Real-time analysis method and system for deep-sea formation impedance

Abstract

The application discloses a deep-sea stratum impedance real-time analysis method and system, and relates to the technical field of deep-sea exploration. The method forms a position-bound multi-source mechanical data set by fusing cone tip resistance and multi-channel rope tension data; an adaptive filtering strategy driven by a change rate is adopted to enhance the weak mutation characteristics near the interface while suppressing noise; based on dynamic adjustment of the fusion weight according to the penetration velocity, the mechanical response coefficient can adapt to different working conditions; combined with probabilistic identification and spatial compensation correction of probe posture deviation, the depth positioning deviation caused by inclination is overcome, and the automatic real-time analysis of stratum impedance is realized by matching with a standard profile. Through multi-source mechanical information depth fusion, change rate driven adaptive filtering, penetration velocity tracking weight adjustment and posture deviation compensation, the accuracy of stratum interface identification under different working conditions is improved, and the automatic and real-time analysis of deep-sea stratum impedance is realized.

Description

Technical Field

[0001] This invention relates to the field of deep-sea exploration technology, specifically to a method and system for real-time analysis of deep-sea strata impedance. Background Technology

[0002] In the field of deep-sea exploration technology, static cone penetration testing (CPPT) is an important means of obtaining in-situ physical and mechanical properties of seabed sediments. As the depth of deep-sea exploration continues to increase, the types of sediments vary by orders of magnitude in intensity, ranging from shallow fluid clay to deep dense sand, calcareous crusts, and other hard interlayers. This places increasingly higher demands on the accurate identification of stratigraphic interfaces.

[0003] Existing formation impedance analysis methods primarily rely on single-dimensional time-series data of end-point probe cone resistance for formation stratification, which significantly reduces accuracy in alternating or heterogeneous formations. While some methods introduce auxiliary parameters such as sidewall friction, the lack of a deeply integrated fusion analysis mechanism between multi-source mechanical data makes it difficult to form a comprehensive characterization of formation mechanical responses. In signal processing, traditional schemes typically employ fixed-parameter filtering and feature extraction strategies, failing to adaptively adjust based on the time-varying characteristics of mechanical signals during penetration. This results in a difficulty in simultaneously suppressing noise in stable formations and enhancing high-frequency abrupt changes near interfaces.

[0004] In actual operation, deep-sea penetration devices are affected by factors such as ocean current disturbances, bottom contact attitude deviations, and sediment heterogeneity, often resulting in a significant tilting deviation of the probe tip. Most existing impedance analysis methods are based on the assumption of vertical probe penetration, lacking an effective compensation mechanism for penetration attitude deviations. This leads to a systematic deviation between the calculated stratigraphic interface depth and the actual vertical depth, directly interfering with the reliability of geological stratification conclusions. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for real-time analysis of deep-sea strata impedance, solving the problems mentioned in the background section.

[0006] To achieve the above objectives, this invention provides the following technical solution: a real-time deep-sea strata impedance analysis method applied to deep-sea penetration devices. Existing penetration devices often employ segmented transmission, resulting in significant step-like jumps in penetration force during segment switching, especially in the transition zone between segments, where the force-velocity curves exhibit a step-like abrupt change. This leads to probe instability, signal distortion, and severely affects interface identification. The deep-sea penetration device proposed in this invention adopts a stepless transmission architecture. Through continuous proportional adjustment of the topological states of multiple channels in a distributed flexible transmission matrix, the concept of fixed segments is weakened, allowing the transmission ratio to vary steplessly across the entire range. This results in a smooth transition of penetration force and velocity, completely eliminating the mechanical steps caused by segmented switching. The deep-sea penetration device includes an end-point penetration probe and multiple flexible transmission channels arranged circumferentially along the central control axis, comprising: Step 1: Multi-source data alignment and binding: The timing data of the cone tip resistance during the penetration process of the end-penetrating probe and the timing data of the rope tension of multiple flexible transmission channels arranged circumferentially along the central control axis are obtained. The timestamps are aligned and the depth is resampled to obtain a set of multi-source mechanical timing data with position binding. Step 2, Adaptive Filtering and Enhancement: Calculate the tension change rate at each depth node in the multi-source mechanical time series data set; when the tension change rate is lower than the preset smoothing threshold, perform sliding window mean filtering on the rope tension time series data; when it is higher than the preset smoothing threshold, perform high-frequency feature enhancement to obtain the multi-source mechanical feature sequence. Step 3: Dynamic weighted feature fusion: Based on the multi-source mechanical feature sequence, the cone tip resistance gradient and multi-channel tension variance are extracted; according to the penetration depth change rate, the fusion weight of the two is adjusted in real time, and the comprehensive mechanical response coefficient of each depth node is calculated. Step 4, Gradient Extremum Extraction Interface: Based on the comprehensive mechanical response coefficient, a mechanical response gradient curve along the penetration depth direction is constructed, and local extreme points are extracted as a set of candidate formation interface depth nodes. Step 5: Calculate the probability of interface mutation. For the candidate stratigraphic interface depth node set, the impedance change significance index of each node is calculated, and the initial stratigraphic interface change probability distribution is constructed by combining the mechanical response correlation of adjacent depth nodes. Step 6, Attitude Misalignment Compensation Correction: Based on the penetration attitude offset of the end penetration probe, the initial formation interface abrupt change probability distribution is spatially compensated and corrected to obtain the corrected formation interface abrupt change probability distribution. Step 7: Standard Profile Matching and Identification: Based on the modified stratigraphic interface abrupt change probability distribution, depth nodes with probability values ​​exceeding the preset abrupt change threshold are extracted. Spatial topological matching is then performed in conjunction with the preset deep-sea sediment physical and mechanical property standard profile to obtain real-time analysis results of deep-sea stratigraphic impedance within the penetration depth range.

