Method and system for analyzing four-component borehole strain observation data

Abstract

A method and system for analyzing four-component borehole strain observation data, includes phenomenon determination and phenomenon analysis. Observation phenomena are determined from a strain observa

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of four-component borehole strain data processing, and particularly relates to a method and system for analyzing four-component borehole strain observation data. BACKGROUND

[0002] The study of borehole strain observation phenomena and their reflection of tectonic stress changes are still limited to disturbance exclusion for a specific phenomenon. Scientifically processing and interpreting borehole strain observation data is a critical scientific issue that needs to be urgently addressed in both geodynamics and seismology.

[0003] The process of processing and analysis of the borehole strain observation data is as follows: the observation data usually requires data pre-processing of reliability analysis, accuracy analysis, data calibration, etc., followed by filtering analysis and time frequency analysis of specific phenomena, whether the data obtained from the analysis respond to a tectonic stress, a non-tectonic stress or a residual stress in response to the observed phenomena, and finally evolution of a distribution law of the tectonic stress is studied through tectonic stress analysis methods.

[0004] Compared with other observation data, the borehole strain observation data is obtained by long-term observation from one monitoring station, which monitors far more geological phenomena than seismometers, radar, satellites, etc. In contrast, the existing time frequency analysis of the borehole strain observation data cannot explain the affecting factors of the complex observation phenomena. Therefore, the improvement and research on data analysis methods not only provide an important reference for improvement in monitoring of a crustal stress tensor and a research method system of numerical earthquake prediction, but also have important scientific significance and application value on improving the understanding of a regional crustal stress field and features and laws of fault stress evolution, relative ground stress big data processing and automated analysis, and understanding of earthquake dynamics process and earthquake breeding. SUMMARY

[0005] An objective of the present disclosure is to provide a method and system for analyzing four-component borehole strain observation data, which may achieve qualitative and quantitative analysis of the borehole strain observation data.

[0006] In order to achieve the above objective, the present disclosure provides the following technical solutions:

[0007] A method for analyzing four-component borehole strain observation data includes:

[0008] obtaining a four-component borehole strain observation data segment in a target time period;

[0009] performing feature analysis on the four-component borehole strain observation data segment, and forming all features obtained from the feature analysis into a feature index vector;

[0010] determining observation phenomena in the target time period by searching a feature index vector-observation phenomenon comparison table according to the feature index vector;

[0011] determining abnormal phenomena from the strain observation data segment according to features of the abnormal phenomena determined in an experiment;

[0012] determining a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively; and

[0013] constructing a borehole strain observation data evaluation system through a network analysis method according to the strain observation data segment, replacing an importance scale value of the observation phenomena compared to each factor in the borehole strain observation data evaluation system in a judgment matrix with the ratio of the observation phenomena, and replacing an importance scale value of the abnormal phenomena compared to each factor in the borehole strain observation data evaluation system with the ratio of the abnormal phenomena, to obtain a weight of each factor in the borehole strain observation data evaluation system as a ratio of each factor in the strain observation data segment.

[0014] A system for analyzing four-component borehole strain observation data includes:

[0015] an observation data obtaining module configured to obtain a four-component borehole strain observation data segment in a target time period;

[0016] a feature analysis module configured to perform feature analysis on the four-component borehole strain observation data segment, and form all features obtained from the feature analysis into a feature index vector;

[0017] an observation phenomenon determination module configured to determine observation phenomena in the target time period by searching a feature index vector-observation phenomenon comparison table according to the feature index vector;

[0018] an abnormal phenomenon determination module configured to determine abnormal phenomena from the strain observation data segment according to features of the abnormal phenomena determined in an experiment;

[0019] a ratio computation module configured to determine a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively; and

[0020] a weight computation module configured to construct a borehole strain observation data evaluation system through a network analysis method according to the strain observation data segment, replace importance of the observation phenomena compared to each factor in the borehole strain observation data evaluation system in a judgment matrix with the ratio of the observation phenomena, and replace importance of the abnormal phenomena compared to each factor in the borehole strain observation data evaluation system with the ratio of the abnormal phenomena, to obtain a weight of each factor in the borehole strain observation data evaluation system as a ratio of each factor in the strain observation data segment.

[0021] According to particular embodiments provided in the present disclosure, the present disclosure provides the following technical effects:

[0022] The present disclosure provides a method and system for analyzing four-component borehole strain observation data, which mainly includes phenomenon determination and phenomenon analysis. Observation phenomena are determined by performing feature analysis on a strain observation data segment. During phenomenon analysis, importance evaluation values of corresponding elements in a judgment matrix are directly replaced with a ratio of the observation phenomena and a ratio of abnormal phenomena, to obtain a ratio of each affecting factor in the strain observation data, so as to achieve qualitative and quantitative analysis of the borehole strain observation data. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] To describe the technical solutions in embodiments of the present disclosure or in the prior art more clearly, the accompanying drawings required in the embodiments are briefly described below. Apparently, the accompanying drawings in the following description show merely some embodiments of the present disclosure, and other drawings can be derived from these accompanying drawings by those of ordinary skill in the art without creative efforts.

