A Landslide Deformation Monitoring Method Based on GNSS and MEMS Collaborative Judgment

Abstract

This invention discloses a landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination, belonging to the field of geological disaster location monitoring technology. It extracts candidate states from a collaboratively reliable state, performs synchronous response locking on rapidly changing candidate states, retains responses to missing conditions on slowly changing candidate states, intercepts signals to fluctuate candidate states, and performs windowed re-injection on verified candidate states, forming gated diversion segments. Based on these gated diversion segments, it performs response binding and solidification on rapidly changing deformation segments, continuously accumulates and solidifies slowly changing deformation segments, performs sequence continuity processing on intercepted signal fluctuation segments, and verifies and retains windowed re-injection segments, forming a deformation classification segment set. This invention improves the reliability of initial deformation judgment, reduces misjudgments, enhances sequence continuity, and improves the overall stability, continuity, and traceability of monitoring results.

Description

Technical Field

[0001] This invention relates to the field of geological disaster location and monitoring technology, and in particular to a landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination. Background Technology

[0002] With the development of continuous observation via global navigation satellite systems, microelectromechanical inertial sensing, time synchronization, and time-series data processing technologies, landslide monitoring is gradually shifting from manual inspections and periodic measurements to automated, continuous, and multi-source sensing. Related technologies typically deploy GNSS monitoring points at key locations within the landslide body to acquire positioning times, displacement changes, and calculation status. Simultaneously, MEMS accelerometers, gyroscopes, or attitude sensors are configured to collect local dynamic responses, and a continuous observation sequence is formed using a unified time reference. This provides a data foundation for landslide displacement evolution analysis, deformation trend identification, and disaster early warning.

[0003] GNSS positioning results are susceptible to satellite obstruction, multipath effects, signal quality, and changes in solution status. Sudden jumps, discontinuities, or short-term fluctuations may occur in the positioning sequence. When relying solely on positioning changes for judgment, the stability of distinguishing between signal disturbances and actual slip is insufficient. On the other hand, MEMS response is more suitable for characterizing vibration, tilt, or instantaneous dynamic changes, but has limited ability to express low-dynamic, continuous cumulative deformations such as slow creep. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination to solve the problem of collaboratively discriminating the true deformation of landslides using GNSS positioning status and MEMS response evidence.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination, comprising: real-time acquisition of navigation satellite positioning data and MEMS response data of landslide monitoring points, and time alignment to generate synchronous observation segments; credible labeling of navigation satellite positioning data in synchronous observation segments to obtain credible label segments, and credible response coupling with MEMS response data in synchronous observation segments to form a collaborative credible state; extraction of candidate states from the collaborative credible state, performing synchronous response locking on fast-changing candidate states, performance response missing retention on slow-changing candidate states, performance signal fluctuation interception on signal fluctuation candidate states, and performance windowed refeedback on verification candidate states to form gated diversion segments; based on the gated diversion segments, performance response binding and solidification of fast-deformation segments, performance continuous accumulation and solidification of slow-deformation segments, performance sequence continuity processing of intercepted signal fluctuation segments, and performance verification retention of windowed refeedback segments to form a deformation classification segment set; and temporal embedding and combination of sudden slip sequences, slow creep sequences, and continuous positioning sequences in the deformation classification segment set to form landslide deformation.

[0007] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the generation of synchronous observation segments is specifically as follows: Navigation satellite positioning data and microelectromechanical response data are acquired through GNSS and MEMS respectively, forming a dual-source time-series data stream; Using the positioning output time in the dual-source time-series data stream as the binding reference time, the response data segments are extracted, numbered, and bound to form positioning time data segments; The navigation satellite positioning data and positioning time data segments are bound to a reference time, and synchronous observation segments are encapsulated. A segment index identifier is written for each synchronous observation segment to generate a synchronous observation segment.

[0008] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the acquisition of the reliable marker fragment is specifically as follows: Navigation satellite positioning data is obtained from synchronous observation segments. When the conditions of continuous output, stable change and available positioning solution are met, a positioning trust mark is written. When the navigation satellite positioning data shows jumps, discontinuities and abnormal positioning solutions, a positioning untrust mark is written, forming a positioning mark data segment. The location marker data segments are layered into trusted segments and arranged according to the time order of synchronous observation segments to generate trusted marker segments.

[0009] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the formation of a collaboratively reliable state is specifically as follows: Based on trusted marker fragments, response evidence is attached. Synchronization response markers and response missing markers are attached to trusted location fragments and suspected fluctuation fragments, respectively, to form a response evidence record. Based on the response evidence record, the trusted location segment is matched with the synchronous response marker and response missing marker within the same time range, and fast-changing candidate state and slow-changing candidate state are generated according to the matching relationship to form a trusted location candidate record. Obtain response evidence records corresponding to suspected fluctuation segments in the trusted positioning candidate records within the same time range, retain suspected fluctuation segments with missing response markers as signal fluctuation candidate states, and transfer suspected fluctuation segments with synchronization response markers to the review candidate state to form candidate state records; Candidate state records are arranged according to the time sequence of synchronous observation segments, and states are merged. Adjacent different candidate states are retained as independent state segments to form a cooperative and reliable state.

[0010] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the step of performing response missing retention for slowly changing candidate states is as follows: Candidate states are extracted from the cooperative trust state, and the fast-changing candidate states, slow-changing candidate states, signal fluctuation candidate states, and verification candidate states in the cooperative trust state are organized into a candidate state sequence. Based on the fast-changing candidate states in the candidate state sequence, the navigation satellite positioning data corresponding to the fast-changing candidate states are bound to the microelectromechanical response data in the same time range according to the binding reference time, and the fast-changing candidate states are updated to fast-changing transformation segments to form a fast-changing locking sequence. Based on the slow-change candidate states in the fast-change locking sequence, navigation satellite positioning data with reliable positioning and missing response markers are separated from the signal fluctuation interception path, and continuous reliable positioning change segments are retained along the time sequence of synchronous observation segments to form a fast-slow diversion sequence.