[0007] As a further aspect of the present invention, step 1 is implemented as follows: The original cone tip resistance time series data of the end-penetrating probe tip during the penetration process is obtained, as well as the original rope tension time series data of multiple flexible transmission channels arranged circumferentially along the central control axis. The acquisition timestamp of each data and the real-time depth coordinate of the end-penetrating probe are extracted to obtain multi-source data spatiotemporal reference information. Based on the spatiotemporal reference information of multi-source data, the original cone tip resistance time series data and the original rope tension time series data are interpolated and aligned on the time axis to obtain unified time base cone tip resistance time series data and unified time base rope tension time series data. Based on the time series data of cone tip resistance and the time series data of rope tension under the same time base, data slicing and channel splicing are performed at preset depth intervals to obtain a set of multi-source mechanical time series data with location binding.

[0008] As a further aspect of the present invention, step 2 is implemented as follows: In a multi-source mechanical time-series dataset bound to a calculation location, the ratio of the rope tension difference between adjacent depth nodes to the depth difference is used to obtain the tension change rate of each depth node. When the rate of change of tension is lower than the preset smoothing threshold, the rope tension time series data within the preset penetration depth range is extracted as a sliding window with the current depth node as the center. The arithmetic mean of the data within the sliding window is calculated to replace the original rope tension value of the current depth node, thus completing the sliding window mean filtering. When the rate of change of tension is higher than or equal to the preset smoothing threshold, a discrete wavelet transform is performed on the rope tension time series data within the sliding window to extract high-frequency wavelet coefficients and perform threshold quantization amplification. Then, an inverse wavelet transform is performed to reconstruct the data, thereby enhancing the high-frequency fluctuation characteristics in the rope tension time series data and obtaining a multi-source mechanical feature sequence.

[0009] As a further aspect of the present invention, step 3 is implemented as follows: Calculate the cone tip resistance difference between the current depth node and the previous adjacent depth node to obtain the cone tip resistance gradient; calculate the standard deviation of the rope tension values ​​of multiple flexible transmission channels at the current depth node to obtain the multi-channel tension variance. Obtain the penetration depth change rate of the current penetration cycle. If the penetration depth change rate is greater than the preset velocity threshold, the first fusion weight of the cone tip resistance gradient is reduced by a preset ratio, and the second fusion weight of the multi-channel tension variance is increased by the same preset ratio. If the penetration depth change rate is less than or equal to the preset velocity threshold, the initial calibration values ​​of the first fusion weight and the second fusion weight are maintained. The adjusted first fusion weight is multiplied by the cone tip resistance gradient, and the adjusted second fusion weight is multiplied by the multi-channel tension variance. The products of the two are summed and normalized to obtain the comprehensive mechanical response coefficients of each depth node.

[0010] As a further aspect of the present invention, step 4 is implemented as follows: Perform a first-order forward difference operation on the sequence of comprehensive mechanical response coefficients arranged along the penetration depth direction to obtain the mechanical response gradient curve along the penetration depth direction. Set a preset depth for the extreme value search window, traverse the extreme value search window in the mechanical response gradient curve, and find the sign reversal point where the gradient value changes from positive to negative or from negative to positive. Extract the depth coordinates of the sign reversal points where the absolute gradient value is greater than the preset gradient significance threshold, and package the depth coordinates and their corresponding absolute gradient values ​​to form a set of candidate stratigraphic interface depth nodes.

[0011] As a further aspect of the present invention, step 5 is implemented as follows: For each candidate node in the candidate stratigraphic interface depth node set, the ratio of its corresponding absolute gradient value to the preset background noise gradient mean is calculated to obtain the impedance change significance index. Taking the current candidate node as the center, extract the comprehensive mechanical response coefficient of a preset number of adjacent depth nodes before and after it, and calculate the Pearson correlation coefficient between adjacent nodes as the mechanical response correlation. The correlation between impedance mutation significance index and mechanical response is nonlinearly mapped and calculated using a preset weighted fusion function to obtain the formation interface mutation probability value at the current candidate node. Traverse the set of candidate stratigraphic interface depth nodes, bind the depth coordinates of each candidate node to its corresponding stratigraphic interface mutation probability value, and construct the initial stratigraphic interface mutation probability distribution.

[0012] As a further aspect of the present invention, step 6 is implemented as follows: The real-time three-dimensional attitude angle of the end-penetrating probe is obtained, the angle between the probe axis and the vertical baseline is calculated to obtain the tilt angle, and the product of the sine of the tilt angle and the current penetration depth is used as the horizontal offset distance to form the penetration attitude offset. Based on the horizontal offset distance and the current penetration depth, the actual penetration hypotenuse length of the probe is calculated using the Pythagorean theorem, and the ratio of the current penetration depth to the hypotenuse length is used as the depth compensation coefficient. The depth coordinates of each candidate node in the initial stratigraphic interface abrupt change probability distribution are multiplied by the depth compensation coefficient to obtain the corrected vertical depth coordinates. Based on the degree of deviation between the horizontal offset distance and the preset allowable offset threshold, a confidence attenuation factor is calculated. The confidence attenuation factor is then multiplied by the corresponding formation interface mutation probability value, and combined with the vertical depth coordinates to generate a corrected formation interface mutation probability distribution.

[0013] As a further aspect of the present invention, step 7 is implemented as follows: From the modified formation interface abrupt change probability distribution, candidate nodes whose formation interface abrupt change probability values ​​are greater than or equal to the preset abrupt change threshold are selected as target formation interface depth nodes. Obtain a pre-defined standard profile of the physical and mechanical properties of deep-sea sediments. The standard profile includes the standard impedance range and prior thickness constraints of multiple sediment layers. Using the target stratigraphic interface depth node as the dividing boundary, the penetration depth is divided into multiple stratigraphic sub-intervals, and the mean value of the comprehensive mechanical response coefficient in each stratigraphic sub-interval is extracted as the measured impedance characteristic. The measured impedance characteristics were compared one by one with the standard impedance range of each sediment layer in the standard profile, and dynamic programming optimization was performed in combination with the thickness prior constraint to determine the target sediment sequence corresponding to each stratum sub-range. By integrating and outputting the target stratigraphic interface depth node, target sediment sequence and corresponding measured impedance characteristics, the real-time analysis results of deep-sea stratigraphic impedance corresponding to the target penetration depth are obtained.

