Open caisson sinking speed self-adaptive regulation and control system based on geological information model

Abstract

The present application belongs to the technical field of sinking well construction, and provides a sinking well sinking speed self-adaptive regulation and control system based on a geological information model, which comprises: a three-dimensional geological information model based on geological exploration stratification data, which is used to calculate a sinking resistance probability distribution interval and output a feedforward resistance prediction atlas; a resistance change trend is analyzed according to the atlas, three stages of initial sinking, normal sinking and final sinking are automatically divided, the upper limit of the pot bottom depth and the target speed threshold value of each stage are calculated, and a segmented self-adaptive strategy set is output; four-corner height difference and blade foot elevation data are collected in real time according to the monitoring frequency set by the strategy set, the speed deviation and the inclination are calculated, and a deviation state vector is output; the priority is judged by analyzing the deviation state vector, a regulation and control instruction is generated and fed back to the real-time monitoring deviation analysis module, and a closed loop is formed until the sinking well is in place. The present application realizes precise self-adaptive control of the sinking well sinking speed.

Description

Technical Field

[0001] This invention belongs to the field of caisson construction technology, specifically an adaptive control system for caisson sinking speed based on a geological information model. Background Technology

[0002] Traditional caisson construction faces numerous technical bottlenecks in projects with complex geological conditions. The most critical issue lies in the contradiction between the discretization of geological information and the open-loop mode of sinking speed control.

[0003] Firstly, current construction primarily relies on the weighted average skin friction values ​​of each soil layer provided in the geological survey report for settlement verification and velocity prediction. However, in highly heterogeneous strata such as mixed fill, silty soil, and sandstone interbedded layers, the lateral thickness of the soil layers varies dramatically, and local phase transitions are frequent. A single average parameter completely masks the spatial distribution characteristics of abrupt changes in resistance. This makes it impossible to accurately predict the resistance change points of the cutting edge when traversing different soil layers before construction. The calculated settlement coefficient deviates significantly from the actual results, and the feedforward basis for velocity control is severely inaccurate. Often, sudden settlement occurs when entering weak interlayers or the soil becomes stagnant and difficult to settle in hard soil layers.

[0004] Secondly, the construction process relies on manual measurement using equipment such as levels and total stations, with observations every 10 minutes during the initial sinking stage and every 30 minutes during the normal sinking stage. This discrete sampling method has significant data blind spots. Furthermore, the entire process of manual recording, analysis, command issuance, and execution results in a severe lag in control response compared to the real-time changes in the caisson's attitude. In actual projects, correction is often only passively initiated when the cumulative tilt reaches the warning threshold of 30cm. By this time, the caisson's center of gravity has already shifted significantly, greatly increasing the difficulty and risk of correction.

[0005] Furthermore, there are inherent conflicts among the current control rules. Correcting deviation requires digging more soil on the higher side of the cutting edge. While this corrects the tilt, it also locally deepens the bottom of the pot, accelerating the sinking speed on that side, creating a coupling contradiction between speed control and attitude control. Operators rely on experience to weigh the pros and cons, lacking quantitative decoupling criteria.

[0006] To address this, the present invention provides an adaptive control system for caisson sinking speed based on a geological information model. Summary of the Invention

[0007] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0008] The technical solution adopted by this invention to solve its technical problem is: an adaptive control system for caisson sinking speed based on a geological information model, comprising the following modules: The geological modeling resistance prediction module, based on geological exploration layered data, quantifies the spatial distribution and friction coefficient of interlayered miscellaneous fill, silty soil, and sandstone, establishes a three-dimensional geological information model, analyzes and calculates the probability distribution range of subsidence resistance, and outputs a feedforward resistance prediction map. The phase division strategy generation module analyzes the resistance change trend based on the feedforward resistance prediction map, automatically divides the sinking process into three stages: initial sinking, normal sinking, and final sinking, sets the upper limit of the bottom depth and the target speed threshold according to the resistance characteristics of each stage, and outputs a segmented adaptive strategy set. The real-time monitoring deviation analysis module collects data on the height difference at the four corners and the elevation of the cutting edge in real time according to the monitoring frequency set in the segmented adaptive strategy, and analyzes and calculates the deviation between the actual sinking speed and the target speed, as well as the current tilt, and outputs the deviation status vector. The deviation control closed-loop feedback module analyzes the deviation state vector, determines whether the current deviation is speed deviation or tilting limit, and generates control commands based on the segmented adaptive strategy set, including adjusting the bottom depth or initiating deviation correction. The corrected data is fed back to the real-time monitoring deviation analysis module, and the closed loop continues until the caisson is in place.