[0011] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the formation of the gated diversion segment is specifically as follows: Signal fluctuation interception is performed based on the candidate states of signal fluctuations in the fast and slow flow splitting sequence to form a fluctuation interception sequence; The candidate states for verification in the fluctuation interception sequence are subjected to windowed backfeed and combined to form a gated diversion segment.

[0012] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the step of responding and binding the fast deformation segments is specifically as follows: Based on the routing attributes carried in the gated routing segments, the gated routing segments are extracted and a routing solidification queue is formed. The fast deformation segments in the diversion and solidification queue are responded to and bound to solidify, and the corresponding positioning change segments are solidified into sudden slip segments to form the first solidification queue. Based on the slow deformation segments in the first curing queue, continuous cumulative curing is performed to solidify the continuously positioning and changing segments into slow creeping segments, thus forming the second curing queue.

[0013] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the formation of the deformation classification segment set is specifically as follows: The signal fluctuation segments in the second solidified queue are processed to achieve sequence continuity. The signal fluctuation segments are isolated and the corresponding positions are filled in to obtain a continuous positioning sequence. For the windowed re-irrigation segments retained in the continuous positioning sequence, the time connection with the adjacent synchronous observation segments is checked. When the windowed re-irrigation segments have not yet been classified into the synchronous response marker and the response missing marker, the windowed re-irrigation segments are retained and combined to form a deformation classification segment set.

[0014] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the step of performing temporal embedding and combination is as follows: The sudden slip sequence in the deformation classification fragment set is marked as a fast insertion fragment, the slow creep sequence is marked as a trend embedding fragment, and the continuous localization sequence is marked as a localization main line fragment. These are combined to generate a sequence splicing benchmark. Based on the sequence splicing benchmark, the continuous processing of the positioning sequence is laid out as the basic positioning main line, the position intercepted by signal fluctuations is retained, and the fast insertion position and trend embedding position are set to generate the positioning main line.

[0015] As a preferred embodiment of the landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination described in this invention, the formation of landslide deformation specifically includes the following: The slow creep sequence is embedded into the trend embedding position in the positioning main line, and continuous displacement change is performed to replace it. The slow change continuation mark is written into the replaced positioning segment to generate the trend-embedded positioning sequence. The sudden slip sequence is inserted into the rapid insertion position in the trend-embedded positioning sequence, and then connected with the adjacent positioning segments in the trend-embedded positioning sequence to form a landslide deformation.

[0016] The beneficial effects of this invention are as follows: by coupling trusted tags with MEMS responses, a collaborative trusted state is formed to distinguish between fast changes, slow changes and fluctuations, thereby improving the reliability of initial deformation judgment. Through gating and differential solidification, fast slip, slow creep and signal fluctuations are processed and embedded for output separately, reducing misjudgment and enhancing sequence continuity, and improving the overall stability, continuity and traceability of monitoring results. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination.

[0019] Figure 2 A flowchart for synchronous observation fragment generation and trusted labeling.

[0020] Figure 3 This is a flowchart illustrating the formation of trusted response coupling and collaborative trusted state.

[0021] Figure 4 A flowchart for generating landslide deformation for the gated diversion segment. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 This is one embodiment of the present invention, which provides a landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination, comprising the following steps: S1: Real-time acquisition of navigation satellite positioning data and microelectromechanical response data of landslide monitoring points, time alignment, and generation of synchronous observation segments; S1.1: Navigation satellite positioning data and microelectromechanical response data are acquired through GNSS and MEMS respectively to form a dual-source time-series data stream; Navigation satellite positioning data includes the positioning time and corresponding location change information.

[0026] Microelectromechanical response data includes the response time and the corresponding dynamic response information.

[0027] Furthermore, navigation satellite positioning data of landslide monitoring points are continuously acquired via GNSS, and microelectromechanical response data of landslide monitoring points are continuously acquired via MEMS. The navigation satellite positioning data and microelectromechanical response data are then sequentially arranged according to the acquisition time, so that the same landslide monitoring point forms a dual-source time-series data stream containing navigation satellite positioning data and microelectromechanical response data within a continuous time period.

[0028] S1.2: Using the positioning output time in the dual-source time-series data stream as the binding reference time, extract the response data segment, number and bind it to form the positioning time data segment; The binding reference time refers to the positioning output time generated by navigation satellite positioning data in the dual-source time-series data stream. It is used as a reference point for time alignment and to extract response data segments within the corresponding time range from the microelectromechanical response data accordingly.

[0029] Furthermore, using the positioning output time in the dual-source time-series data stream as the binding reference time, the response interception start and end points are determined in the microelectromechanical response data according to the binding reference time, and the response data segment located between the response interception start and end points is intercepted; a corresponding binding number is generated based on the positioning output time, and the binding number is written into both the navigation satellite positioning data and the response data segment. A one-to-one correspondence is established between the positioning output time of the navigation satellite positioning data, the time range of the response data segment, and the binding number to form the positioning time data segment.

[0030] It should be noted that the corresponding number generated according to the positioning output time is used to match the navigation satellite positioning data and the extracted response data segment under the same binding reference time, so that the navigation satellite positioning data and the response data segment can be matched according to the same number when encapsulating synchronous observation segments in the later stage.

[0031] S1.3: Encapsulate navigation satellite positioning data and positioning time data segments into synchronous observation segments according to the bound reference time, and write segment index identifiers for synchronous observation segments to generate synchronous observation segments; According to the binding reference time, retrieve the corresponding navigation satellite positioning data and positioning time data segment, align the navigation satellite positioning data and positioning time data segment under the same binding reference time, and encapsulate the time-aligned navigation satellite positioning data and positioning time data segment into a synchronous observation segment.