[0024] FIG. 1 is a diagram showing relative positions during borehole strain observation data processing through a method for analyzing four-component borehole strain observation data according to an embodiment of the present disclosure and a conventional data processing method;

[0025] FIG. 2 is a flowchart of a method for analyzing four-component borehole strain observation data according to an embodiment of the present disclosure;

[0026] FIG. 3 is a schematic diagram of a four-component borehole strain observation data segment in a target time period according to an embodiment of the present disclosure;

[0027] FIG. 4 is a schematic diagram of an existing dual bushing model for borehole strain observation;

[0028] FIG. 5 is a schematic diagram of a borehole strain observation data evaluation system for simplified coseismic observation data according to an embodiment of the present disclosure; and a is a network structure diagram of the borehole strain observation data evaluation system, and b is a schematic diagram of grading evaluation indexes of the borehole strain observation data evaluation system;

[0029] FIG. 6 is a schematic diagram of a ratio of each affecting factor in a target evaluation data segment according to an embodiment of the present disclosure; and a is a diagram showing a result of weights of grading evaluation indexes of a borehole strain observation data evaluation system, and b is a diagram showing a result of target data segment visualization evaluation; and

[0030] FIG. 7 is a block diagram of a principle of a method for analyzing four-component borehole strain observation data according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] The technical solutions of the embodiments of the present disclosure are clearly and completely described below with reference to the drawings. Apparently, the described embodiments are merely some embodiments rather than all embodiments of the present disclosure. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0032] An objective of the present disclosure is to provide a method and system for analyzing four-component borehole strain observation data, which may achieve qualitative and quantitative analysis of the borehole strain observation data.

[0033] To make the above objectives, features, and advantages of the present disclosure clearer and more comprehensible, the present disclosure will be further described in detail below with reference to the accompanying drawings and the specific examples.

[0034] FIG. 1 shows a position of the method of the present disclosure in existing borehole strain observation data processing methods. Contents in a rectangular box are main research contents of the method of the present disclosure. Contents outside the rectangular box are research contents of the existing methods. The method for analyzing four-component borehole strain observation data provided in the present disclosure mainly includes phenomenon determination and phenomenon analysis.

[0035] As shown in FIG. 2, a method for analyzing four-component borehole strain observation provided in an embodiment of the present disclosure includes:

[0036] Step 1: obtain a four-component borehole strain observation data segment in a target time period.

[0037] FIG. 3 shows a four-component borehole strain observation data segment including four strain observation data segments that are borehole strain observation data measured in four directions. An angle between the observation directions of strain 1 and strain 3 is 90° during observation. An angle between the observation directions of strain 2 and strain 4 is 90° during observation.

[0038] After a target analysis observation data segment is obtained, data inspection is performed first. By means of self-inspection and calibration, it is determined that a change of crustal stress is observed by means of the data (not instrument failure).

[0039] The borehole strain observation data should satisfy the following self-inspection condition:

[0040] Ae1 + Ae'" = Ae11 + Aeiv (1)

[0041] The self-inspection condition is generally be referred as “ 1 +3=2+4” for short. In response to determining that the observation data does not satisfy this condition, it is indicated that at least one or more components have problems of measurements or calibrations. Ae 1. Ae1'. demand Aew are borehole strain observation data in the four directions at the same time.

[0042] Calibration: during actual observation data processing, it is also necessary to consider an effect of a double-ring sleeve on observation, strain gauge component reading (FIG. 4), probe reading, and an effect of a steel cylinder and cement.

[0043] When there are horizontal main strains Aer and Ae2 at a distance, a relative change of a hole diameter of a borehole in a direction 0 is:

[0044] Ae9 = A(Aet + Ae2) + B(Aer — Ae2)cos2(0 — <p) (2)

[0045] In the formula, As is a value given by actual observation in an angle 0, q> is an azimuth angle of Aen and absolute calibration of the borehole gives values of A and B. It can be seen from formula (2) that results of four-component borehole strain observation still satisfy the selfinspection formula (1). Furthermore, the values of A and B may also be computed by means of complicated formulas with physical parameters of each layer and a hole size of each layer.

[0046] Step 2: perform feature analysis on the four-component borehole strain observation data segment, and form all features obtained from the feature analysis into a feature index vector.