[0009] As a further aspect of the present invention: in the geological modeling resistance prediction module, the process of quantifying the spatial distribution and friction coefficient of interlayered miscellaneous fill, silty soil, and sandstone is as follows: Extract the friction coefficient and ultimate bearing capacity of the well wall and soil corresponding to each soil layer from the geotechnical test report. Use the borehole number and the elevation range of the top and bottom plates of the layer as the joint primary key to establish a structured data table containing the layer, depth range, friction coefficient and bearing capacity. Using the plane projection area of ​​the caisson as the boundary, the area is divided into planar grids with a spacing of two meters by two meters, and vertical layer spacing of half a meter. The ordinary Kriging interpolation method is used to generate a three-dimensional geological grid, thus completing the geological information model.

[0010] As a further aspect of the present invention: in the geological modeling resistance prediction module, the process of analyzing and calculating the probability distribution range of subsidence resistance includes: For each three-dimensional mesh cell that the caisson wall passes through, the friction coefficient and cell height are read, and the side friction of the micro-element segment is calculated. From the starting elevation of the cutting edge downwards to the current sinking depth, the side friction resistance of all the units passed through is summed to obtain the total cumulative friction resistance; At each sinking depth, the 15th and 85th percentiles of the cumulative total frictional resistance are taken as the lower and upper limits of the sinking resistance, respectively, forming a probability distribution interval.

[0011] As a further aspect of the present invention: the process of automatically dividing the three stages of initial settling, normal settling, and final settling in the stage division strategy generation module is as follows: The resistance mean curve is obtained from the feedforward resistance prediction map, and the rate of change of resistance with respect to depth is calculated point by point with a step size of half a meter to generate a slope sequence. The depth location where the absolute value of the slope change exceeds a preset threshold is marked as a resistance mutation node, and the interval between adjacent mutation nodes is divided into a stable segment. The section before the first abrupt drop in resistance is defined as the initial settling stage, the section up to the first two meters of the design depth is defined as the normal stage, and the last two meters are defined as the final settling stage.

[0012] As a further aspect of the present invention: in the stage division strategy generation module, the process of setting the upper limit of the pot bottom depth and the target speed threshold is as follows: Extract the characteristic resistance values ​​of all stable sections within each stage, calculate the average resistance, maximum resistance, and minimum resistance to obtain the resistance fluctuation amplitude; calculate the upper limit of the pot bottom depth based on the average resistance to achieve the reverse adjustment where the greater the resistance, the lower the upper limit of the pot bottom depth. The upper limit of the target speed is calculated based on the magnitude of the resistance fluctuation. The greater the resistance fluctuation, the lower the upper limit of the target speed. The lower limit of the target speed is set to 50% of the upper limit.

[0013] As a further aspect of the present invention: the process of real-time acquisition of the height difference at the four corners and the elevation of the cutting edge in the real-time monitoring deviation analysis module is as follows: Based on the current sinking depth, determine the current stage and extract the corresponding monitoring frequency from the segmented adaptive strategy set: once every ten minutes in the initial and final sinking stages, and once every thirty minutes in the normal stage. Displacement sensors are installed at the four corners of the top surface of the caisson. The elevation values ​​of the four corner points are read synchronously at set time intervals. At the same time, the current elevation value of the cutting edge position sensor is read as the basis data for subsequent deviation calculation.

[0014] As a further aspect of the present invention: in the real-time monitoring deviation analysis module, the process of calculating speed deviation and tilt is as follows: The actual sinking speed is calculated based on the difference in the elevation of the cutting edge and the time interval between two adjacent data collections. The speed deviation is obtained by comparing it with the upper and lower limits of the target speed at the current stage. Calculate the diagonal elevation difference using the elevations of the four corner points, take the maximum value and divide it by the horizontal distance between the corner points to obtain the inclination. Based on the current stage, the preset tilt warning threshold is read. If the tilt exceeds the threshold, the tilt is determined to be out of limit, a tilt alarm flag is generated, and all calculation results are encapsulated into a deviation state vector.

[0015] As a further aspect of the present invention: in the deviation control closed-loop feedback module, the priority determination process is as follows: Analyze the deviation state vector to extract the velocity deviation, tilt over-limit flag, maximum diagonal height difference, and the upper limit of the pot bottom depth at the current stage; If the tilt exceeding limit flag is true, calculate the tilt exceeding limit value and set it as tilt priority; if the tilt exceeding limit flag is false, calculate the speed deviation rate and set it as speed priority; under speed priority, generate instructions to adjust the bottom depth or use weight to assist settling according to the speed, and generate instructions to correct deviation from the soil under tilt priority.

[0016] As a further aspect of the present invention: in the deviation control closed-loop feedback module, the process of generating tilt-priority control commands is as follows: The side with more subsidence and the side with less subsidence are determined based on the maximum diagonal height difference. The correction force is calculated as the ratio of the tilt exceeding the limit to the warning threshold, and does not exceed 2.0. Increase excavation at the cutting edge on the side with more subsidence, with the excavation volume increasing by 30% of the correction force; stop excavation at the cutting edge on the side with less subsidence and backfill with sand and gravel, with the backfill height being 0.1 meters of the correction force, thus forming quantitative correction operation parameters.