[0032] Write a segment index identifier to the synchronous observation segment. The segment index identifier is used to record the binding reference time, arrangement order and binding relationship of the synchronous observation segment. The synchronous observation segments are stored in order according to the segment index identifier to generate synchronous observation segments.

[0033] S2: The navigation satellite positioning data in the synchronous observation segment is credibly labeled, the credibly labeled segment is obtained, and the microelectromechanical response data in the synchronous observation segment is combined to perform credible response coupling to form a cooperative credible state; S2.1: Obtain navigation satellite positioning data from synchronous observation segments. When the conditions of continuous output, stable change, and usable positioning solution are met, write a positioning trustworthy flag. When the navigation satellite positioning data shows jumps, discontinuities, or abnormal positioning solutions, write a positioning untrustworthy flag to form a positioning flag data segment. Furthermore, navigation satellite positioning data is extracted according to the segment index identifier of the synchronous observation segment, and the continuous output status, adjacent positioning change status, and positioning solution availability status of the navigation satellite positioning data are checked along the time sequence of the synchronous observation segment. When the navigation satellite positioning data is continuously output, the adjacent positioning change does not exceed the preset positioning change threshold, and the positioning solution is in an available state, a positioning trust mark is written to the corresponding navigation satellite positioning data.

[0034] The navigation satellite positioning data is checked again along the time sequence of the synchronous observation segments. If the navigation satellite positioning data does not produce a corresponding positioning output at the predetermined sampling time, and there is a missing time interval between two adjacent positioning outputs, it is considered as discontinuous output. If the position change corresponding to two adjacent navigation satellite positioning data exceeds the preset positioning change threshold, it is considered as abnormal adjacent positioning change. If the positioning solution corresponding to the navigation satellite positioning data is in an unavailable state, the solution state is abnormal, or the positioning result is invalid, it is considered as abnormal positioning solution. Navigation satellite positioning data with discontinuous output, abnormal adjacent positioning change, or abnormal positioning solution are written with a positioning untrusted flag, and the navigation satellite positioning data carrying the positioning trusted flag and the positioning untrusted flag are arranged in the time sequence of the synchronous observation segments to form a positioning flag data segment.

[0035] It should be noted that the positioning change threshold (example range: 10 mm to 50 mm) is set based on the normal noise fluctuation and positioning accuracy of navigation satellite positioning data during the stable monitoring period, and the upper limit is set based on the normal slow displacement growth range of the landslide, the on-site geological deformation rate and false alarm control requirements within adjacent sampling periods.

[0036] S2.2: Perform trusted segment layering on the location marker data segments and arrange them according to the time order of the synchronous observation segments to generate trusted marker segments; Furthermore, based on the reliable and unreliable positioning markers carried in the positioning marker data segments, navigation satellite positioning data carrying reliable positioning markers are classified into reliable positioning segments, and navigation satellite positioning data carrying unreliable positioning markers are classified into suspected fluctuation segments. The segment index identifiers of the synchronous observation segments corresponding to the reliable positioning segments and suspected fluctuation segments are retained.

[0037] The reliable location segments and suspected fluctuation segments are arranged according to the time sequence of the synchronous observation segments. When adjacent segments carry the same marker, they are merged into continuous segments. When adjacent segments carry different markers, a layer boundary is set at the marker change position. The arranged reliable location segments and suspected fluctuation segments are combined to generate reliable marker segments.

[0038] S2.3: Response evidence attachment is performed based on trusted marker fragments. Synchronization response markers and response missing markers are attached to trusted location fragments and suspected fluctuation fragments respectively to form response evidence records; Furthermore, based on the segment index identifier in the trusted marker segment, the time range corresponding to the trusted location segment and the suspected fluctuation segment is determined, and the response data segment within the corresponding time range is extracted from the microelectromechanical response data in the synchronous observation segment.

[0039] The response data segments are time-correlated with the corresponding trusted positioning segments or suspected fluctuation segments. When the response data segment shows a change in sensor response within the corresponding time range, a synchronous response mark is affixed to the corresponding trusted positioning segment or suspected fluctuation segment. When the response data segment does not show a change in sensor response within the corresponding time range, a response missing mark is affixed to the corresponding trusted positioning segment or suspected fluctuation segment. The marks are arranged in the time order of the synchronous observation segments, and the trusted positioning segments and suspected fluctuation segments after the synchronous response mark or response missing mark are affixed are retained. The correspondence between the segments and the marks is preserved to form a response evidence record.

[0040] It should be noted that the presence of sensor response changes within the corresponding time range of the response data segment means that at least one of the changes in the composite amplitude of the three-axis acceleration, the composite amplitude of the angular velocity, and the attitude angle within the response data segment exceeds the normal fluctuation range of the corresponding response quantity during the stable monitoring period, and remains in the same direction or in a continuous change state in continuous sampling points; otherwise, it is considered as a missing response.

[0041] The normal fluctuation range is determined based on the noise level of the microelectromechanical response data, the field vibration background, or the field calibration results during the stable monitoring period.

[0042] S2.4: Based on the response evidence record, match the trusted location segment with the synchronous response marker and response missing marker within the same time range, and generate fast-changing candidate states and slow-changing candidate states according to the matching relationship to form a trusted location candidate record; Furthermore, based on the response evidence record, a reliable positioning segment is extracted, and according to the segment index identifier corresponding to the reliable positioning segment, the synchronization response marker and response missing marker carried by the reliable positioning segment are obtained. The reliable positioning segments and corresponding markers are paired and collected, so that the navigation satellite positioning data in the reliable positioning segment and the microelectromechanical response data in the same time range maintain a correspondence, forming a reliable positioning response pair.

[0043] Based on the type of marker carried by the trusted location response pair, the trusted location response pairs are classified into candidate states. Trusted location response pairs carrying synchronization response markers are classified into fast-changing candidate states, and trusted location response pairs carrying response missing markers are classified into slow-changing candidate states. The fast-changing candidate states and slow-changing candidate states are arranged according to the time order of the synchronous observation segments. The correspondence between the fast-changing candidate states, slow-changing candidate states and trusted location segments is preserved to form a trusted location candidate record.