[0047] Illustratively, a detailed implementation process of this step includes: obtain a time range and a sampling rate of the strain observation data segment, and determine a time range feature representation mi and a sampling rate feature representation m4; determine a frequency range of the strain observation data segment according to the time range of the strain observation data segment, and determine a frequency range feature representation m3; convert the strain observation data segment into a crustal stress tensor, extract an amplitude range from the crustal stress tensor, and determine an amplitude range feature representation m2; compare the strain observation data segment with observation data of a seismic network in the same time range, and first determine an observation result whether a phenomenon ms exists; and form the time range, the amplitude range, the frequency range, the sampling rate and the phenomena into features of the strain observation data segment, and construct the feature index vector as D=D(mi, m2, m3, nu, ms), where D is the feature index vector, and D(mi, m2, m3, m4, ms) is a feature index vector function.

[0048] A computation process of converting the strain observation data segment into the crustal stress tensor is as follows:

[0049] A least-squares solution of a plane crustal stress is obtained directly from observation data of four components:

[0050] Ae^As"1,Ae'^Ae1:

[0051] = — (df^ + Ae'" + Ae" + Ae,v) + — (Ae^ — Ae"")2 + (Ae^ — Ae" )2 (2)

[0052] 02 = — {Ae1 + Ae1" + Ae11 + Aeiv) - — MAe1 - Ae1")2 + {Ae1v - Ae11)2 (3) 8^4 45 1 1 Ae11—Aeiv

[0053] ¢ = 6 - - arctan- {—----) (4) * As —As11

[0054] The observation data segments are subjected to feature analysis (mainly including mi of phenomenon duration, m2 of maximum amplitude range crustal stress conversion, m3 of frequency range, and m4 of data sampling rate). For the target data segment in the example of FIG. 3. the data segment has features: (a time range of minute level, an amplitude range of about 200 Pa (crustal stress conversion), a frequency range of 5 Hz-20 Hz, a data sampling rate of 100 times / second, and a specific phenomenon (earthquake) is observed). The features can be sequentially taken as a feature index vector D=D(1, 1, 1, 1), which points to the earthquake (phenomenon ms) in a comparison table, that is, the data is affected by the earthquake. Similarly, since seismic waves also affect observation of other frequency bands, D=D(2, 1, 2, 2) also indexes a seismic phenomenon. Although an m5 index phenomenon is consistent, processing methods of different frequency bands are different, which provides the basis for further data processing.

[0055] As an example, as shown in Table 1, the feature index vector has only four dimensions. mi=2 indicates that a target observation time range is in the minute level. m2=l indicates an amplitude range of about 200 Pa. m3=l indicates a frequency range of 5 Hz-20 Hz. m4=l indicates a sampling rate of 100 times / second. me is a target data representation.

[0056] Table 1 Feature index corresponding table Value range Index value mi 0-1 min 1 Several seconds-24 hours 2 m2 0-1000 1 m3 10-100 1 0.01-50 2 rru 100 1 1 2 m6 Minimal ratio 1 Part ratio 2 All data 3

[0057] An order and indicated meanings of ml, m2, m3, m4 and m5 may be adjusted accordingly according to analysis data or study subjects. By searching a feature representation corresponding table, it is determined whether the phenomenon m5 has been quantitatively studied.

[0058] Step 3 : determine observation phenomena in the target time period by searching a feature index vector-observation phenomenon comparison table according to the feature index vector.

[0059] The feature index vector-observation phenomenon comparison table is shown in Table 2.

[0060] Table 2 Feature index vector-observation phenomenon comparison table mi m2 m3 m4 m5 Phenomena corresponding to ms 1 1 1 1 1 Instrument noise 1 1 1 1 2 Coseismic decoupling 1 1 1 1 3 Earthquake 1 1 1 1 4 Rock micro-fracture 1 1 1 2 / No phenomenon 1 1 2 1 1 Instrument noise 1 1 2 1 2 Decoupling 1 1 2 1 3 Earthquake 1 1 2 1 4 Solid tide 1 1 2 2 1 Decoupling 1 1 2 2 2 Earthquake 1 1 2 2 3 Solid tide 2 1 1 1 1 Earthquake 2 1 1 1 2 Solid tide 2 1 1 2 1 Earthquake 2 1 1 2 2 Solid tide 2 1 2 1 1 Earthquake 2 1 2 1 2 Solid tide 2 1 2 2 1 Earthquake 2 1 2 2 2 Solid tide

[0061] The feature index vector-observation phenomenon comparison table further provides a classification of crustal stress corresponding to each observation phenomenon and a computation grade of importance of analysis phenomena, ms is determined by means of the features mi, m2, m3, and m4 of the observation phenomena, and finally me is determined, that is, an order of obtaining quantitative data preferentially is determined. Grade 1 is higher than grade 2, and grade 2 is higher than grade 3.

[0062] By searching the feature index vector comparison table, a phenomenon observed according to a target observation curve is judged according to features of the target observation curve. An m& comparison table is used to determine the grade of possible observed phenomena. For example, seismic waves are obtained in the example and correspond to a phenomenon grade of 2, which indicates that the seismic waves are observed by means of some data. Therefore, in a subsequent analysis process, besides the less abnormal phenomena in all the observation data, an effect of seismic phenomena on the observation curve is preferentially analyzed.