[0017] As a further aspect of the present invention: in the deviation control closed-loop feedback module, the process of closed-loop feedback until the caisson is in place is as follows: The generated control commands are packaged into a standard format and sent to the on-site operation terminal for execution; the corrected four-corner elevation and cutting edge elevation data are fed back to the real-time monitoring deviation analysis module to recalculate the new deviation state vector; the priority judgment and control command generation steps are repeated until the current sinking depth reaches the design depth, and the tilt over-limit flag is zero and the speed deviation is zero, the caisson arrival signal is output, and the closed-loop control process ends.

[0018] The beneficial effects of this invention are as follows: First, in response to the problem that traditional construction methods use a single weighted average frictional resistance to calculate settlement and obscure the strong heterogeneous distribution characteristics of the soil layers, this invention establishes a three-dimensional geological information model and calculates the probability distribution range of settlement resistance, quantifies the spatial heterogeneity of mixed fill, silty soil, and sandstone interlayers, and outputs a feedforward resistance prediction map, providing a refined geological basis for velocity control.

[0019] Secondly, in response to the problem that traditional construction relies on manual periodic measurement and control with a serious lag in response, this invention automatically collects data according to the monitoring frequency set by the strategy set, calculates speed deviation and tilt in real time, and compresses the entire measurement-judgment-control cycle from minutes to seconds, thus achieving timely response.

[0020] Third, in response to the inherent coupling conflict between speed control and tilt correction in traditional construction, this invention analyzes the deviation state vector to determine the priority and generates speed control commands or correction commands respectively. It quantifies parameters such as excavation adjustment amount, backfill height, and compaction tonnage, decoupling the side effect of "correcting soil deviation" while accelerating settlement.

[0021] Fourth, in response to the problem that traditional construction lacks quantitative criteria and the risk of sudden sinking is difficult to predict, this invention establishes a quantitative model of the degree of deviation by calculating the upper limit of the bottom depth and the target speed threshold, thereby upgrading the control command from experience-based judgment to quantitative calculation, and ensuring the safety and accuracy of the caisson sinking process. Attached Figure Description

[0022] The invention will now be further described with reference to the accompanying drawings.

[0023] Figure 1 This is a flowchart of the module of the adaptive control system for caisson sinking speed based on geological information model as described in an embodiment of the present invention; Figure 2 This is a flowchart illustrating the implementation steps of the adaptive control system for caisson sinking speed based on a geological information model, as described in this embodiment of the invention. Figure 3 This is a logic judgment diagram of the adaptive control system for caisson sinking speed based on a geological information model, as described in an embodiment of the present invention. Detailed Implementation