[0014] A real-time deep-sea stratigraphic impedance analysis system, which performs real-time deep-sea stratigraphic impedance analysis methods, including: The power input unit is connected to the central control shaft drive and is used to provide reference power to the central control shaft and the distributed flexible drive matrix; the output end of the power input unit is coupled to the central control shaft through a reduction mechanism or a hydraulic cylinder. The central control shaft is arranged along the longitudinal axis of the deep-sea penetration device. One end of the central control shaft is connected to the output end of the power input unit, and the other end establishes a force transmission path with the end penetration probe through a distributed flexible transmission matrix. The distributed flexible transmission matrix includes multiple independent and controllable flexible transmission channels arranged circumferentially along the central control axis; each flexible transmission channel includes a flexible rope, a pulley assembly, a clutch mechanism, a tension detection unit, and a channel displacement detection unit. The end-penetration probe is installed at the front end of the deep-sea penetration device and is used to directly penetrate deep-sea sediments; The controller is communicatively connected to the power input unit, the central control shaft, the distributed flexible transmission matrix, and the end penetration probe.

[0015] As a further aspect of the present invention: the central control shaft is equipped with an angular velocity detection unit and a rotation angle detection unit, which are used to detect the real-time angular velocity and cumulative rotation angle of the central control shaft, respectively. The angular velocity of the central control shaft is converted into the baseline velocity of the distributed flexible transmission matrix, and the cumulative rotation angle of the central control shaft is converted into the corresponding baseline displacement.

[0016] As a further embodiment of the present invention: one end of the flexible rope is coupled to the central control shaft or the end probe via a pulley assembly, and the other end is switched to different transmission paths via a clutch mechanism; Pulley assemblies are used to change the direction of flexible ropes and distribute force flow; The clutch mechanism is used to control the current topological state of the flexible transmission channel. The topological state includes at least one of the following: moving pulley coupling state, fixed pulley constraint state, bypass follow-up state, and tension convergence state. By adjusting the number of flexible transmission channels in different topological states and their combination ratio, the transmission ratio can be continuously and steplessly adjusted, so that the deep-sea penetration device does not have a step-like jump between penetration force and penetration speed in the full penetration depth range, forming a smooth and continuous stepless penetration characteristic.

[0017] The tension detection unit is used to detect the rope tension borne by the flexible rope in the flexible transmission channel in real time and feed the rope tension data back to the controller; The channel displacement detection unit is used to detect the displacement of the flexible rope end or pulley assembly in the flexible transmission channel in real time and feed the channel displacement data back to the controller.

[0018] As a further aspect of the present invention: a cone tip is provided at the front end of the end-penetrating probe, and a cone tip force sensor is installed at the cone tip to collect the cone tip resistance time-series data in real time during the penetration process; a force sensor is installed on the rod body of the end-penetrating probe or at the penetration actuator to collect the penetration resistance signal; the end-penetrating probe is also equipped with a displacement sensor and an attitude sensor, the displacement sensor is used to detect the penetration displacement of the end-penetrating probe and convert it into real-time depth coordinates, and the attitude sensor is used to detect the three-dimensional attitude angle of the end-penetrating probe to calculate the penetration attitude offset.

[0019] As a further aspect of the present invention: the controller is used to receive the reference power parameters of the power input unit, the penetration state parameters of the end penetration probe, and the matrix state parameters of each flexible transmission channel, and execute steps 1 to 7 of the aforementioned real-time deep-sea stratum impedance analysis method based on the received data, and output the real-time deep-sea stratum impedance analysis results corresponding to the target penetration depth position.

[0020] This invention provides a method and system for real-time analysis of deep-sea formation impedance. Compared with existing technologies, it has the following advantages: This invention acquires time-series data of cone tip resistance and multi-channel rope tension simultaneously, and forms a multi-source mechanical data set with location binding through timestamp alignment and depth resampling, effectively integrating multi-dimensional information such as cone tip resistance and distributed channel mechanical response.

[0021] This invention achieves stepless transmission of penetration force and speed by continuously adjusting the topological state combination ratio of each flexible transmission channel, thus overcoming the mechanical step problem of traditional step-by-step switching.

[0022] This invention utilizes the adaptive selection of sliding window mean filtering or wavelet high-frequency enhancement based on the tension change rate to effectively suppress measurement noise in stable formations and enhance weak abrupt changes near the interface, thus avoiding the feature smoothing or spurious peak amplification problems caused by fixed filtering strategies.

[0023] Extract the cone tip resistance gradient and multi-channel tension variance, and dynamically adjust the fusion weight according to the penetration depth change rate. When penetrating rapidly, increase the tension variance weight to compensate for the lag in the cone tip resistance response, and maintain the standard weight when penetrating slowly, so that the comprehensive mechanical response coefficient can adapt to different penetration conditions.

[0024] By performing gradient calculations and local extremum searches on the comprehensive mechanical response coefficient sequence, the set of candidate stratigraphic interface depth nodes at the location of abrupt response changes is automatically extracted. This eliminates the need for manually presetting the number of layers or empirical threshold curves, thus achieving automated and objective positioning of the initial location of the stratigraphic interface.

[0025] For candidate interface nodes, the correlation between impedance mutation significance index and mechanical response at adjacent depths is integrated. An initial formation interface mutation probability distribution is constructed through nonlinear mapping to quantify the credibility of the interface in probabilistic form, thus avoiding the defect of hard threshold discrimination being easily disturbed by local anomalies.

[0026] Based on the probe attitude offset, the depth compensation coefficient and confidence attenuation factor are calculated to spatially correct the interface depth and probability value, effectively suppressing the depth positioning error and interface misjudgment caused by probe tilt, and ensuring that the final obtained formation interface depth is close to the true vertical depth.

[0027] By dynamically programming and matching the corrected mutation probability distribution with the standard deep-sea sediment profile, and optimizing the sequence division using thickness prior constraints, the measured impedance characteristics and sequence assignments of each sediment layer are output, thus realizing an automated closed loop from mechanical data to geological stratification interpretation. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the real-time impedance analysis method for deep-sea formations according to the present invention.

