighted normal equation to complete the parameter iterative update, and update the event weights according to the iterative residuals. Repeat the iteration until the convergence condition is met, and output the final error parameter estimate. Step S5: Based on the final error parameter estimation, the positioning output results of DAS events and wireless events are dynamically corrected in real time to obtain the corrected coordinates. The confidence level is determined based on the historical residuals of the corrected coordinates, and the final fusion positioning result of the co-judgment events is obtained by weighted averaging. Step S6: Maintain a fixed-size sliding window to retain the latest co-judgment event data. When the latest co-judgment event enters the window, use a recursive least squares algorithm combined with a forgetting factor to update the error parameter estimate. Adjust the forgetting factor according to the covariance trace or residual sequence of the error parameter estimate to achieve online adaptive optimization of the parameters. The joint observation model satisfies: ； ； in, and For parameterized error functions, Indicates random observation noise. This is the DAS error parameter vector. This is the wireless network error parameter vector. For the first event in the co-judgment event pool The actual geographical location of the event For fiber optic location, and For a co-determined event instance The observation data is relevant to information. For the first DAS observation noise for each event, For the first Wireless network observation noise for each event; The expression for the residual equation is: ； in, The residual equations are represented as the DAS error parameter vector and the wireless network error parameter vector; The DAS error parameter vector is used to correct the mapping deviation from the fiber location to the geographic coordinates, and the wireless network error parameter vector is used to correct the positioning deviation caused by the deviation in sound wave propagation speed.

2. The joint error dynamic correction method based on optical fiber and wireless sensing co-judgment events as described in claim 1, characterized in that, Before performing step S1, the joint error dynamic correction method further includes an initialization and optical cable identification phase: A specially coded vibration signal is applied to the target optical cable at a known geographic coordinate point using a micro-perturbation device; The DAS host detects and records the fiber location and vibration characteristics of each vibration event. The wireless sensor network synchronously detects and records the preliminary location coordinates of each acoustic event; The central processing unit spatiotemporally correlates DAS events and wireless events from the same physical stimulus, confirms the co-judgment relationship, and establishes a reference mapping library of optical cable identity ID, optical fiber location, and measured coordinates of the wireless sensor network for optical cable identification during online monitoring.

3. The joint error dynamic correction method based on optical fiber and wireless sensing co-judgment events as described in claim 1, characterized in that, The DAS host in step S1 is a distributed acoustic wave sensing host. The DAS host obtains the fiber location and DAS local timestamp of the vibration event list by processing the back Rayleigh scattering signal in real time and by phase demodulation and event detection.

4. The joint error dynamic correction method based on co-judgment of events by optical fiber and wireless sensing according to claim 1, characterized in that, The preset time window in step S2 is a symmetrical interval before and after the absolute time of the DAS event. The spatial distance is obtained by calculating the Euclidean distance between the estimated geographic coordinates and the preliminary location coordinates of the wireless event, and the expression is: ,in, For spatial distance, To estimate geographic coordinates, These are the initial coordinates for locating the wireless event.

5. The joint error dynamic correction method based on optical fiber and wireless sensing co-judgment events according to claim 1, characterized in that, The error function model in step S4 includes a linear drift model on the fiber optic sensing side and a regional translation model on the wireless sensing side. The expression for the linear drift model is: ,in, , This is a fixed offset vector in the DAS error. This is the vector of linear drift coefficients in the DAS error. The fiber location for the DAS event; The expression for the regional translation model is: ,in, , This is the regional translation vector in the wireless network error.

6. The joint error dynamic correction method based on co-judgment of events by optical fiber and wireless sensing according to claim 1, characterized in that, The iterative reweighted least squares algorithm in step S4 uses the Huber weight function and the Tukey weight function to update the event weights.

7. The joint error dynamic correction method based on co-judgment of events by optical fiber and wireless sensing according to claim 5, characterized in that, The formula for calculating the corrected coordinates in step S5 is as follows: For newly arrived DAS events, the corrected coordinates are: ; For newly arrived wireless events, the corrected coordinates are: ; in, The coordinates after DAS event correction. The coordinates are corrected for the wireless event. For the final error parameter estimation of DAS, This is for estimating the final error parameters of the wireless network.

8. The joint error dynamic correction method based on co-judgment of events by optical fiber and wireless sensing according to claim 1, characterized in that, The update formula for the error parameter estimation in step S6 includes: ； ； in, It is the regression vector corresponding to the new event. It is the observation difference of the new event. For the updated error parameter estimation, For the error parameter estimation before the update, It is a forgetting factor and satisfies , For the updated covariance, Covariance before update This represents the number of iterations.

Images