[0024] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0025] For examples, please refer to Figures 1-3 As shown in the embodiment of the present invention, the adaptive control system for caisson sinking speed based on a geological information model includes the following modules: The geological modeling resistance prediction module, based on geological exploration layered data, quantifies the spatial distribution and friction coefficient of interlayered miscellaneous fill, silty soil, and sandstone, establishes a three-dimensional geological information model, analyzes and calculates the probability distribution range of subsidence resistance, and outputs a feedforward resistance prediction map. In the geological modeling resistance prediction module, the geological exploration stratification data includes stratigraphic stratification information for each borehole point, specifically including the top elevation, bottom elevation, and soil / rock name for each stratum. The state description of the soil / rock name is as follows: for miscellaneous fill, its looseness and content of fragmented stones are recorded; for silty soil, its fluid state, saturation, and organic matter content are recorded; and sandstone is classified according to its weathering degree as completely weathered (sandy), strongly weathered (fragmented), and moderately weathered (short columnar). In the geological modeling resistance prediction module, the process of quantifying the spatial distribution and friction coefficient of interlayered fill, silty soil, and sandstone is as follows: Extract the friction coefficient (in kPa) and ultimate bearing capacity (in kPa) between the wellbore and the soil for each soil layer from the geotechnical test report. Add the friction coefficient and ultimate bearing capacity as attribute fields to the stratigraphic stratification information. Specifically, the binding method is to establish a structured data table of "stratum-depth interval-friction coefficient-bearing capacity" using the borehole number and the elevation range of the top and bottom plates of the stratum as the joint primary key. In the geological modeling resistance prediction module, the process of establishing a three-dimensional geological information model is as follows: The planar grid area of ​​the caisson is used as the boundary, and the grid is divided with a 2m×2m spacing. Vertical layer spacing is 0.5m, and the area is discretized from the ground elevation to 5m below the designed bottom elevation of the caisson. Using ordinary Kriging interpolation, with structured data tables as sample points, spatial interpolation is performed on the soil layer type and mechanical properties of each grid unit to generate a three-dimensional geological grid, thus completing the geological information model. In the geological modeling resistance prediction module, the process of analyzing and calculating the probability distribution interval of subsidence resistance includes: S101: Calculate the side friction resistance of a single grid cell: For each three-dimensional grid cell that the caisson wall passes through, read the friction coefficient f_i (unit kPa) and cell height h_i (unit m) of the cell, and calculate the side friction resistance of the micro-segment F_i = f_i × h_i × unit perimeter (take 1m width). S102: Calculate the cumulative total frictional resistance by integrating along the depth: Starting from the initial elevation of the caisson cutting edge, and moving downwards to the current sinking depth, sum the side frictional resistance of all grid cells traversed: F_total(d) = ΣF_i, where d is the current sinking depth. Since the frictional coefficient of each grid cell in the geological information model is a spatial interpolation result, there are multiple possible frictional coefficient values ​​at each depth location. Therefore, F_total(d) presents a probability distribution rather than a single numerical value. S103: Generate the probability distribution interval of sinking resistance: At each sinking depth d, take the 15th percentile and 85th percentile of the cumulative total frictional resistance F_total(d) as the lower limit and upper limit of the sinking resistance at that depth, respectively, to form the probability distribution interval of sinking resistance as a function of depth. In the geological modeling resistance prediction module, the process of outputting the feedforward resistance prediction map is as follows: Plot three curves on the same coordinate system, with depth as the ordinate (in meters) and resistance value as the abscissa (in kN / m): the lower resistance limit curve (15th quantile), the mean resistance curve (50th quantile), and the upper resistance limit curve (85th quantile). Mark the locations of each soil layer interface on the curves (e.g., the boundary between miscellaneous fill and silty soil, and the boundary between silty soil and sandstone), forming a feedforward resistance prediction map that can be directly read by the stage division strategy generation module. Understandably, the significance of the geological modeling resistance prediction module lies in the fact that traditional caisson construction relies solely on a single weighted average skin friction provided by the geological survey report for sinking verification. This completely masks the highly heterogeneous spatial distribution characteristics of interlayered fill, silty soil, and sandstone, leading to severely inaccurate resistance predictions. This step, by establishing a three-dimensional geological information model, quantifies the spatial distribution and mechanical parameters of each soil layer, calculates the probability distribution range of sinking resistance instead of a single value, and outputs a feedforward resistance prediction map. This provides a refined geological basis for subsequent sinking speed control and solves the fundamental defect of traditional methods that "substitute uniformity for heterogeneity." The phase division strategy generation module analyzes the resistance change trend based on the feedforward resistance prediction map, automatically divides the sinking process into three stages: initial sinking, normal sinking, and final sinking, sets the upper limit of the bottom depth and the target speed threshold according to the resistance characteristics of each stage, and outputs a segmented adaptive strategy set. In the stage division strategy generation module, based on the analysis of the resistance change trend using the feedforward resistance prediction map, the sinking process is automatically divided into three stages: initial sinking, normal sinking, and final sinking. Read the feedforward resistance prediction map output by the geological modeling resistance prediction module to obtain the resistance mean curve (50th quantile curve). Calculate the rate of change of resistance with respect to depth point by point from the initial subsidence depth to the design subsidence depth, with a step size of 0.5m: k(d) = [F_total(d+0.5) - F_total(d)] / 0.5, in kN / m 2 Generate a sequence of slope changes with depth; Traverse the slope sequence k(d), and when the absolute value of the slope change between two adjacent points exceeds a preset threshold (e.g., 30 kN / m)... 2 When the slope changes from positive to negative or the positive value decreases beyond the threshold, mark the location at that depth as a resistance mutation node and determine the mutation type: if the slope changes from positive to negative or the positive value decreases beyond the threshold, mark it as a resistance decrease mutation; if the slope changes from negative to positive or the negative value increases beyond the threshold, mark it as a resistance increase mutation. The depth interval between adjacent mutation nodes is divided into a stable segment. For each stable segment, the arithmetic mean of the drag mean curve F_total(d) within that segment is calculated and used as the characteristic drag value F_avg of that segment. The starting depth, ending depth, and segment length of the segment are also recorded. Starting from the initial sinking depth, scan downwards to the first abrupt change in drag. The depth range preceding this abrupt change is defined as the initial sinking stage. The continuous stable section within 2m after the end of the initial sinking stage and before the designed sinking depth is defined as the normal sinking stage. The interval from 2m before the designed sinking depth to the designed sinking depth is divided into the final sinking stage; It should be noted that the initial settling stage is characterized by: resistance increasing rapidly with depth (large positive slope), the caisson not yet fully penetrating the stable soil layer, and poor stability. The output of this stage is the starting and ending depths of the initial settling stage. The normal settling stage is characterized by: resistance changing gradually with depth (absolute slope close to zero), the characteristic resistance value F_avg of each stable section fluctuating by no more than 15%, and the caisson sinking under uniform conditions. The output of this stage is the starting and ending depths of the normal settling stage. The final settling stage is characterized by: resistance potentially experiencing another abrupt change (e.g., entering the bearing layer), and strict control of oversinking is required; the sinking speed must be actively reduced. The output of this stage is the starting depth and designed sinking depth of the final settling stage. The division results are summarized to generate a three-stage (initial settlement, normal, and final settlement) division table, recording the starting depth, ending depth, segment length, and a list of characteristic resistance values ​​F_avg for all stable segments included in each stage. In the stage segmentation strategy generation module, the process of calculating and setting the upper limit of the pot bottom depth and the target speed threshold based on the resistance characteristics of each stage, and outputting the segmented adaptive strategy set is as follows: For each stage, extract the F_avg values ​​of all stable segments within that stage, and calculate the arithmetic mean of the resistance for that stage, F_stage_avg. Simultaneously, take the maximum F_stage_max and minimum F_stage_min values ​​of F_avg for each stable segment within that stage. Calculate the resistance fluctuation amplitude ΔF_stage = (F_stage_max - F_stage_min) / F_stage_avg × 100%; Using the average resistance F_stage_avg at each stage as the independent variable, the upper limit of the pot bottom depth D_bottom_max is calculated according to the following formula: D_bottom_max=D_base×(1-α×(F_stage_avg-F_ref) / F_ref); Where D_base is the baseline bottom depth (1.2m), F_ref is the reference resistance for normal sinking (150kPa), and α is the adjustment coefficient (0.25). Boundary constraints are applied to D_bottom_max after calculation: the upper limit for the initial sinking stage is no more than 0.9m, the upper limit for the normal stage is no more than 1.3m, and the upper limit for the final sinking stage is no more than 0.6m. This formula