[0029] Figure 2 This is a system block diagram of the real-time deep-sea formation impedance analysis system of the present invention.

[0030] Figure 3 This is a flowchart of the calculation of the comprehensive mechanical response coefficient to the formation matching in this invention.

[0031] Figure 4 This is a schematic diagram of the flexible transmission channel in the deep-sea penetration device of the present invention.

[0032] Figure 5 This is a schematic diagram of the end penetration probe in the deep-sea penetration device of the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Please see Figures 1 to 5 As shown, the embodiments of the present invention provide the following technical solutions: As an embodiment of the present invention: This invention relates to a real-time method for analyzing the impedance of deep-sea formations. This method is applied to deep-sea penetration devices, which employ a continuously variable transmission architecture. Instead of segmented switching, a continuously variable transmission ratio is used to achieve smooth, stepless adjustment of penetration force and speed. The deep-sea penetration device includes an end-point penetration probe and multiple flexible transmission channels arranged circumferentially along the central control axis. The method includes the following steps: Step 1: Obtain the timing data of the cone tip resistance during the penetration process of the end-penetrating probe, as well as the timing data of the rope tension of multiple flexible transmission channels arranged circumferentially along the central control axis. Perform timestamp alignment and depth resampling to obtain a set of multi-source mechanical timing data with position binding.

[0035] In a specific implementation, this step further includes the following sub-steps: Acquire the original cone tip resistance time series data during the penetration process of the end-penetrating probe tip. And the original rope tension timing data of multiple flexible transmission channels arranged circumferentially along the central control axis. Extract the acquisition timestamps of the above data and the real-time depth coordinates of the end-penetrating probe. This allows us to obtain spatiotemporal reference information from multi-source data. Where j=1,2,...,M, M is the total number of flexible transmission channels; Original cone tip resistance time series data Data is collected by a cone-tip force sensor installed at the front end of the end-penetrating probe. Raw rope tension time series data The data is collected by a tension detection unit installed on each flexible transmission channel; Real-time depth coordinates It is calculated by the penetration displacement sensor or the rope release length detection unit.

[0036] Based on the spatiotemporal reference information of multi-source data, the original cone tip resistance time series data... Compared with the original rope tension time series data Time axis interpolation and alignment were performed to obtain time series data of cone tip resistance with a unified time base. With unified time-based rope tension time series data ; In this embodiment, when the sampling frequencies of the various sensors are inconsistent, the two types of data are mapped to the same time base through linear interpolation or nearest neighbor interpolation, ensuring that the cone tip resistance and rope tension correspond one-to-one in time at each timestamp.

[0037] Then, based on the unified time-base cone tip resistance time series data With unified time-based rope tension time series data Data slices and channels are stitched together at preset depth intervals to obtain a location-bound multi-source mechanical time-series data set. ; in, Let be the depth coordinates at the i-th depth node. This represents the cone tip resistance value at that depth. Let be the rope tension value of the j-th flexible transmission channel at that depth; The target penetration depth is determined by the real-time depth coordinates of the end-penetrating probe; data slicing refers to extracting a set of data at a preset depth interval Δz; channel splicing refers to combining the rope tension data of multiple flexible transmission channels at the same depth as different dimensional features.

[0038] Step 2: Calculate the multi-source mechanical time series data set The rate of change of tension at each depth node; when the rate of change of tension is lower than the preset smoothing threshold γ th At that time, a sliding window mean filter is applied to the rope tension time series data; When the rate of change of tension is higher than or equal to the preset smoothing threshold γ th At that time, high-frequency feature enhancement is performed to obtain a multi-source mechanical feature sequence.

[0039] In a specific implementation, this step further includes: In a multi-source mechanical time-series dataset bound to a computational location, the ratio of the rope tension difference to the depth difference between adjacent depth nodes is used to obtain the values ​​for each depth node. rate of change of tension The formula is Where M is the total number of flexible transmission channels. and These are the rope tension values ​​of the j-th flexible transmission channel at the i-th and (i-1)-th depth nodes, respectively. and These are the corresponding depth coordinates.

[0040] When the rate of change of tension Below the preset smoothing threshold γ th At that time, with the current depth node Centered on the target area, extract the preset penetration depth range. The rope tension time series data within the window is used as a sliding window, where, For the preset half-width of the penetration depth interval, calculate the arithmetic mean of the data within the sliding window to replace the original rope tension value of the current depth node, thus completing the sliding window mean filtering; where i is the depth node index, i=1,2,...,N, and N is the total number of depth nodes.

[0041] For each flexible transmission channel j, the filtered tension value is ; in, This represents the number of depth nodes contained within the sliding window. is the original rope tension value of each depth node within the window; k is the half width of the sliding window, and the total length of the window is 2k+1 depth nodes; When the rate of change of tension Higher than or equal to the preset smoothing threshold γ th At that time, a discrete wavelet transform is performed on the rope tension time series data within the sliding window to extract high-frequency wavelet coefficients and then perform threshold quantization amplification. ; In the formula, A is the preset amplification bias, and λ is the preset wavelet coefficient threshold. Subsequently, inverse wavelet transform reconstruction was performed to obtain the enhanced rope tension value. ; The cone tip resistance data after the above processing Compared with enhanced rope tension data Together they constitute a multi-source mechanical characteristic sequence .

[0042] Step 3: Based on the multi-source mechanical feature sequence, extract the cone tip resistance gradient and multi-channel tension variance; according to the penetration depth change rate, adjust the fusion weight of the two in real time, and calculate the comprehensive mechanical response coefficient of each depth node.