[0063] Step 4: determine abnormal phenomena from the strain observation data segment according to features of the abnormal phenomena determined in an experiment.

[0064] The features of the abnormal phenomena are obtained according to abnormal observation phenomena which have been quantitatively studied.

[0065] The abnormal phenomena may be coseismic decoupling, sudden jump before and after the earthquake. The grade of the abnormal phenomena is higher than that of the seismic waves.

[0066] Phenomenon determination is achieved from step 1 to step 4.

[0067] Step 5 : determine a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively in the target time period.

[0068] In the target example, a proportion of data points of a seismic phenomenon to data points of the observation curve is about 55.7%. Therefore, in the observation curve, the importance of the earthquake compared with a different phenomenon except the earthquake is 1.26 first, that is, compared with a different phenomenon, the earthquake is 1.26 times more important than a different phenomenon in the target observation data segment to be analyzed. Such a different phenomenon mainly is a solid tide, which affects all the observation data.

[0069] The ratio of the observation phenomena includes a ratio of the observation phenomena to each observed phenomenon in the feature index vector-observation phenomenon comparison table, and a ratio of the observation phenomena to the abnormal phenomena. The ratio of the abnormal phenomena includes a ratio of the abnormal phenomena to each observation phenomenon in the feature index vector-observation phenomenon comparison table.

[0070] Since a network analysis method needs to be used in the subsequent steps, when a judgment matrix is constructed in the network analysis method, each element in the matrix represents relative importance of two factors, and a value of each element is usually determined through the 1-9 scale method. The 1-9 scale method is obtained through empirical subjective evaluation from experts. The quantitative importance comparison provided in the feature analysis of the present disclosure is an important constraint in step 6, and serves as a supplement to the 1-9 scale method, strongly constraining a final evaluation result.

[0071] The 1-9 scale method is described in Table 3.

[0072] Table 3 Scales and meanings of judgment matrix____________________________________ Scale Definition and description Sij=l Factor Si and factor Sj have a same effect degree on data under specific affecting factors Sij=3 Factor Si is slightly more important than factor Sj sy=5 Factor Si is significantly more important than factor Sj Sij=7 Factor Si is much more important than factor Sj Sij=9 Factor Si is extremely more important than factor Sj Sij=2n, n=l,2,3,4 The importance of factor Si relative to Sj is between Sij=2n-1 and Sij=2n+1 Sjj=l / n, n=l,2,...,9 If and only if sy=n

[0073] Step 6: construct a borehole strain observation data evaluation system through a network analysis method according to the strain observation data segment, replace importance of the observation phenomena compared to each factor in the borehole strain observation data evaluation system in a judgment matrix with the ratio of the observation phenomena, replace importance of the abnormal phenomena compared to each factor in the borehole strain observation data evaluation system with the ratio of the abnormal phenomena, and supplement a qualitative importance ratio provided by a 1-9 scale method or other relevant methods for factors having no quantitative importance ratio, to obtain a weight of each factor in the borehole strain observation data evaluation system as a value of a weight vector of each factor in the strain observation data segment.

[0074] 1) Basic principles of the network analysis method are as follows. The network analysis model divides the system into: the control layer and the network layer. The control layer includes a decision-making target and decision-making criteria. At least one target is provided, but no decision-making criteria may be provided. The network layer is composed of element groups. These element groups are controlled by the control layer. The element groups and the internal elements are interdependent and interact with each other to form a network structure.

[0075] A typical analytic network process (ANP) model is shown in FIG. 7. The borehole strain observation data evaluation system includes a control layer and a network layer. The control layer includes a general target and a plurality of criteria. The general target is strain observation data , and the plurality of criteria include a tectonic stress, a non-tectonic stress, a residual stress and other criteria that may be increased or decreased according to actual research data. The network layer includes a coseismic cluster, a solid tide cluster, a site effect cluster, an instrument parameter cluster, an abnormal tectonic cluster, and other element clusters that need to be increased emphatically according to the actual research data or analysis needs. The coseismic cluster includes P waves, S waves and other sub-elements that are increased or decreased according to the analysis needs. The solid tide cluster include M2 waves, 01 waves and other sub-elements that are increased or decreased according to station location and the analysis needs. The site effect cluster includes terrain, rivers, faults, and sub-elements that are increased or decreased according to analysis needs of the station location, basic geology, meteorology, hydrology, temperature, and human factors. The instrument parameter cluster includes cement creep, coseismic decoupling, and other sub-elements that are increased or decreased according to an analysis time period.