achieves a reverse adjustment relationship where the greater the resistance, the lower the upper limit of the bottom depth. Using the resistance fluctuation amplitude ΔF_stage at each stage as the independent variable, the target speed upper limit V_max is calculated using the following formula: V_max=V_base×(1-β×ΔF_stage / 100%); Where V_base is the upper limit of the baseline velocity (taken as 0.8 m / d), and β is the adjustment coefficient (taken as 0.3). The lower limit of the target velocity, V_min, is taken as 50% of V_max, and V_max is forced to not exceed 0.25 m / d during the final settling stage. This formula implements the control logic that the greater the drag fluctuation, the lower the upper limit of the target velocity, in order to ensure settling stability; The three-stage (initial sinking, normal, and final sinking) partitioning table is combined and encapsulated with the calculated upper limit of the bottom depth D_bottom_max, the upper limit of the target velocity V_max, and the lower limit of the target velocity V_min to obtain a segmented adaptive strategy set. The format of each strategy record in the segmented adaptive strategy set is: {stage number, starting depth, ending depth, F_stage_avg, ΔF_stage, D_bottom_max, V_min, V_max, priority control target}; where the priority control target is automatically assigned according to the stage type: "anti-deviation" for the initial sinking stage, "velocity" for the normal stage, and "anti-oversinking" for the final sinking stage. Understandably, the significance of the stage division strategy generation module lies in the fact that traditional construction relies on manual experience to determine settlement stages, lacking quantitative basis, and the settings for the bottom depth and target velocity are often fixed, failing to adapt to changes in resistance in different soil layers. This step automatically identifies nodes of sudden resistance changes based on the feedforward resistance prediction map, quantitatively divides the settlement into three stages: initial settlement, normal settlement, and final settlement, and calculates and sets the upper limit of the bottom depth and the target velocity threshold based on the average resistance and resistance fluctuation amplitude of each stage, achieving adaptive adjustment of control parameters according to changes in geological conditions. The real-time monitoring deviation analysis module collects data on the height difference at the four corners and the elevation of the cutting edge in real time according to the monitoring frequency set in the segmented adaptive strategy, and analyzes and calculates the deviation between the actual sinking speed and the target speed, as well as the current tilt, and outputs the deviation status vector. In the real-time monitoring deviation analysis module, the process of collecting the four corner elevation differences and the edge elevation data in real time according to the monitoring frequency set in the segmented adaptive strategy is as follows: The current sinking depth d_current determines the sinking stage (initial sinking / normal / final sinking), and the preset monitoring frequency f_monitor is extracted from the strategy record of that stage. The specific definition of the monitoring frequency is: once every 10 minutes in the initial sinking stage, once every 30 minutes in the normal stage, and once every 10 minutes in the final sinking stage; High-precision displacement sensors are installed at the four corners of the caisson's top surface, with each sensor monitoring the vertical elevation of that corner point in real time. Strain gauges are installed every 1 meter along the height of the outer side of the caisson wall to check the sinking resistance. According to the time interval set by f_monitor, synchronously read the current elevation values of the displacement sensors at four corner points, which are recorded as Z_A, Z_B, Z_C and Z_D respectively (unit: m). Simultaneously read the current elevation value Z_blade (unit: m) of the cutting edge position sensor; In the real-time monitoring deviation analysis module, the process of analyzing and calculating the deviation between the actual sinking speed and the target speed, as well as the current inclination, and outputting the deviation state vector is: Read the cutting edge elevation Z_blade(t) collected at the current time t, and the cutting edge elevation Z_blade(t-Δt) collected at the previous collection time t-Δt, where Δt is the time interval between two adjacent collections (unit: day). Calculate the current actual sinking speed V_actual by the following formula: V_actual=|Z_blade(t)-Z_blade(t-Δt)| / Δt; According to the sinking stage corresponding to the current sinking depth d_current, read the lower limit of target speed V_min and the upper limit of target speed V_max of this stage from the segmented adaptive strategy set, and calculate the deviation between the actual sinking speed and the target speed; If V_actual<V_min, then ΔV=V_actual-V_min (negative value indicates the speed is too slow); if V_actual>V_max, then ΔV=V_actual-V_max (positive value indicates the speed is too fast); if V_min≤V_actual≤V_max, then ΔV=0 (indicates that the speed is within the target range and no regulation is required) Taking the four corner elevations Z_A, Z_B, Z_C, Z_D as inputs, calculate the open caisson inclination δ by the following formula: First calculate the diagonal elevation difference: Δh_AC=|Z_A-Z_C|, Δh_BD=|Z_B-Z_D|; take the larger value as the maximum diagonal elevation difference Δh_max=max(Δh_AC,Δh_BD); inclination δ=Δh_max / L_diag, where L_diag is the horizontal distance between opposite corner points (unit: m), and δ is a dimensionless ratio; Read the preset inclination warning threshold δ_threshold in the segmented adaptive strategy set. According to different current sinking stages, δ_threshold takes different values: 0.5% for initial sinking and normal stages (corresponding to 30cm / 10m), and 0.04% for final sinking stage (corresponding to 4cm / 10m). If δ>δ_threshold, it is determined that the inclination exceeds the limit, and an inclination alarm flag Flag_tilt=1 is generated; otherwise Flag_tilt=0; Encapsulate all calculation results into a deviation state vector E, and the vector format is defined as: E=(d_current,V_actual,V_min,V_max,ΔV,δ,Δh_max,Flag_tilt, T_stamp), where d_current: current sinking depth (unit: m), V_actual: actual sinking speed (unit: m / d), V_min, V_max: target speed lower limit, upper limit (unit: m / d), ΔV: speed deviation (unit: m / d), δ: current tilt (dimensionless), Δh_max: maximum diagonal height difference (unit: m), Flag_tilt: tilt over-limit flag (0 or 1), T_stamp: data acquisition timestamp; Understandably, the significance of the real-time monitoring deviation analysis module lies in the fact that traditional construction relies on manual periodic measurements using leveling instruments, with measurement intervals as long as 10 to 30 minutes. Furthermore, the cumulative effect of manual recording, analysis, and command transmission throughout the entire cycle leads to a severe lag in control response, often resulting in passive correction only when the accumulated tilt reaches the warning value. This step automatically collects the elevation data of the four corners and the cutting edge according to the monitoring frequency set in the strategy set, calculates the deviation between the actual sinking speed and the target speed and the current tilt in real time, and outputs a structured deviation state vector, providing timely and accurate feedback input for closed-loop control. The deviation control closed-loop feedback module analyzes the deviation state vector, determines whether the current deviation is speed deviation or tilting limit, and generates control instructions based on the segmented adaptive strategy set, including adjusting the bottom depth or initiating deviation correction. The corrected data is fed back to the real-time monitoring deviation analysis module, and the closed loop continues until the caisson is in place. In the deviation control closed-loop feedback module, the process of generating control commands is as follows: S401, analyze the deviation state vector and extract the velocity deviation ΔV, tilt over-limit flag Flag_tilt, maximum diagonal height difference Δh_max, and current stage bottom depth limit D_bottom_max. S402, determine the priority and calculate the degree of deviation. If Flag_tilt=1, calculate the tilt over-limit value ε_tilt=δ-δ_threshold, set it as "tilt priority" and go to S404; if Flag_tilt=0, calculate the speed deviation rate η_v=|ΔV| / V_max, set it as "speed priority" and go to S403. S403: Generation of control instructions under speed priority. No control is needed when ΔV=0. When ΔV<0 (too slow), calculate the increase value of the pot bottom depth ΔH_avail=D_bottom_max-H_current. If ΔH_avail>0, generate the instruction to "deepen the pot bottom by 0.1-0.2m", otherwise generate the instruction to "start the weight-assisted sinking". When ΔV>0 (too fast), generate the instruction to "decrease the pot bottom depth by 0.1-0.2m". S404: Control command generation under tilt priority. Based on Δh_max, determine the side with more subsidence (side X) and the side with less subsidence (side Y); calculate the correction force γ=min(ε_tilt / δ_threshold,2.0), and generate the "Soil Extrusion Correction" command: increase excavation at the cutting edge of side X by γ×30%; ​​stop excavation at the cutting edge of side Y and backfill with sand and gravel to a height of γ×0.1m. If γ>1.5, additionally generate the "Apply Compactor to Side Y" command with a compactor weight of (γ-1)×5 tons. S405: Encapsulates and issues control commands. It encapsulates the commands generated by S403 or S404 into a standard format and issues them to the field operation terminal for execution. S406: Feedback the corrected data to the real-time monitoring deviation analysis module, wait for the next monitoring cycle, re-collect the four corner elevations and the cutting edge elevation data, and input them into the real-time monitoring deviation analysis module to calculate the new deviation state vector E_new; S407: Loop until the caisson is in place, repeat S401 to S406 until the current sinking depth reaches the design depth and Flag_tilt=0, ΔV=0, and output the "caisson in place" signal; Understandably, the significance of the deviation control closed-loop feedback module lies in the inherent conflict between sinking speed control and tilt correction in traditional construction. While correcting the tilt, soil excavation correction can locally deepen the bottom of the caisson, accelerating sinking on that side. Operators rely on experience to weigh these factors, lacking quantitative decoupling criteria. This step analyzes the deviation state vector to determine priorities, generating speed control or correction commands accordingly. It quantifies parameters such as excavation adjustment, backfill height, and counterweight tonnage, feeding the corrected data back to the real-time monitoring deviation analysis module to form a closed loop until the caisson is precisely in place.