[0043] In a specific implementation, this step further includes: Calculate the current depth node adjacent depth node The difference in cone tip resistance is used to obtain the cone tip resistance gradient. The formula is ; Calculate multiple flexible transmission channels at the current depth node The standard deviation of the rope tension value is used to obtain the multi-channel tension variance, as shown in the formula: ,in, , This represents the average tension of the ropes in each channel at that depth node; Obtain the rate of change of penetration depth during the current penetration cycle. ,in, The time interval between adjacent depth nodes; If the rate of change of penetration depth Greater than the preset speed threshold v th Then the cone tip drag gradient G i The first fusion weight w1 is reduced by the preset ratio β, and the multi-channel tension variance V is also reduced. i The second fusion weight w2 is increased by the same preset ratio β; where w1 + w2 = 1, w1 is the initial fusion weight of the cone tip resistance gradient; w2 is the initial fusion weight of the multi-channel tension variance, and w1 and w 2均为 The preset value, β, refers to the preset weight adjustment ratio. When the penetration speed exceeds the threshold, it is subtracted from the weight corresponding to the cone tip resistance gradient and added to the weight corresponding to the tension variance.

[0044] If the rate of change of penetration depth Less than or equal to the preset speed threshold v th Then, the initial calibration values ​​w1 and w2 of the first fusion weight and the second fusion weight are maintained.

[0045] The adjusted first fusion weight is compared with the cone tip resistance gradient. Multiply, and then multiply the adjusted second fusion weights with the multi-channel tension variance. Multiply the two, sum the products and normalize them to obtain the comprehensive mechanical response coefficients of each depth node; Normalization is performed using either maximum-minimum normalization or Z-score normalization.

[0046] Step 4: Based on the comprehensive mechanical response coefficient, construct the mechanical response gradient curve along the penetration depth direction, and extract local extreme points as a set of candidate formation interface depth nodes.

[0047] In a specific implementation, this step further includes: The sequence of comprehensive mechanical response coefficients arranged along the penetration depth direction Performing a first-order forward difference operation yields the mechanical response gradient curve along the penetration depth direction. ; For depth node i, its gradient value ,in, This is the comprehensive mechanical response coefficient of the node.

[0048] Set a preset depth for the extreme value search window, traverse the extreme value search window in the mechanical response gradient curve, and find the sign reversal point where the gradient value changes from positive to negative or from negative to positive. If there is a condition in the window that satisfies For depths less than 0, record the depth coordinates as z. cand .

[0049] Extract the depth coordinates of points where the absolute gradient value is greater than a preset gradient significance threshold from the sign reversal points. Package the depth coordinates and their corresponding absolute gradient values ​​to form a set of candidate stratigraphic interface depth nodes. Where p = 1, 2, ... , The total number of candidate nodes. Let be the depth coordinates corresponding to candidate node p. The absolute value of the mechanical response gradient at candidate node p.

[0050] Step 5: For the candidate stratigraphic interface depth node set, calculate the impedance change significance index of each node, and construct the initial stratigraphic interface change probability distribution by combining the mechanical response correlation of adjacent depth nodes.

[0051] In a specific implementation, this step further includes: For the candidate stratigraphic interface depth node set For each candidate node p, calculate the ratio of its corresponding absolute gradient value to the preset mean gradient of background noise to obtain the impedance change significance index. The preset background noise gradient mean is obtained by statistical analysis of the mechanical response gradient of the stable sediment section in the initial stage of penetration. The preset background noise gradient mean can be obtained by selecting a section with stable fluctuation of comprehensive mechanical response coefficient in the initial stage of penetration (such as the first 0.5 meters of depth) and calculating the arithmetic mean or median of the absolute value of the mechanical response gradient in that section as the value of this parameter.

[0052] With the current candidate node Centered on the target node, extract the comprehensive mechanical response coefficients of a predetermined number of adjacent depth nodes. Calculate the Pearson correlation coefficient between adjacent nodes as the mechanical response correlation; The formula for the Pearson correlation coefficient is: ,in, This represents the average value of the comprehensive mechanical response coefficients within the window.

[0053] The correlation between impedance mutation significance index and mechanical response is nonlinearly mapped using a preset weighted fusion function to obtain the probability value of formation interface mutation at the current candidate node. ; The calculation formula is Where a, b, and c are the calibration coefficients of the nonlinear mapping. It reflects the decorrelation of the mechanical response; the lower the correlation, the higher the probability of abrupt change.

[0054] Traverse the set of candidate stratigraphic interfaces depth nodes The depth coordinates of each candidate node Its corresponding probability value of abrupt change at the stratigraphic interface Binding was performed to construct the initial stratigraphic interface abrupt change probability distribution. .

[0055] Step 6: Based on the penetration attitude offset of the end-penetrating probe, perform spatial compensation correction on the initial formation interface abrupt change probability distribution to obtain the corrected formation interface abrupt change probability distribution.

[0056] In a specific implementation, this step further includes: The real-time three-dimensional attitude angle of the end-penetrating probe is obtained, and the tilt angle is calculated by measuring the angle between the probe axis and the vertical baseline. The penetration length measured along its own axis is taken as... The sine of the tilt angle is compared with the current penetration depth. The product of these is used as the horizontal offset distance. This constitutes the penetration attitude offset, and the formula is: ; The real-time three-dimensional attitude angle is obtained by an attitude sensor installed inside the probe.

[0057] Based on the horizontal offset distance and the current penetration depth, the actual penetration hypotenuse length of the probe is calculated using the Pythagorean theorem. The formula is as follows: ; Current penetration depth With the length of the hypotenuse The ratio of the two values ​​is used as the depth compensation coefficient. The formula is: .

[0058] The corrected vertical depth coordinates are obtained by multiplying the depth coordinates of each candidate node in the initial formation interface abrupt change probability distribution by a depth compensation coefficient. .

[0059] Based on horizontal offset distance With the preset allowed offset threshold The degree of deviation between them is used to calculate the confidence decay factor. ;in, ; Multiply the confidence decay factor by the corresponding formation interface abrupt change probability value, and combine this with the vertical depth coordinates to generate the corrected formation interface abrupt change probability distribution. .

[0060] Step 7: Based on the modified stratigraphic interface abrupt change probability distribution, extract depth nodes with probability values ​​exceeding the preset abrupt change threshold, and perform spatial topological matching in conjunction with the preset deep-sea sediment physical and mechanical property standard profile to obtain real-time analysis results of deep-sea stratigraphic impedance within the penetration depth range.

[0061] In a specific implementation, this step further includes: From the modified formation interface mutation probability distribution, candidate nodes with formation interface mutation probability values ​​greater than or equal to the preset mutation threshold are selected as target formation interface depth nodes.

