Hydraulic ring geological drilling parameter self-adaptive adjustment method

Abstract

The invention discloses a hydraulic ring geological drilling parameter adaptive adjustment method, which relates to the technical field of geology, and comprises the following steps: before drilling, based on geological survey data, obtaining a first stratum sequence of a to-be-drilled stroke, the first stratum sequence comprising a plurality of sequentially arranged silty clay layers and corresponding first characteristic parameters thereof, the first characteristic parameter at least comprises a first slurry making potential parameter; pre-generating a dynamic flow control sequence according to the first stratum sequence and the first slurry making potential parameter of each stratum in the first stratum sequence; according to the method, self-adaptive optimization of the clean water drilling parameters of the silty clay stratum is achieved, the problems of drill jamming and sediment accumulation caused by slurry viscosity in traditional fixed flow operation can be solved, energy consumption waste and hole wall collapse risks caused by too large flow can be solved, and drilling operation is more efficient and safer.

Description

Technical Field

[0001] This invention belongs to the field of geological drilling technology, specifically a method for adaptive adjustment of hydrogeological drilling parameters. Background Technology

[0002] Hydrogeological drilling is a core method in geological exploration and groundwater remediation, and its quality directly affects the accuracy of geological data and the rationality of subsequent engineering designs. With increasingly stringent environmental regulations, operations in sensitive areas such as urban suburbs and drinking water source protection zones face stringent requirements for groundwater pollution control. Traditional chemically modified drilling mud is prone to leakage and pollution and is no longer suitable for such scenarios. Clear water drilling, due to its advantages of being free of chemical additives and environmentally friendly, has become the preferred technology for sensitive areas.

[0003] Specifically, the pressure inside the borehole is balanced by circulating clean water, which cools the drill bit and carries rock cuttings. Continuous operation is achieved by relying on the fluidity of the clean water, effectively avoiding pollution risks and meeting environmental protection requirements.

[0004] Silty clay strata are common in sensitive areas, and their inherent characteristics include high clay particle content and easy softening and disintegration when exposed to water.

[0005] Based on the above, existing clear water drilling still uses conventional circulation parameters (such as fixed flow rate) for sand and rock layers, without specifically adapting to the changes in characteristics of silty clay after contact with clear water.

[0006] During operation, after water seeps in, the clay particles absorb water, soften, and disperse. Some of them adhere to the borehole wall to form a dense mud cake, causing a series of problems: the borehole diameter narrows, which makes it difficult to lower the drill rod, difficult to pull up the drill, and even jam the drill; the mud cake covers the rock core, resulting in incomplete sampling and attached impurities, affecting the accuracy of geological stratification and soil bearing capacity calculation.

[0007] Current operations often involve increasing the flow rate to flush away the mud cake. However, silty clay has low strength, and increased water flow impact can exacerbate borehole wall collapse, creating a technical contradiction: "Clearing the mud cake leads to borehole collapse, while not clearing it results in reduced diameter and distorted sampling." Summary of the Invention

[0008] The purpose of this invention is to provide an adaptive adjustment method for hydrogeological drilling parameters to solve the problems mentioned in the background art.

[0009] An adaptive adjustment method for hydrogeological drilling parameters is applied to a drilling system for clear water drilling in silty clay formations. The method includes: Before drilling begins, based on geological survey data, the first stratigraphic sequence of the drilling path is obtained. The first stratigraphic sequence contains multiple silty clay layers arranged in sequence and their corresponding first characteristic parameters. The first characteristic parameters include at least the first slurry-making potential parameter. Based on the first formation sequence and the first slurry-making potential parameters of each formation therein, a dynamic flow control sequence is pre-generated; the dynamic flow control sequence defines the target flow rate value that should be used for clear water circulation when drilling at different depths or in different formations; During drilling operations, the operating system is controlled to dynamically adjust the clean water circulation flow rate to the corresponding target flow rate value according to the dynamic flow control sequence.