[0026] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An adaptive control system for caisson sinking speed based on a geological information model, characterized in that: Includes the following modules: The geological modeling resistance prediction module, based on geological exploration layered data, quantifies the spatial distribution and friction coefficient of interlayered miscellaneous fill, silty soil, and sandstone, establishes a three-dimensional geological information model, analyzes and calculates the probability distribution range of subsidence resistance, and outputs a feedforward resistance prediction map. The phase division strategy generation module analyzes the resistance change trend based on the feedforward resistance prediction map, automatically divides the sinking process into three stages: initial sinking, normal sinking, and final sinking, sets the upper limit of the bottom depth and the target speed threshold according to the resistance characteristics of each stage, and outputs a segmented adaptive strategy set. The process of automatically dividing the three stages of initial settling, normal settling, and final settling in the stage division strategy generation module is as follows: The resistance mean curve is obtained from the feedforward resistance prediction map, and the rate of change of resistance with respect to depth is calculated point by point with a step size of half a meter to generate a slope sequence. The depth location where the absolute value of the slope change exceeds a preset threshold is marked as a resistance mutation node, and the interval between adjacent mutation nodes is divided into a stable segment. The section before the first abrupt drop in resistance is defined as the initial settling stage, the section up to the first two meters of the design depth is defined as the normal stage, and the last two meters are defined as the final settling stage. In the phase division strategy generation module, the process of setting the upper limit of the pot bottom depth and the target speed threshold is as follows: Extract the characteristic resistance values ​​of all stable segments within each stage, calculate the average resistance, maximum resistance, and minimum resistance, and obtain the resistance fluctuation amplitude; The upper limit of the pot bottom depth is calculated based on the average resistance, thus achieving a reverse adjustment where the higher the resistance, the lower the upper limit of the pot bottom depth. The upper limit of the target speed is calculated based on the magnitude of the resistance fluctuation. The greater the resistance fluctuation, the lower the upper limit of the target speed. The lower limit of the target speed is set to 50% of the upper limit. The real-time monitoring deviation analysis module collects data on the height difference at the four corners and the elevation of the cutting edge in real time according to the monitoring frequency set in the segmented adaptive strategy, and analyzes and calculates the deviation between the actual sinking speed and the target speed, as well as the current tilt, and outputs the deviation status vector. The deviation control closed-loop feedback module analyzes the deviation state vector, determines whether the current deviation is speed deviation or tilting limit, and generates control commands based on the segmented adaptive strategy set, including adjusting the bottom depth or initiating deviation correction. The corrected data is fed back to the real-time monitoring deviation analysis module, and the closed loop continues until the caisson is in place.