[0044] It should be noted that the correspondence between fast-changing candidate states, slow-changing candidate states and trusted positioning segments means that each fast-changing candidate state or slow-changing candidate state originates from a specific trusted positioning segment, and maintains a correspondence with the trusted positioning segment through segment index identifier, time range and positioning data location.

[0045] S2.5: Obtain the response evidence records corresponding to the suspected fluctuation segments in the trusted positioning candidate records within the same time range, retain the suspected fluctuation segments with response missing markers as signal fluctuation candidate states, and transfer the suspected fluctuation segments with synchronization response markers to the verification candidate states to form candidate state records; Furthermore, suspected fluctuation segments are extracted from the trusted marker segments associated with trusted positioning candidate records, and the synchronization response markers or response missing markers carried by the suspected fluctuation segments in the response evidence records are retrieved according to the segment index identifier; the suspected fluctuation segments are paired with the synchronization response markers or response missing markers to bind the navigation satellite positioning data in the suspected fluctuation segments with the microelectromechanical response data in the same time range, forming suspected fluctuation response pairs.

[0046] Based on the response markers corresponding to the suspected fluctuation segments in the suspected fluctuation response pairs, the suspected fluctuation response pairs are classified into candidate states. When the response markers corresponding to the suspected fluctuation segments are missing, the suspected fluctuation response pairs are classified into signal fluctuation candidate states. When the suspected fluctuation segments correspond to the synchronization response markers, the suspected fluctuation response pairs are classified into verification candidate states. The signal fluctuation candidate states, verification candidate states, and fast-changing candidate states and slow-changing candidate states in the reliable positioning candidate records are merged according to the time order of the synchronization observation segments to form a candidate state record.

[0047] It should be noted that a suspected fluctuation response pair refers to the corresponding data formed by binding a suspected fluctuation segment with microelectromechanical response data within the same time range according to the segment index identifier. This data is used to determine whether the suspected fluctuation segment belongs to a signal fluctuation candidate state or a verification candidate state.

[0048] S2.6: Arrange the candidate state records according to the time order of the synchronous observation segments, merge the states, and retain adjacent different candidate states as independent state segments to form a cooperative and reliable state. Furthermore, the candidate state records are arranged in chronological order according to the synchronous observation segments. The fast-changing candidate states, slow-changing candidate states, signal fluctuation candidate states, and verification candidate states in the candidate state records are bound to the corresponding synchronous observation segments according to the segment index identifiers, and a candidate state time sequence chain is established according to the order of the segment index identifiers.

[0049] Merge consecutive identical candidate states along the candidate state time sequence chain. Merge consecutive fast-changing candidate states into fast-changing state segments, consecutive slow-changing candidate states into slow-changing state segments, consecutive signal fluctuation candidate states into signal fluctuation state segments, and consecutive verification candidate states into verification state segments. When the candidate state types at adjacent positions are different, set state boundaries at the state change position and retain adjacent different candidate states as independent state segments to form a collaborative reliable state.

[0050] S3: Extract candidate states from the collaborative trusted state, perform synchronous response locking on fast-changing candidate states, perform response missing retention on slow-changing candidate states, perform signal fluctuation interception on signal fluctuation candidate states, and perform windowed backfeeding on verification candidate states to form gated shunting segments. S3.1: Extract candidate states from the cooperative trust state and organize the fast-changing candidate states, slow-changing candidate states, signal fluctuation candidate states, and verification candidate states in the cooperative trust state into a candidate state sequence; Furthermore, according to the segment index identifiers corresponding to each state segment in the collaborative trust state, fast-changing candidate states, slow-changing candidate states, signal fluctuation candidate states, and verification candidate states are extracted, and each candidate state is indexed and bound to the corresponding synchronous observation segment to form a candidate state extraction record.

[0051] Candidate states with rapid changes, slow changes, signal fluctuations, and verification are arranged in the order of their segment index identifiers. Candidate states that appear consecutively and are of the same type are arranged in a continuous sequence, while adjacent candidate states of different types are separated at the position of the state change, thus forming a candidate state sequence.

[0052] It should be noted that the boundary between adjacent candidate states of different types at the state change location refers to the segmentation marker written at the location where the candidate state type changes, which is used to distinguish the time range of the previous candidate state and the next candidate state.

[0053] S3.2: Based on the fast-changing candidate states in the candidate state sequence, the navigation satellite positioning data corresponding to the fast-changing candidate states are bound to the microelectromechanical response data in the same time range according to the binding reference time, and the fast-changing candidate states are updated to fast-changing segments to form a fast-changing locking sequence. Furthermore, based on the rapidly changing candidate states in the candidate state sequence, the navigation satellite positioning change segments corresponding to the rapidly changing candidate states are extracted according to the segment index identifier, and the microelectromechanical response data is extracted from the same time range; the navigation satellite positioning change segments and the microelectromechanical response data are time-registered according to the binding reference time to establish the synchronous response correspondence between the navigation satellite positioning change segments and the microelectromechanical response data.

[0054] Align the start and end times of the navigation satellite positioning change segment with the response time range of the microelectromechanical response data, write the segment index identifier of the navigation satellite positioning change segment into the corresponding microelectromechanical response data, and write the response mark of the microelectromechanical response data into the corresponding navigation satellite positioning change segment. This ensures that the navigation satellite positioning change segment and the microelectromechanical response data are stored in pairs under the same segment index identifier. Update the fast change candidate state to a fast change segment, and form a fast change locking sequence according to the time order of the candidate state sequence.