[0062] Obtain a pre-defined standard profile of the physical and mechanical properties of deep-sea sediments. The standard profile includes the standard impedance range and thickness prior constraints of multiple sediment layers.

[0063] The standard profile is established through regional geological surveys, borehole data, or prior in-situ test data. In one embodiment, the standard profile is obtained by: establishing a correspondence table between different sediment types (such as deep-sea clay, siliceous mud, calcareous mud, turbidite sand layers, etc.) and standard impedance ranges and typical thicknesses based on laboratory test data of core samples from previous geological boreholes in the local sea area, or publicly available geological survey data from adjacent areas. This correspondence table is pre-installed in the controller in the form of a database as the standard profile. Using the target stratigraphic interface depth node as the dividing boundary, the penetration depth is divided into multiple stratigraphic sub-intervals. The average comprehensive mechanical response coefficient within each stratigraphic sub-interval is extracted as the measured impedance characteristic.

[0064] The measured impedance characteristics were compared one by one with the standard impedance intervals of each sediment layer in the standard profile, and dynamic programming optimization was performed in conjunction with the prior thickness constraint to determine the target sedimentary sequence corresponding to each stratigraphic sub-interval. The dynamic programming optimization aimed to minimize the deviation between the measured impedance characteristics and the center value of the standard impedance interval, while satisfying the thickness constraint, in order to find the optimal stratigraphic division scheme.

[0065] By integrating and outputting the target stratigraphic interface depth node, target sediment sequence and corresponding measured impedance characteristics, real-time analysis results of deep-sea stratigraphic impedance within the penetration depth range are obtained.

[0066] As a second embodiment of the present invention: In specific implementation, compared with Embodiment 1, the only difference between the technical solution of this embodiment and Embodiment 1 is that: this embodiment provides a real-time deep-sea stratum impedance analysis system, which is used to execute the aforementioned real-time deep-sea stratum impedance analysis method. The system includes a power input unit 100, a central control shaft 200, a distributed flexible transmission matrix 300, an end penetration probe 400, and a controller 500. The distributed flexible drive matrix 300 includes multiple independently controllable flexible drive channels arranged circumferentially along the central control axis 200.

[0067] The power input unit 100 is connected to the central control shaft 200 and is used to provide reference power to the central control shaft 200 and the distributed flexible transmission matrix 300. The power input unit 100 adopts either a pressure-resistant electro-hydraulic servo system or a motor-driven winch, and its output end is coupled to the central control shaft 200 through a reduction mechanism or a hydraulic cylinder. The power input unit 100 is equipped with at least one of the following sensing devices: a motor encoder, a hydraulic cylinder displacement sensor, a winch speed sensor, a hydraulic pressure sensor, a torque detection unit, and a main rope tension detection unit, for collecting or converting reference power parameters such as reference speed, reference stroke, and reference thrust or reference tension. The central control shaft 200 is arranged along the longitudinal axis of the deep-sea penetration device; One end of the central control shaft 200 is connected to the output end of the power input unit 100, and the other end establishes a force transmission path with the end penetration probe 400 through the distributed flexible transmission matrix 300. The central control axis 200 is equipped with an angular velocity detection unit and a rotation angle detection unit, which are used to detect the real-time angular velocity and cumulative rotation angle of the central control axis 200, respectively. The angular velocity of the central control axis 200 is converted into the baseline linear velocity of the distributed flexible transmission matrix 300, and the cumulative rotation angle of the central control axis 200 is converted into the corresponding baseline displacement.

[0068] The distributed flexible transmission matrix 300 includes multiple independently controllable flexible transmission channels. Each flexible transmission channel includes a flexible rope, a pulley assembly, a clutch mechanism, a tension detection unit, and a channel displacement detection unit.

[0069] One end of the flexible rope is coupled to the central control shaft 200 or the end probe 400 via a pulley assembly, and the other end can be switched to different transmission paths via a clutch mechanism; Pulley assemblies are used to change the direction of flexible ropes and distribute force flow; The clutch mechanism is used to control the current topological state of the flexible transmission channel, which includes at least one of the following: moving pulley coupling state, fixed pulley constraint state, bypass follow-up state, and tension convergence state. The tension detection unit is used to detect the rope tension borne by the flexible rope in the flexible transmission channel in real time and feed the rope tension data back to the controller 500. The channel displacement detection unit is used to detect the displacement of the flexible rope end or pulley assembly in the flexible transmission channel in real time, and feeds back the channel displacement data to the controller 500.

[0070] The terminal penetration probe 400 is installed at the front end of the deep-sea penetration device for direct penetration into deep-sea sediments; The front end of the end penetration probe 400 is provided with a cone tip, and a cone tip force sensor is installed at the cone tip to collect the cone tip resistance time series data in real time during the penetration process. A force sensor is installed on the rod of the end-penetrating probe 400 or at the penetration actuator to collect penetration resistance signals; The end-penetrating probe 400 is also equipped with a displacement sensor and an attitude sensor. The displacement sensor is used to detect the penetration displacement of the end-penetrating probe 400 and convert it into real-time depth coordinates. The attitude sensor is used to detect the three-dimensional attitude angle of the end-penetrating probe 400 in order to calculate the penetration attitude offset. The controller 500 is communicatively connected to the power input unit 100, the central control shaft 200, the distributed flexible transmission matrix 300, and the end penetration probe 400. The controller 500 is used to receive the reference power parameters of the power input unit 100, the penetration state parameters of the end penetration probe 400 and the matrix state parameters of each flexible transmission channel, and execute steps 1 to 7 of the aforementioned real-time deep-sea stratum impedance analysis method based on the received data, and finally output the real-time deep-sea stratum impedance analysis results corresponding to the target penetration depth position. The controller 500 is also used to send control commands to the clutch mechanism of each flexible transmission channel to adjust the number of flexible transmission channels in the moving pulley coupling state or the tension aggregation state, so as to realize the stepless smooth switching of the distributed flexible transmission matrix 300 between the coaxial motion coupling extension mode and the tension torque reverse aggregation force amplification mode.