2. The adaptive control system for caisson sinking speed based on a geological information model according to claim 1, characterized in that: In the geological modeling resistance prediction module, the process of quantifying the spatial distribution and friction coefficient of interlayered miscellaneous fill, silty soil, and sandstone is as follows: Extract the friction coefficient and ultimate bearing capacity of the well wall and soil corresponding to each soil layer from the geotechnical test report. Use the borehole number and the elevation range of the top and bottom plates of the layer as the joint primary key to establish a structured data table containing the layer, depth range, friction coefficient and bearing capacity. Using the plane projection area of ​​the caisson as the boundary, the area is divided into planar grids with a spacing of two meters by two meters, and vertical layer spacing of half a meter. The ordinary Kriging interpolation method is used to generate a three-dimensional geological grid, thus completing the geological information model.

3. The adaptive control system for caisson sinking speed based on a geological information model according to claim 2, characterized in that: In the geological modeling resistance prediction module, the process of analyzing and calculating the probability distribution interval of subsidence resistance includes: For each three-dimensional mesh cell that the caisson wall passes through, the friction coefficient and cell height are read, and the side friction of the micro-element segment is calculated. From the starting elevation of the cutting edge downwards to the current sinking depth, the side friction resistance of all the units passed through is summed to obtain the total cumulative friction resistance; At each sinking depth, the 15th and 85th percentiles of the cumulative total frictional resistance are taken as the lower and upper limits of the sinking resistance, respectively, forming a probability distribution interval.