[0055] S3.3: Based on the slow-change candidate states in the fast-change locking sequence, the navigation satellite positioning data with reliable positioning and carrying response missing markers are separated from the signal fluctuation interception path, and continuous reliable positioning change segments are retained along the time sequence of the synchronous observation segments to form a fast-slow diversion sequence; Furthermore, based on the fast-changing locking sequence, the continuous reliable positioning change segments corresponding to the slow-changing candidate states are extracted according to the segment index identifier, and the microelectromechanical response data within the same time range are retrieved; when the microelectromechanical response data does not carry a synchronization response marker, the segment index identifier, time sequence and sequential connection relationship of the continuous reliable positioning change segments are preserved to form a slow-changing retention record.

[0056] Reliable positioning change segments that lack synchronous response but remain continuous are retained in the deformation transmission path. These continuous reliable positioning change segments are then linked together according to the segment index identifier. The linked continuous reliable positioning change segments are written with slow deformation tags to form slow deformation segments. The fast deformation segments and slow deformation segments in the fast deformation locking sequence are then arranged according to the time order of the synchronous observation segments to form a fast-slow split sequence.

[0057] S3.4: Based on the candidate states of signal fluctuations in the fast and slow splitting sequence, signal fluctuations are intercepted to form a fluctuation interception sequence; Furthermore, by using the fast and slow split sequence, the signal fluctuation location jump segment corresponding to the candidate state of the signal fluctuation is extracted according to the segment index identifier, and the microelectromechanical response data within the same time range is retrieved; when the microelectromechanical response data carries a response missing marker, the segment index identifier, time position and adjacent reliable location segment position of the signal fluctuation location jump segment are retained.

[0058] The signal fluctuation location jump segment is marked as a non-deformation transmission data segment and separated from the deformation transmission path; the separated position is retained as the position for subsequent sequence continuity processing; the signal fluctuation location jump segment after the completion of non-deformation transmission marking and path separation is taken as a signal fluctuation segment; and the fast deformation segment, slow deformation segment and signal fluctuation segment in the fast and slow diversion sequence are arranged in the time order of the synchronous observation segment to form a fluctuation interception sequence.

[0059] S3.5: Extend the window and re-inject the candidate states of the fluctuation interception sequence to form a gated diversion segment.

[0060] Furthermore, based on the fluctuation interception sequence, the data segments corresponding to the candidate states for review are extracted according to the segment index identifier, and the adjacent synchronous observation segments before and after the candidate states for review are located. The data segments corresponding to the candidate states for review are connected with the adjacent synchronous observation segments in chronological order, and the temporal connection relationship between the navigation satellite positioning data, microelectromechanical response data and the adjacent synchronous observation segments corresponding to the candidate states for review is preserved to form a windowed backfeed record.

[0061] The data segment corresponding to the candidate state of the verification is written into the refeeding mark, and the data segment written into the refeeding mark is placed at the next round of trusted response coupling position to form a windowed refeeding segment. The windowed refeeding segment retains a time placeholder in this round of deformation combination, but is not used as the actual deformation output. The segment state at the corresponding time position will be updated after the next round of trusted response coupling confirmation. Then, the fast deformation segment, slow deformation segment, signal fluctuation segment and windowed refeeding segment are arranged in the time order of the synchronous observation segment and combined to form a gated shunt segment.

[0062] It should be noted that the temporal connection relationship refers to the connection relationship between the data segment corresponding to the candidate state and the previous synchronous observation segment and the next synchronous observation segment in time sequence. This includes the start and end times of the data segment corresponding to the candidate state, the end time of the previous synchronous observation segment, the start time of the next synchronous observation segment, and whether the three are continuous or have a time interval.

[0063] S4: Based on the gated diversion segment, fast deformation segments are responded to and bound, slow deformation segments are continuously accumulated and bound, intercepted signal fluctuation segments are processed into a sequence of continuous segments, and windowed recharge segments are reviewed and retained to form a set of deformation classification segments. S4.1: Extract the gated switching segments according to the switching attributes carried in the gated switching segments to form a switching solidification queue; Furthermore, according to the shunting attributes carried in the gated shunting segments, the fast deformation segment, slow deformation segment, signal fluctuation segment, and windowed backfeed segment in the gated shunting segments are respectively written with solidified path tags; the fast deformation segment is written with a response binding solidified path tag, the slow deformation segment is written with a continuous accumulation solidified path tag, the signal fluctuation segment is written with a sequence continuity path tag, and the windowed backfeed segment is written with a verification retention path tag. After writing the solidified path tags, each type of segment is assigned to the corresponding solidified path according to the segment index identifier.

[0064] The segments in each solidified path are arranged according to the time sequence of the synchronous observation segments, and the solidification order is set according to the processing order of response-bound solidified path, continuous cumulative solidified path, sequence continuous path and verification retention path; the fast deformation segment, slow deformation segment, signal fluctuation segment and windowed recharge segment that have completed solidified path marking and solidification order setting are combined to form a diversion solidification queue.

[0065] It should be noted that the shunting attribute is a processing path marker carried by the gated shunting segment, used to indicate whether the corresponding segment in the gated shunting segment belongs to a fast deformation segment, a slow deformation segment, a signal fluctuation segment, or a windowed refeeding segment, and accordingly determine whether the corresponding segment enters response binding and solidification, continuous accumulation and solidification, sequence continuous processing, or review and retention.

[0066] S4.2: Respond to the fast deformation segment in the diversion solidification queue, and solidify the corresponding positioning change segment into a sudden slip segment to form the first solidification queue; Furthermore, based on the response binding and solidification path markers in the diversion and solidification queue, the fast deformation segment is located, and the positioning change segment and the microelectromechanical response segment under the same segment index are extracted from the fast deformation segment; the start and end positions of the positioning change segment are aligned with the response time range of the microelectromechanical response segment, and the positioning change segment and the microelectromechanical response segment are bound in pairs according to the segment index to form a fast change response binding record.

[0067] The paired and bound positioning change segments are written into the sudden slip mark, and the segment index identification binding relationship between the positioning change segment and the microelectromechanical response segment is retained; the positioning change segment written into the sudden slip mark is solidified into a sudden slip segment, and the sudden slip segment is combined with the slow deformation segment, signal fluctuation segment and windowed recharge segment that have not yet been solidified in the shunt solidification queue in chronological order to form the first solidification queue.