[0076] The element cluster is Ch, h=l, 2, ..., N. Ch has nk elements, respectively denoted as ehi, eh2,..., ehnk. A priority vector obtained by pairwise comparison represents an effect of a given series of elements in a certain component on other elements in a system. The priority vectors obtained from pairwise comparison matrices are part of a super-matrix column. A super-matrix reflects a degree to effects of elements on a left side of the matrix on top elements of the matrix. Formulas (5) and (6) show the super-matrix and its sub-matrix Wij. A right side of the element cluster Ch shows all priority vectors computed according to its upper layer criterion nodes.

[0077]

[0078] ,.,00 _ / ,.,01

[0079] «u «12 -e^ «21 «22 «2n. eWi eN2 -eNn «n Cl «1», / Wu ^12 Wiw\ C2 «21 «2,1 2 W21 W^22 WA «21 «2ii2 wWi ^2 W^1 ) ,,.02) ... OndX Wil \ w ij = ,,, Oi Wi2 \ ini ) ,,,00) Wi2 ) w°'2) (J nt) w.^ J il w. / ini / (5) (6) ,( / 0 «.....( / =1,2. nj) is a normalized weight vector.

[0080] According to a number of the criteria N in the control layer, m super-matrices are provided, which are all non-negative matrices, to form column-normalized sub-matrices but the super matrix IF is not normalized. With N, as a criterion, the importance of each element cluster under Ni to a criterion Q ( / = 1, 2, ..., N) is compared. Normalized feature vectors are obtained after io pairwise comparison of Q. After combination, a matrix A=aij may be obtained, z=l, 2, ..., N, and j=l, 2, ..., N. Weighting the super-matrix W results in a weighted super-matrix = , wij - aij / =1, 2,.,., W and j=l, 2, ..., N. A sum of W columns is 1, which is a column random matrix.

[0081] A weight vector of Ni, z=l, 2, ..., m in the control layer to the target is set to be ivCO = — , and a weight vector of each element cluster in the network layer to the criterion layer is set to be ws - (¾ >wis ''"’wns ) s - —such that a weight vector of each solution to the target is obtained by means of combination of w^w^Ci = 1,2, —, m / and finally a weight vector of the network layer to the target layer is obtained as follows:

[0082] ^ = (7^,^ / -,7^ (7)

[0083] Weights can be computed according to this computation process for more hierarchical problems.

[0084] 2) Hierarchical order and consistency test are as follows. The judgment matrix may be constructed through the 1-9 scale method in Table 1. A maximal eigenvalue of each judgment matrix and a corresponding feature vector W are computed through a sum-product method. The feature vectors are normalized to obtain a weight value of each factor for the last layer.

[0085] A main process of the sum-product method for computing the maximal eigenvalue and the feature vectors is as follows: first, the constructed judgment matrices X are normalized in columns to obtain a matrix — O^Anx-ss ; matrices Y are added in rows to obtain a matrix ■ matrices Z are normalized to obtain the feature vector — (¾¾ *** f ; and finally the maximal eigenvalue is computed, and the feature vector corresponding to the maximal eigenvalue is a corresponding weight value. Specifically: - : ~ — 1,2,,-

[0086] (8) Q = = 1,2,---

[0087] J ' (9) W; = --, (l J = 1.,2,-- - ,

[0088] (10) ___ J- "Jj .v-

[0089] R (11)

[0090] Because of complexity of the decision problems and one-sidedness of the recognition for objects, after the weight value indicating the importance order is computed, it is necessary to test the consistency of each judgment matrix, and the test is performed mainly by means of consistency indexes, random consistency indexes and consistency ratios, where the formula is as follows:

[0091] CR = CI / R1 (12)

[0092] Cl = (Amas — n) / (n — 1) (13)

[0093] In the formula, CR-random consistency ratio;

[0094] Ci-consistency index;

[0095] Ri-random consistency index, values are shown in table 4 and related to a matrix order number n;

[0096] -maximal eigenvalue of the judgment matrix; and

[0097] when CR is less than 0.1, it may be construed as that the constructed judgment matrix passes the consistency test, and otherwise, it is necessary to reconstruct the judgment matrix. [0C 98] able 4 Average random consistency index RI of udgment matrix n 1 2 3 4 5 6 7 8 9 RI 0 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45

[0099] A borehole strain observation data evaluation system is constructed through the network analysis method according to the borehole strain observation data. According to the importance judgment of other pairwise sub-elements under the condition of sub-elements, the super-matrix is constructed, and a scale value obtained through the 1-9 scale method is replaced with the ratio of the observation phenomena and the ratio of the abnormal phenomena, to obtain a weight of each factor in the borehole strain observation data evaluation system as a ratio of each factor in the strain observation data segment.

[0100] During pairwise judgment in the process of constructing the super-matrix, one of the features of this method is to replace the subjective importance evaluation with the 1-9 scale method according to a quantitative ratio of important data, and finally a borehole strain observation phenomenon evaluation method system combining quantitative and qualitative analysis is obtained.

[0101] Phenomenological analysis is achieved from step 5 to step 6.