4. The adaptive control system for caisson sinking speed based on a geological information model according to claim 1, characterized in that: In the real-time monitoring deviation analysis module, the process of collecting real-time data on the height difference at the four corners and the elevation of the cutting edge is as follows: Based on the current sinking depth, determine the current stage and extract the corresponding monitoring frequency from the segmented adaptive strategy set: once every ten minutes in the initial and final sinking stages, and once every thirty minutes in the normal stage. Displacement sensors are installed at the four corners of the top surface of the caisson. The elevation values ​​of the four corner points are read synchronously at set time intervals. At the same time, the current elevation value of the cutting edge position sensor is read as the basis data for subsequent deviation calculation.

5. The adaptive control system for caisson sinking speed based on a geological information model according to claim 4, characterized in that: In the real-time monitoring deviation analysis module, the process of calculating speed deviation and tilt is as follows: The actual sinking speed is calculated based on the difference in the elevation of the cutting edge and the time interval between two adjacent data collections. The speed deviation is obtained by comparing it with the upper and lower limits of the target speed at the current stage. Calculate the diagonal elevation difference using the elevations of the four corner points, take the maximum value and divide it by the horizontal distance between the corner points to obtain the inclination. Based on the current stage, the preset tilt warning threshold is read. If the tilt exceeds the threshold, the tilt is determined to be out of limit, a tilt alarm flag is generated, and all calculation results are encapsulated into a deviation state vector.

6. The adaptive control system for caisson sinking speed based on a geological information model according to claim 1, characterized in that: In the deviation control closed-loop feedback module, the priority determination process is as follows: Analyze the deviation state vector to extract the velocity deviation, tilt over-limit flag, maximum diagonal height difference, and the upper limit of the pot bottom depth at the current stage; If the tilt exceeding limit flag is true, calculate the tilt exceeding limit value and set it as tilt priority; if the tilt exceeding limit flag is false, calculate the speed deviation rate and set it as speed priority; under speed priority, generate instructions to adjust the bottom depth or use weight to assist settling according to the speed, and generate instructions to correct deviation from the soil under tilt priority.

7. The adaptive control system for caisson sinking speed based on a geological information model according to claim 6, characterized in that: In the deviation control closed-loop feedback module, the process of generating tilt-priority control commands is as follows: The side with more subsidence and the side with less subsidence are determined based on the maximum diagonal height difference. The correction force is calculated as the ratio of the tilt exceeding the limit to the warning threshold, and does not exceed 2.

0. Increase excavation at the cutting edge on the side with more subsidence, with the excavation volume increasing by 30% of the correction force; stop excavation at the cutting edge on the side with less subsidence and backfill with sand and gravel, with the backfill height being 0.1 meters of the correction force, thus forming quantitative correction operation parameters.

8. The adaptive control system for caisson sinking speed based on a geological information model according to claim 7, characterized in that: In the deviation control closed-loop feedback module, the process from closed-loop feedback until the caisson is in place is as follows: The generated control commands are packaged into a standard format and sent to the on-site operation terminal for execution; the corrected four-corner elevation and cutting edge elevation data are fed back to the real-time monitoring deviation analysis module to recalculate the new deviation state vector; the priority judgment and control command generation steps are repeated until the current sinking depth reaches the design depth, and the tilt over-limit flag is zero and the speed deviation is zero, the caisson arrival signal is output, and the closed-loop control process ends.

Images