[0068] S4.3: Based on the slow deformation segments in the first curing queue, perform continuous cumulative curing to solidify the continuously positioning change segments into slow creep segments and form the second curing queue; Furthermore, based on the slow deformation segments in the first solidification queue, continuous cumulative solidification is performed to solidify the continuous positioning change segments into slow creep segments, forming a second solidification queue. Further, based on the continuous cumulative solidification path markers in the first solidification queue, the slow deformation segments are located, and continuous positioning change segments within the slow deformation segments are extracted according to the segment index identifier and time sequence. The temporal continuity and displacement direction continuity of adjacent continuous positioning change segments are checked. In the case where the microelectromechanical response segment does not carry a synchronization response marker, the cumulative relationship between continuous positioning change segments is preserved, and a slow-change cumulative consistency value is calculated. This slow-change cumulative consistency value characterizes the degree to which continuous positioning change segments continuously accumulate along the same slow creep direction.

[0069] The formula for calculating the slow-varying cumulative consensus value is: ; In the formula, This represents the slow-change cumulative consistency value, used to determine whether the cumulative relationship between continuously changing positioning segments meets the curing requirements of the slow-creep segment; This represents the total number of segments involved in the continuous cumulative solidification of positioning changes. Indicates the first The displacement change vector of a series of continuous positioning change segments is used to represent the direction and magnitude of positional change between adjacent positioning segments. It indicates the main direction of continuation of the slow creeping segment and is used as a reference for whether continuous positioning change segments accumulate along the same creeping direction; This is used to avoid the minimum value where the denominator is zero; This indicates the number of consecutive positioning change segments that satisfy the condition of continuous continuation. This indicates the total number of continuously changing positioning segments participating in continuous cumulative solidification; when A value greater than 0 indicates that the location change segment has a positive contribution along the main continuation direction, so the value is retained. If the value is less than or equal to 0, it means that the direction is opposite or there is no positive contribution, so it is treated as 0.

[0070] When the slow-varying cumulative consistency value reaches the preset cumulative threshold, the continuous positioning change segments within the same continuous time range are connected one after the other according to the segment index, and the continuous positioning change segments after connection are written with a slow-varying continuation mark, thus solidifying the continuous positioning change segments after connection into slow-creep segments; when the slow-varying cumulative consistency value does not reach the preset cumulative threshold, the corresponding continuous positioning change segments are retained to the next round of trusted response coupling position; the slow-creep segments are combined with the sudden slip segments, unprocessed signal fluctuation segments and windowed backfeed segments already formed in the first solidification queue in chronological order to form the second solidification queue.

[0071] It should be noted that the cumulative threshold (example range: 0.65 to 0.85) is set with the lower limit based on the random fluctuation level of navigation satellite positioning data and false alarm control requirements during the stable monitoring period, and the upper limit based on the cumulative stability of the actual slow creep segment under the influence of measurement noise, the on-site geological deformation rate, and the omission control requirements.

[0072] It should be noted that the formula for the slow-varying cumulative consistency value is based on the technical characteristic that slow-varying deformation segments may still belong to true slow creep when the microelectromechanical response segment does not carry a synchronous response marker. The core basis includes the displacement direction continuity, temporal continuity, and cumulative stability of the continuous positioning change segment. In the formula, the positive contribution of the continuous positioning change segment in the main continuity direction of the slow creep segment is used to distinguish between continuous unidirectional accumulation and random reciprocating jumps. By the ratio between the number of continuous positioning change segments that meet the continuous continuity condition and the total number of continuous positioning change segments participating in the continuous accumulation and solidification, the slow-varying deformation segment is constrained to have sufficient temporal continuity. Therefore, the slow-varying cumulative consistency value does not simply calculate the displacement magnitude, but simultaneously measures whether it accumulates in the same direction and whether it occurs continuously. It is used to support the solidification of continuous positioning change segments into slow creep segments and avoids the missolution of random positioning fluctuations or short-term isolated changes into slow creep segments.

[0073] S4.4: Perform sequence continuity processing on the signal fluctuation segments in the second solidified queue, isolate the signal fluctuation segments, and supplement the corresponding positional continuity relationship to obtain a continuous positioning sequence; Furthermore, based on the signal fluctuation segments in the second solidified queue, the segment index identifier and time position of the signal fluctuation segment are read to locate the previous and next reliable positioning segments of the signal fluctuation segment; abnormal positioning changes in the signal fluctuation segment are removed from the positioning sequence, and a vacancy mark is written at the removal position, while retaining the end position of the previous reliable positioning segment, the start position of the next reliable positioning segment, and the original time position of the signal fluctuation segment.

[0074] Based on the original time position of the signal fluctuation segment, a supplementary chain position is established between the previous and subsequent credible positioning segments. The end position of the previous credible positioning segment is used as the starting point of the supplementary chain, and the beginning position of the subsequent credible positioning segment is used as the ending point of the supplementary chain. Continuous positioning points are written at the supplementary chain positions, and the segment index identifiers of the continuous positioning points are matched with the original time positions of the signal fluctuation segments. This ensures that the abnormal positioning changes that are removed do not participate in the landslide deformation combination. At the same time, the previous credible positioning segment, continuous positioning points, and subsequent credible positioning segments are connected in chronological order to obtain a continuous positioning sequence.

[0075] S4.5: For the windowed re-irrigation segments retained in the continuous positioning sequence, check the time connection with the adjacent synchronous observation segments. When the windowed re-irrigation segments have not yet been classified into the synchronous response marker and the response missing marker, retain the windowed re-irrigation segments and combine them to form a deformation classification segment set.

[0076] S5: The sudden slip sequence, slow creep sequence and continuous localization sequence in the deformation classification fragment set are temporally interlocked to form landslide deformation.