[0102] Feasibility of an algorithm is verified through preliminary experiments and computation. With a simple network structure established in FIG. 7 as an example, by further simplifying the network structure ((a) in FIG. 5), observation data, from Youyu Station at 1: 48-1: 53, of M4.3 earthquake in Pingshan County (38.35°N, 113.73°E), Shijiazhuang City, Hebei Province, China at 1: 49 on October 3, 2022 are analyzed. A sampling rate of the observation data is 100 times / s, and a seismic process is well recorded. In order to further discuss a possible tectonic stress effect observed by the observation data, other tectonic clusters (Ot) that can only affect coseismic observations is added during network layer design, to represent some factors caused by tectonic stress that do not take into account the effect of seismic observations.

[0103] According to an analytic network process in (a) of FIG. 5 and grading evaluation indexes ((b) of FIG. 5), the corresponding super-matrix and corresponding sub-matrix are constructed. The study of coseismic data observation may be affected by other constructive and non-constructive factors. Computation results are shown in FIG. 6, (a) of FIG. 6 shows a hierarchical evaluation weight ratio of each level. In FIG. 3 of target evaluation data segment, about 55.7% of observation data points may be affected by earthquake, but after comprehensive evaluation, the impact on the entire observation data decreases to 49.34%. FIG. 6(b) shows a final visual evaluation result of the target data segment. The basis is provided for data collation and quantitative evaluation of specific phenomena and rule extraction of tectonic stress phenomena.

[0104] An embodiment of the present disclosure further provides a system for analyzing four-component borehole strain observation, including:

[0105] an observation data obtaining module configured to obtain a four-component borehole strain observation data segment in a target time period;

[0106] a feature analysis module configured to perform feature analysis on the four-component borehole strain observation data segment, and form all features obtained from the feature analysis into a feature index vector;

[0107] an observation phenomenon determination module configured to determine observation phenomena in the target time period by searching a feature index vector-observation phenomenon comparison table according to the feature index vector;

[0108] an abnormal phenomenon determination module configured to determine abnormal phenomena from the strain observation data segment according to features of the abnormal phenomena determined in an experiment;

[0109] a ratio computation module configured to determine a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively; and

[0110] a weight computation module configured to construct a borehole strain observation data evaluation system through a network analysis method according to the strain observation data segment, replace an importance scale value of the observation phenomena compared to each factor in the borehole strain observation data evaluation system in a judgment matrix with the ratio of the observation phenomena, and replace an importance scale value of the abnormal phenomena compared to each factor in the borehole strain observation data evaluation system with the ratio of the abnormal phenomena, to obtain a weight of each factor in the borehole strain observation data evaluation system as a ratio of each factor in the strain observation data segment.

[0111] The weight computation module includes a super-matrix construction sub-module, which is configured to compare relative importance degrees of pairwise sub-elements under different sub-element conditions, and perform scoring through a 1-9 scale method or a 1-3 scale method. For quantitative research contents, corresponding scale evaluation values are replaced with the ratio of the observation phenomena and the ratio of the abnormal phenomena.

[0112] In an example, the analysis system further includes a self-inspection module and a crustal stress change determination module.

[0113] The self-inspection module is configured to perform self-inspection on the four-component borehole strain observation data segment, and pass, in response to determining that Ae 7 + As111 = Ae n + AEn / is satisfied, the self-inspection, where Ae 7, Ae^, Ae111 and Ae^ are borehole strain observation data in four directions at the same time.

[0114] The crustal stress change determination module is configured to determine that a change of crustal stress is observed by means of the data in the strain observation data segment passing the self-inspection.

[0115] Illustratively, the feature analysis module includes a time range obtaining unit, a frequency range determination unit, an amplitude range extraction unit, a phenomenon determination unit, and a feature index vector construction unit.

[0116] The time range obtaining unit is configured to obtain a time range and a sampling rate of the strain observation data segment, and determine a time range feature representation mi and a sampling rate feature representation m4.

[0117] The frequency range determination unit is configured to determine a frequency range of the strain observation data segment according to the time range of the strain observation data segment, and determine a frequency range feature representation m3.

[0118] The amplitude range extraction unit is configured to convert the strain observation data segment into a crustal stress tensor, extract an amplitude range from the crustal stress tensor, and determine an amplitude range feature representation m2.

[0119] The phenomenon determination unit is configured to compare the strain observation data segment with observation data of a seismic network in the same time range, and first determine an observation result whether a phenomenon ms exists.

[0120] The feature index vector construction unit is configured to form the time range, the amplitude range, the frequency range, the sampling rate and the phenomena into features of the strain observation data segment, and construct the feature index vector as D=D(mi, m2, m3, nu, ms), where D is the feature index vector, and D(mi, m2, m3, nu, ms) is a feature index vector function.

[0121] Illustratively, the ratio computation module specifically includes an abnormal phenomenon ratio determination unit and an observation phenomenon ratio determination unit.