[0071] During the penetration operation, the power input unit 100 outputs a reference power, which is transmitted to the distributed flexible transmission matrix 300 via the central control shaft 200. Based on the topology state commands issued by the controller 500, the distributed flexible transmission matrix 300 distributes the input power to the flexible transmission channels in different topology states through the state switching of the clutch mechanisms in each flexible transmission channel.

[0072] When multiple flexible transmission channels are in a state of dynamic pulley coupling, the distributed flexible transmission matrix 300 forms a transmission topology with amplified speed and stroke, enabling the end-penetrating probe 400 to obtain a higher penetration speed and a larger single penetration stroke.

[0073] When multiple flexible transmission channels are in a state of tension aggregation, the rope tension of each channel is reversely aggregated through the pulley assembly and the central control shaft 200 to form a combined thrust acting on the end penetration probe 400, so that the end penetration probe 400 obtains a large instantaneous penetration force to penetrate hard interlayers or dense formations.

[0074] When transitions between different penetration stages are required, the controller 500 gradually changes the ratio of channels in the dynamic pulley coupling state to the tension convergence state by progressively switching the topology of individual channels, thus achieving continuous stepless adjustment of the transmission ratio. During this process, the penetration force and penetration speed of the end-penetration probe 400 change smoothly and continuously, ensuring the mechanical continuity of the entire penetration process.

[0075] Throughout the entire penetration process, the controller 500 uses multi-source data such as the cone tip resistance of the end penetration probe 400, the rope tension of each flexible transmission channel, and the rate of change of penetration depth to execute the aforementioned real-time formation impedance analysis method in real time, dynamically adjust the topology of the distributed flexible transmission matrix 300, and output the real-time formation impedance analysis results.

[0076] It should be noted that the preset thresholds (such as preset smoothing threshold, preset velocity threshold, preset gradient significance threshold, preset mutation threshold, etc.) and preset ratios and quantities involved in the above steps can all be predetermined through calibration tests or statistical analysis based on factors such as the actual operating conditions of the deep-sea penetration device, sensor accuracy, and sediment characteristics of the target sea area, and do not refer to a specific fixed value.

[0077] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0078] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0079] All the above formulas use dimensionless numerical calculations, that is, by introducing a reference benchmark, the original physical quantities are transformed into dimensionless relative values, eliminating the influence of units and retaining the relative size relationship of physical quantities; the formula is the closest to the real situation obtained by software simulation of a large amount of data, and the preset parameters and threshold selection in the formula are set by those skilled in the art according to the actual situation.

[0080] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0081] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for real-time analysis of deep-sea formation impedance, characterized in that, This method, applied to deep-sea penetration devices, which include an end-point penetration probe and multiple flexible transmission channels arranged circumferentially along a central control axis, comprises the following steps: Step 1: Obtain the timing data of the cone tip resistance of the end-penetrating probe tip during the penetration process, as well as the timing data of the rope tension of multiple flexible transmission channels arranged circumferentially along the central control axis. Perform timestamp alignment and depth resampling to obtain a set of multi-source mechanical timing data with position binding. Step 2: Calculate the tension change rate at each depth node in the multi-source mechanical time series data set; when the tension change rate is lower than the preset smoothing threshold, perform sliding window mean filtering on the rope tension time series data; when it is higher than the preset smoothing threshold, perform high-frequency feature enhancement to obtain the multi-source mechanical feature sequence. Step 3: Based on the multi-source mechanical feature sequence, extract the cone tip resistance gradient and multi-channel tension variance; according to the penetration depth change rate, adjust the fusion weight of the two in real time, and calculate the comprehensive mechanical response coefficient of each depth node. Step 4: Based on the comprehensive mechanical response coefficient, construct the mechanical response gradient curve along the penetration depth direction, and extract local extreme points as a set of candidate formation interface depth nodes. Step 5: For the candidate stratigraphic interface depth node set, calculate the impedance change significance index of each node, and construct the initial stratigraphic interface change probability distribution by combining the mechanical response correlation of adjacent depth nodes. Step 6: Based on the penetration attitude offset of the end penetration probe, perform spatial compensation correction on the initial formation interface abrupt change probability distribution to obtain the corrected formation interface abrupt change probability distribution. Step 7: Based on the modified stratigraphic interface abrupt change probability distribution, extract depth nodes with probability values ​​exceeding the preset abrupt change threshold, and perform spatial topological matching in conjunction with the preset deep-sea sediment physical and mechanical property standard profile to obtain real-time analysis results of deep-sea stratigraphic impedance within the penetration depth range.

2. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 1 are as follows: The original cone tip resistance time series data of the end-penetrating probe tip during the penetration process is obtained, as well as the original rope tension time series data of multiple flexible transmission channels arranged circumferentially along the central control axis. The acquisition timestamp of each data and the real-time depth coordinate of the end-penetrating probe are extracted to obtain multi-source data spatiotemporal reference information. Based on the spatiotemporal reference information of multi-source data, the original cone tip resistance time series data and the original rope tension time series data are interpolated and aligned on the time axis to obtain unified time base cone tip resistance time series data and unified time base rope tension time series data. Based on the unified time-base cone tip resistance time-series data and the unified time-base rope tension time-series data, data slicing and channel splicing are performed at preset depth intervals to obtain a set of location-bound multi-source mechanical time-series data.

3. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 2 are as follows: In a multi-source mechanical time-series dataset bound to a calculation location, the ratio of the rope tension difference between adjacent depth nodes to the depth difference is used to obtain the tension change rate of each depth node. When the rate of change of tension is lower than the preset smoothing threshold, the rope tension time series data within the preset penetration depth range is extracted as a sliding window with the current depth node as the center. The arithmetic mean of the data within the sliding window is calculated to replace the original rope tension value of the current depth node, thus completing the sliding window mean filtering. When the rate of change of tension is higher than or equal to the preset smoothing threshold, a discrete wavelet transform is performed on the rope tension time series data within the sliding window to extract high-frequency wavelet coefficients and perform threshold quantization amplification. Then, an inverse wavelet transform is performed to reconstruct the data, thereby enhancing the high-frequency fluctuation characteristics in the rope tension time series data and obtaining a multi-source mechanical feature sequence.

4. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 3 are as follows: Calculate the cone tip resistance difference between the current depth node and the previous adjacent depth node to obtain the cone tip resistance gradient; calculate the standard deviation of the rope tension values ​​of multiple flexible transmission channels at the current depth node to obtain the multi-channel tension variance. Obtain the penetration depth change rate of the current penetration cycle. If the penetration depth change rate is greater than the preset velocity threshold, the first fusion weight of the cone tip resistance gradient is reduced by a preset ratio, and the second fusion weight of the multi-channel tension variance is increased by the same preset ratio. If the penetration depth change rate is less than or equal to the preset velocity threshold, the initial calibration values ​​of the first fusion weight and the second fusion weight are maintained. The adjusted first fusion weight is multiplied by the cone tip resistance gradient, and the adjusted second fusion weight is multiplied by the multi-channel tension variance. The products of the two are summed and normalized to obtain the comprehensive mechanical response coefficients of each depth node.

5. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 4 are as follows: Perform a first-order forward difference operation on the sequence of comprehensive mechanical response coefficients arranged along the penetration depth direction to obtain the mechanical response gradient curve along the penetration depth direction. Set a preset depth for the extreme value search window, traverse the extreme value search window in the mechanical response gradient curve, and find the sign reversal point where the gradient value changes from positive to negative or from negative to positive. Extract the depth coordinates of the sign reversal points where the absolute gradient value is greater than the preset gradient significance threshold, and package the depth coordinates and their corresponding absolute gradient values ​​to form a set of candidate stratigraphic interface depth nodes.

6. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 5 are as follows: For each candidate node in the candidate stratigraphic interface depth node set, the ratio of its corresponding absolute gradient value to the preset background noise gradient mean is calculated to obtain the impedance change significance index. Taking the current candidate node as the center, extract the comprehensive mechanical response coefficient of a preset number of adjacent depth nodes before and after it, and calculate the Pearson correlation coefficient between adjacent nodes as the mechanical response correlation. The correlation between impedance mutation significance index and mechanical response is nonlinearly mapped and calculated using a preset weighted fusion function to obtain the formation interface mutation probability value at the current candidate node. Traverse the set of candidate stratigraphic interface depth nodes, bind the depth coordinates of each candidate node to its corresponding stratigraphic interface mutation probability value, and construct the initial stratigraphic interface mutation probability distribution.

7. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 6 are as follows: The real-time three-dimensional attitude angle of the end-penetrating probe is obtained, the angle between the probe axis and the vertical baseline is calculated to obtain the tilt angle, and the product of the sine of the tilt angle and the current penetration depth is used as the horizontal offset distance to form the penetration attitude offset. Based on the horizontal offset distance and the current penetration depth, the actual penetration hypotenuse length of the probe is calculated using the Pythagorean theorem, and the ratio of the current penetration depth to the hypotenuse length is used as the depth compensation coefficient. The depth coordinates of each candidate node in the initial stratigraphic interface abrupt change probability distribution are multiplied by the depth compensation coefficient to obtain the corrected vertical depth coordinates. Based on the degree of deviation between the horizontal offset distance and the preset allowable offset threshold, a confidence attenuation factor is calculated. The confidence attenuation factor is then multiplied by the corresponding formation interface mutation probability value, and combined with the vertical depth coordinates to generate a corrected formation interface mutation probability distribution.

8. The method for real-time analysis of deep-sea formation impedance according to claim 1, characterized in that: The specific steps for step 7 are as follows: From the modified formation interface abrupt change probability distribution, candidate nodes whose formation interface abrupt change probability values ​​are greater than or equal to the preset abrupt change threshold are selected as target formation interface depth nodes. Obtain a pre-defined standard profile of the physical and mechanical properties of deep-sea sediments. The standard profile includes the standard impedance range and prior thickness constraints of multiple sediment layers. Using the target stratigraphic interface depth node as the dividing boundary, the penetration depth is divided into multiple stratigraphic sub-intervals, and the mean value of the comprehensive mechanical response coefficient in each stratigraphic sub-interval is extracted as the measured impedance characteristic. The measured impedance characteristics were compared one by one with the standard impedance range of each sediment layer in the standard profile, and dynamic programming optimization was performed in combination with the thickness prior constraint to determine the target sediment sequence corresponding to each stratum sub-range. By integrating and outputting the target stratigraphic interface depth node, target sediment sequence and corresponding measured impedance characteristics, the real-time analysis results of deep-sea stratigraphic impedance corresponding to the target penetration depth are obtained.

9. A real-time deep-sea stratigraphic impedance analysis system, the system being used to execute the real-time deep-sea stratigraphic impedance analysis method according to any one of claims 1-8, characterized in that, The system includes: The power input unit (100) is connected to the central control shaft (200) and is used to provide reference power to the central control shaft (200) and the distributed flexible drive matrix (300); The central control shaft (200) is arranged along the longitudinal axis of the deep-sea penetration device. One end of the central control shaft (200) is connected to the output end of the power input unit (100), and the other end establishes a force transmission path with the end penetration probe (400) through the distributed flexible transmission matrix (300). The distributed flexible drive matrix (300) includes multiple independently controllable flexible drive channels arranged circumferentially along the central control axis (200); The end-penetration probe (400) is installed at the front end of the deep-sea penetration device and is used to directly penetrate deep-sea sediments.

10. The real-time deep-sea formation impedance analysis system according to claim 9, characterized in that: The system also includes a controller (500), which is communicatively connected to the power input unit (100), the central control shaft (200), the distributed flexible drive matrix (300), and the end penetration probe (400); The controller (500) is used to receive the reference power parameters of the power input unit (100), the penetration state parameters of the end penetration probe (400) and the matrix state parameters of each flexible transmission channel, and execute steps 1 to 7 of the aforementioned real-time deep-sea stratum impedance analysis method based on the received data, and output the real-time deep-sea stratum impedance analysis results corresponding to the target penetration depth position.

Images