[0077] S5.1: Mark the sudden slip sequence in the deformation classification fragment set as the fast insertion fragment, mark the slow creep sequence as the trend embedding fragment, and mark the continuous localization sequence as the localization main line fragment, and combine them to generate the sequence splicing benchmark; Furthermore, sudden slip sequences, slow creep sequences, and continuous positioning sequences are extracted from the deformation classification fragment set. The positions of the sudden slip sequences, slow creep sequences, and continuous positioning sequences in the deformation classification fragment set are determined according to the fragment index identifier and time order, respectively. A fast insertion fragment tag is written for the sudden slip sequence, a trend embedding fragment tag is written for the slow creep sequence, and a positioning main line fragment tag is written for the continuous positioning sequence, forming a fragment role tag record.

[0078] Establish a correspondence between the occurrence position of the rapid insertion segment, the continuation position of the trend embedding segment, and the continuous position of the positioning main line segment in chronological order; using the positioning main line segment as the combination benchmark, map the rapid insertion segment and the trend embedding segment to the time position in the positioning main line segment to generate a sequence splicing benchmark.

[0079] S5.2: Based on the sequence splicing benchmark, the continuous processing positioning sequence is laid out as the basic positioning main line, the position intercepted by signal fluctuation is retained, and the fast insertion position and trend embedding position are set to generate the positioning main line; Furthermore, based on the sequence splicing benchmark, the continuous positioning sequence is read and arranged in chronological order according to the segment index identifier. The end and start positions of adjacent positioning segments in the continuous positioning sequence are connected sequentially, and the position where the signal fluctuation is intercepted is reserved as a vacant position. A signal fluctuation interception mark is written in the vacant position to form the basic positioning main line.

[0080] Obtain the fast insertion segment and trend embedding segment from the sequence splicing benchmark, and locate the corresponding time position in the basic positioning master line according to the segment index identifiers corresponding to the fast insertion segment and trend embedding segment; write the time position corresponding to the fast insertion segment into the fast insertion position marker, and write the time position corresponding to the trend embedding segment into the trend embedding position marker, so that the basic positioning master line simultaneously retains the signal fluctuation interception position, fast insertion position and trend embedding position, and generates the positioning master line.

[0081] It should be noted that the signal fluctuation interception mark refers to the identification mark written to the location where the signal fluctuation is intercepted. It is used to record that the location change at the corresponding location has been determined to be a signal fluctuation segment and does not participate in the actual landslide deformation combination, but the corresponding time and location are retained for continuous processing.

[0082] S5.3: Embed the slow creep sequence into the trend embedding position in the positioning main line, perform continuous displacement change replacement, and write the slow change continuation mark on the replaced positioning segment to generate a trend-embedded positioning sequence. Furthermore, based on the trend embedding position in the positioning main line, the time range corresponding to the slow creep sequence is obtained, and the original positioning segment within the same time range is determined in the positioning main line; the original positioning segment within the same time range is marked as the replaced segment, and the replaced segment is replaced by the slow creep sequence, so that the slow creep sequence replaces the positioning change of the original positioning segment, forming the creep replacement segment.

[0083] Connect the starting position of the creeping replacement segment with the previous positioning segment in the positioning main line, connect the ending position of the creeping replacement segment with the next positioning segment in the positioning main line, and confirm the slow change continuation of the creeping replacement segment so that the slow creeping sequence maintains a continuous change relationship in the positioning main line, thereby generating a trend-embedded positioning sequence.

[0084] S5.4: Insert the sudden slip sequence into the rapid insertion position in the trend-embedded positioning sequence, and connect it with the adjacent positioning segment in the trend-embedded positioning sequence to form a landslide deformation.

[0085] Furthermore, based on the rapid insertion position in the trend-embedded positioning sequence, the occurrence time and insertion range corresponding to the sudden slip sequence are matched, and adjacent positioning segments within the same time range are defined in the trend-embedded positioning sequence; the rapid insertion position is freed up from the trend-embedded positioning sequence to make room for insertion, and the abrupt displacement change in the sudden slip sequence is embedded into the insertion space, so that the sudden slip sequence replaces the original positioning change segment within the same time range, forming a sudden insertion segment.

[0086] The starting position of the sudden insertion segment is connected to the previous positioning segment in the trend-embedded positioning sequence, and the ending position of the sudden insertion segment is connected to the next positioning segment in the trend-embedded positioning sequence. The time sequence of the sudden insertion segment and the adjacent positioning segment is corrected according to the segment index identifier, so that continuous positioning changes, slow creep changes and sudden slip changes are connected in the same positioning sequence and combined to form landslide deformation.

[0087] It should be noted that the original positioning change segment refers to the positioning change segment retained by the continuous-processing positioning sequence in the trend-embedded positioning sequence that is located within the same time range as the sudden slip sequence but has not yet been embedded into the sudden slip sequence.

[0088] In summary, this invention achieves a collaborative trusted state by coupling trusted markers with MEMS responses, which is used to distinguish between fast changes, slow changes, and fluctuations, thereby improving the reliability of initial deformation judgment. Through gating and differential solidification, fast slip, slow creep, and signal fluctuations are processed and embedded for output separately, reducing misjudgments and enhancing sequence continuity, as well as improving the overall stability, continuity, and traceability of monitoring results.