[0122] The abnormal phenomenon ratio determination unit is configured to determine a ratio of the number of the data points corresponding to the abnormal phenomena to a number of data points of the strain observation data segment as the ratio of the abnormal phenomena according to a priority analysis order of the abnormal phenomena higher than the observation phenomena.

[0123] The observation phenomenon ratio determination unit is configured to compute a ratio of the number of the data points corresponding to the observation phenomena to the number of the data points of the strain observation data segment as the ratio of the observation phenomena.

[0124] Illustratively, the borehole strain observation data evaluation system includes a control layer and a network layer.

[0125] The control layer includes a general target and a plurality of criteria, where the general target is the strain observation data, and the plurality of criteria include a tectonic stress, a nontectonic stress and a residual stress.

[0126] The network layer includes a coseismic cluster, a solid tide cluster, a site effect cluster, an instrument parameter cluster and an abnormal tectonic cluster, where the coseismic cluster includes P waves and S waves; the solid tide cluster includes M2 waves and 01 waves; the site effect cluster includes terrain, rivers and faults; and the instrument parameter cluster includes cement creep and coseismic decoupling.

[0127] In FIG. 7, contents in a smaller rectangular box are main research contents of the method of the present disclosure. Contents in a larger rectangular box are research contents of the system of the present disclosure.

[0128] Each embodiment of the present specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts between the embodiments may refer to each other. Since the system disclosed in an embodiment corresponds to the method disclosed in another embodiment, the description is relatively simple, and reference can be made to the method description.

[0129] Specific examples are used herein to explain the principles and implementations of the present disclosure. The foregoing description of the embodiments is merely intended to help understand the method of the present disclosure and its core ideas; besides, various modification modes may be made by a person of ordinary skill in the art to specific embodiments and the scope of application in accordance with the ideas of the present disclosure. In conclusion, the content of the present specification shall not be construed as limitations to the present disclosure.

Claims

1. A method for analyzing four-component borehole strain observation data, comprising:obtaining a four-component borehole strain observation data segment in a target time period;performing feature analysis on the four-component borehole strain observation data segment, and forming all features obtained from the feature analysis into a feature index vector;determining observation phenomena in the target time period by searching a feature index vector-observation phenomenon comparison table according to the feature index vector;determining abnormal phenomena from the strain observation data segment according to features of the abnormal phenomena determined in an experiment;determining a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively; andconstructing a borehole strain observation data evaluation system through a network analysis method according to the strain observation data segment, replacing an importance scale value of the observation phenomena compared to each factor in the borehole strain observation data evaluation system in a judgment matrix with the ratio of the observation phenomena, and replacing an importance scale value of the abnormal phenomena compared to each factor in the borehole strain observation data evaluation system with the ratio of the abnormal phenomena, to obtain a weight of each factor in the borehole strain observation data evaluation system as a ratio of each factor in the strain observation data segment.

2. The method for analyzing four-component borehole strain observation data according to claim 1, wherein after the obtaining a four-component borehole strain observation data segment in a target time period, the method further comprises:performing self-inspection on the four-component borehole strain observation data segment, and in response to determining that de J + Ae111 = Aeji + Ae1' is satisfied, passing the selfinspection, wherein Ae 1. Ae11 , Ae111 and Ae1' are borehole strain observation data in four directions at the same time; anddetermining that a change of crustal stress is observed by means of the data in the strain observation data segment passing the self-inspection.

3. The method for analyzing four-component borehole strain observation data according to claim 1, wherein the performing feature analysis on the four-component borehole strain observation data segment, and forming all features obtained from the feature analysis into a feature index vector comprises:obtaining a time range and a sampling rate of the strain observation data segment, and determining a time range feature representation mi and a sampling rate feature representation m4;determining a frequency range of the strain observation data segment according to the time range of the strain observation data segment, and determining a frequency range feature representation m3;converting the strain observation data segment into a crustal stress tensor, extracting an amplitude range from the crustal stress tensor, and determining an amplitude range feature representation m2;comparing the strain observation data segment with observation data of a seismic network in the same time range, and first determining an observation result whether a phenomenon ms exists; andforming the time range, the amplitude range, the frequency range, the sampling rate and the phenomena into features of the strain observation data segment, and constructing the feature index vector as D=D(mi, m2, m3, m4, ms), wherein D is the feature index vector, and D(mi, m2, m3, m4, ms) is a feature index vector function.

4. The method for analyzing four-component borehole strain observation data according to claim 1, wherein the determining a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively specifically comprises:determining a ratio of the number of the data points corresponding to the abnormal phenomena to a number of data points of the strain observation data segment as the ratio of the abnormal phenomena according to a priority analysis order of the abnormal phenomena higher than the observation phenomena; andcomputing a ratio of the number of the data points corresponding to the observation phenomena to the number of the data points of the strain observation data segment as the ratio of the observation phenomena.