[0089] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A landslide deformation monitoring method based on GNSS and MEMS collaborative identification, characterized in that, include: Real-time acquisition of navigation satellite positioning data and microelectromechanical response data from landslide monitoring points, time alignment, and generation of synchronous observation segments; The navigation satellite positioning data in the synchronous observation segment is credibly labeled to obtain the credibly labeled segment. The credible response is coupled with the microelectromechanical response data in the synchronous observation segment to form a cooperative credible state. Candidate states are extracted from the collaborative trusted state. Synchronous response locking is performed on fast-changing candidate states, response missing retention is performed on slow-changing candidate states, signal fluctuation interception is performed on signal fluctuation candidate states, and windowed backfeed is performed on verification candidate states to form gated shunting segments. Based on the gated diversion segment, the fast deformation segment is responded and bound, the slow deformation segment is continuously accumulated and bound, the intercepted signal fluctuation segment is processed into a sequence of continuous processing, and the windowed recharge segment is reviewed and retained to form a deformation classification segment set. The sudden slip sequence, slow creep sequence, and continuous localization sequence in the deformation classification fragment set are temporally interlocked to form landslide deformation. 2.The landslide deformation monitoring method based on GNSS and MEMS collaborative identification according to claim 1, wherein, The generation of synchronous observation segments is as follows: Navigation satellite positioning data and microelectromechanical response data are acquired through GNSS and MEMS respectively, forming a dual-source time-series data stream; Using the positioning output time in the dual-source time-series data stream as the binding reference time, the response data segments are extracted, numbered, and bound to form positioning time data segments; The navigation satellite positioning data and positioning time data segments are bound to a reference time, and synchronous observation segments are encapsulated. A segment index identifier is written for each synchronous observation segment to generate a synchronous observation segment.

3. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 1, characterized in that, The acquisition of the trusted marker fragment is specifically as follows: Navigation satellite positioning data is obtained from synchronous observation segments. When the conditions of continuous output, stable change and available positioning solution are met, a positioning trust mark is written. When the navigation satellite positioning data shows jumps, discontinuities and abnormal positioning solutions, a positioning untrust mark is written, forming a positioning mark data segment. The location marker data segments are layered into trusted segments and arranged according to the time order of synchronous observation segments to generate trusted marker segments.

4. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 3, characterized in that, The formation of a collaborative and trustworthy state is specifically as follows: Based on trusted marker fragments, response evidence is attached. Synchronization response markers and response missing markers are attached to trusted location fragments and suspected fluctuation fragments, respectively, to form a response evidence record. Based on the response evidence record, the trusted location segment is matched with the synchronous response marker and response missing marker within the same time range, and fast-changing candidate state and slow-changing candidate state are generated according to the matching relationship to form a trusted location candidate record. Obtain response evidence records corresponding to suspected fluctuation segments in the trusted positioning candidate records within the same time range, retain suspected fluctuation segments with missing response markers as signal fluctuation candidate states, and transfer suspected fluctuation segments with synchronization response markers to the review candidate state to form candidate state records; Candidate state records are arranged according to the time sequence of synchronous observation segments, and states are merged. Adjacent different candidate states are retained as independent state segments to form a cooperative and reliable state.

5. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 1, characterized in that, The specific steps for retaining missing responses to slowly changing candidate states are as follows: Candidate states are extracted from the cooperative trust state, and the fast-changing candidate states, slow-changing candidate states, signal fluctuation candidate states, and verification candidate states in the cooperative trust state are organized into a candidate state sequence. Based on the fast-changing candidate states in the candidate state sequence, the navigation satellite positioning data corresponding to the fast-changing candidate states are bound to the microelectromechanical response data in the same time range according to the binding reference time, and the fast-changing candidate states are updated to fast-changing transformation segments to form a fast-changing locking sequence. Based on the slow-change candidate states in the fast-change locking sequence, navigation satellite positioning data with reliable positioning and missing response markers are separated from the signal fluctuation interception path, and continuous reliable positioning change segments are retained along the time sequence of synchronous observation segments to form a fast-slow diversion sequence.

6. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 5, characterized in that, The formation of the gated shunt segment is as follows: Signal fluctuation interception is performed based on the candidate states of signal fluctuations in the fast and slow flow splitting sequence to form a fluctuation interception sequence; The candidate states for verification in the fluctuation interception sequence are subjected to windowed backfeed and combined to form a gated diversion segment.

7. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 6, characterized in that, The specific steps for responding to and binding the rapidly deformable fragments are as follows: Based on the routing attributes carried in the gated routing segments, the gated routing segments are extracted and a routing solidification queue is formed. The fast deformation segments in the diversion and solidification queue are responded to and bound to solidify, and the corresponding positioning change segments are solidified into sudden slip segments to form the first solidification queue. Based on the slow deformation segments in the first curing queue, continuous cumulative curing is performed to solidify the continuously positioning and changing segments into slow creeping segments, thus forming the second curing queue.

8. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 1 or 7, characterized in that, The set of deformable classification segments is formed as follows: The signal fluctuation segments in the second solidified queue are processed to achieve sequence continuity. The signal fluctuation segments are isolated and the corresponding positions are filled in to obtain a continuous positioning sequence. For the windowed re-irrigation segments retained in the continuous positioning sequence, the time connection with the adjacent synchronous observation segments is checked. When the windowed re-irrigation segments have not yet been classified into the synchronous response marker and the response missing marker, the windowed re-irrigation segments are retained and combined to form a deformation classification segment set.

9. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 1, characterized in that, The temporal interlocking combination is performed as follows: The sudden slip sequence in the deformation classification fragment set is marked as a fast insertion fragment, the slow creep sequence is marked as a trend embedding fragment, and the continuous localization sequence is marked as a localization main line fragment. These are combined to generate a sequence splicing benchmark. Based on the sequence splicing benchmark, the continuous processing of the positioning sequence is laid out as the basic positioning main line, the position intercepted by signal fluctuations is retained, and the fast insertion position and trend embedding position are set to generate the positioning main line.

10. The landslide deformation monitoring method based on GNSS and MEMS collaborative discrimination as described in claim 9, characterized in that, The landslide deformation is specifically as follows: The slow creep sequence is embedded into the trend embedding position in the positioning main line, and continuous displacement change is performed to replace it. The slow change continuation mark is written into the replaced positioning segment to generate the trend-embedded positioning sequence. The sudden slip sequence is inserted into the rapid insertion position in the trend-embedded positioning sequence, and then connected with the adjacent positioning segments in the trend-embedded positioning sequence to form a landslide deformation.

Images