5. The method for analyzing four-component borehole strain observation data according to claim 1, wherein the borehole strain observation data evaluation system comprises a control layer and a network layer;the control layer comprises a general target and a plurality of criteria, wherein the general target is the strain observation data, and the plurality of criteria comprise a tectonic stress, a nontectonic stress and a residual stress; andthe network layer comprises a coseismic cluster, a solid tide cluster, a site effect cluster, aninstrument parameter cluster and an abnormal tectonic cluster, wherein the coseismic cluster comprises P waves and S waves; the solid tide cluster comprises M2 waves and 01 waves; the site effect cluster comprises terrain, rivers and faults; and the instrument parameter cluster comprises cement creep and coseismic decoupling.

6. A system for analyzing four-component borehole strain observation data, comprising:an observation data obtaining module configured to obtain a four-component borehole strain observation data segment in a target time period;a feature analysis module configured to perform feature analysis on the four-component borehole strain observation data segment, and form all features obtained from the feature analysis into a feature index vector;an observation phenomenon determination module configured to determine observation phenomena in the target time period by searching a feature index vector-observation phenomenon comparison table according to the feature index vector;an abnormal phenomenon determination module configured to determine abnormal phenomena from the strain observation data segment according to features of the abnormal phenomena determined in an experiment;a ratio computation module configured to determine a ratio of the observation phenomena and a ratio of the abnormal phenomena according to numbers of data points corresponding to the observation phenomena and abnormal phenomena in the strain observation data segment respectively; anda weight computation module configured to construct a borehole strain observation data evaluation system through a network analysis method according to the strain observation data segment, replace an importance scale value of the observation phenomena compared to each factor in the borehole strain observation data evaluation system in a judgment matrix with the ratio of the observation phenomena, and replace an importance scale value of the abnormal phenomena compared to each factor in the borehole strain observation data evaluation system with the ratio of the abnormal phenomena, to obtain a weight of each factor in the borehole strain observation data evaluation system as a ratio of each factor in the strain observation data segment.

7. The system for analyzing four-component borehole strain observation data according to claim 6, further comprising:a self-inspection module configured to perform self-inspection on the four-component borehole strain observation data segment, and pass, in response to determining that Ae 1 + Ae,,! = Ae11 + Ae^ is satisfied, the self-inspection, wherein Ae1 , Ae11, Ae111 and Ae^ areborehole strain observation data in four directions at the same time; anda crustal stress change determination module configured to determine that a change of crustal stress is observed by means of the data in the strain observation data segment passing the selfinspection.

8. The system for analyzing four-component borehole strain observation data according to claim 6, wherein the feature analysis module comprises:a time range obtaining unit configured to obtain a time range and a sampling rate of the strain observation data segment, and determine a time range feature representation mi and a sampling rate feature representation nu;a frequency range determination unit configured to determine a frequency range of the strain observation data segment according to the time range of the strain observation data segment, and determine a frequency range feature representation m3;an amplitude range extraction unit configured to convert the strain observation data segment into a crustal stress tensor, extract an amplitude range from the crustal stress tensor, and determine an amplitude range feature representation m2;a phenomenon determination unit configured to compare the strain observation data segment with observation data of a seismic network in the same time range, and first determine an observation result whether a phenomenon ms exists; anda feature index vector construction unit configured to form the time range, the amplitude range, the frequency range, the sampling rate and the phenomena into features of the strain observation data segment, and construct the feature index vector as D=D(mi, m2, m3, nu, ms), wherein D is the feature index vector, and D(mi, m2, m3, nu, ms) is a feature index vector function.

9. The system for analyzing four-component borehole strain observation data according to claim 8, wherein the ratio computation module specifically comprises:an abnormal phenomenon ratio determination unit configured to determine a ratio of the number of the data points corresponding to the abnormal phenomena to a number of data points of the strain observation data segment as the ratio of the abnormal phenomena according to a priority analysis order of the abnormal phenomena higher than the observation phenomena; andan observation phenomenon ratio determination unit configured to compute a ratio of the number of the data points corresponding to the observation phenomena to the number of the data points of the strain observation data segment as the ratio of the observation phenomena.

10. The system for analyzing four-component borehole strain observation data according toclaim 6, wherein the borehole strain observation data evaluation system comprises a control layer and a network layer;the control layer comprises a general target and a plurality of criteria, wherein the general target is the strain observation data, and the plurality of criteria comprise a tectonic stress, a nontectonic stress and a residual stress; andthe network layer comprises a coseismic cluster, a solid tide cluster, a site effect cluster, an instrument parameter cluster and an abnormal tectonic cluster, wherein the coseismic cluster comprises P waves and S waves; the solid tide cluster comprises M2 waves and 01 waves; the site effect cluster comprises terrain, rivers and faults; and the instrument parameter cluster comprises cement creep and coseismic decoupling.

Images