[0010] Preferably, the step of pre-generating a dynamic flow control sequence includes: For each stratum in the first stratigraphic sequence, a basic target flow rate value is assigned to the stratum based on the first slurry potential parameter of that stratum; Based on the chronological order of the first formation sequence, the basic target flow rate values ​​of adjacent formations are smoothed and optimized to generate a curve showing the change of the target flow rate value with drilling depth. This curve has the minimum integral value on the drilling depth axis and satisfies the condition that the flow rate change rate does not exceed a preset stability threshold, thereby generating the final target flow rate value.

[0011] Preferably, the step of generating a curve showing the target flow rate as a function of drilling depth, wherein the integral value of this curve on the drilling depth axis is minimized, is achieved in the following manner: The flow rate curve for the entire drilling process is modeled as a polygon composed of flow rate values ​​from multiple formation nodes. The vertex coordinates of this polygon are optimized using an iterative algorithm to minimize the total area of ​​the polygon while satisfying the constraint that the absolute value of the slope of the line connecting all adjacent vertices is not greater than the stability threshold.

[0012] Preferably, the step of optimizing the basic target flow values ​​of adjacent strata based on the chronological order of the first stratigraphic sequence further includes: Before drilling operations begin, a flow transition interval is pre-defined for each pair of adjacent formations; The length of the flow transition interval is set according to the absolute value of the difference between the first slurry potential parameters of two adjacent strata. The larger the difference, the longer the preset transition interval. During drilling, when the drill bit enters the flow transition zone, the clear water circulation flow rate is controlled to gradually change from the target flow rate value of the current formation to the target flow rate value of the next formation before drilling through the interface between the two formations, with a flow rate change rate not exceeding the stability threshold.

[0013] Preferably, the method further includes: For at least one critical stratum node in the dynamic flow control sequence, at least one backup control branch is pre-generated; The step of generating the backup control branch includes: calculating the probability of working condition risk based on the first characteristic parameter of the key formation node using the historical drilling database; and setting a corresponding corrected flow control strategy for abnormal working conditions where the probability exceeds a preset risk threshold.

[0014] Preferably, in the step of calculating the probability of working condition risk based on the historical drilling database, if the key formation node has no match in the historical database or the matching data is insufficient, the following steps are performed: Based on the first feature parameters of the key formation nodes, a similarity search is performed in the drilling knowledge graph, which constructs the correlation between formation features, drilling parameters and working condition risks. Obtain the top N historical stratigraphic nodes with the highest similarity to the current key stratigraphic node features and their associated abnormal working condition records; Based on the frequency of abnormal operating conditions associated with the N historical stratigraphic nodes, and combined with the feature similarity between the current node and each historical node, a weighted calculation is performed to deduce the probability of operating condition risk of the current key stratigraphic node.

[0015] Preferably, the method further includes: During drilling operations, a second real-time working condition parameter is acquired to characterize the current borehole wall condition and cuttings carrying status. The second real-time working condition parameter includes at least drill pipe lifting resistance, return water consistency, and return sand volume. The second real-time operating condition parameter is compared with the expected state based on the dynamic flow control sequence; When the comparison result triggers one of the abnormal operating conditions, the corresponding backup control branch is invoked to overwrite the pre-generated target flow value.

[0016] Preferably, the real-time acquisition of drill pipe lifting resistance is achieved through the following methods: A pressure sensor is installed on the drilling rig lifting system, and the peak hydraulic pressure at the initial moment of the uniform lifting action of the drill rod at the borehole opening when it is stationary is collected as a characterization value of the lifting resistance of the drill rod.

[0017] Preferably, the abnormal operating conditions include at least the excessive mud cake development condition and the borehole wall instability condition; The criteria for judging the excessive mud cake condition are: the drill rod lifting resistance continues to increase, and at the same time, the ratio of return water flow to return sand volume shows a monotonically increasing trend within M consecutive sampling periods. The criteria for determining the instability of the borehole wall are as follows: within a single sampling period, the real-time measured value of the return water consistency increases by more than 20% of its normal operating condition baseline value relative to the previous sampling period, and the percentage of particles larger than the normal cutting rock chip size in the particle size distribution of the current return sand exceeds K times its historical average level, where K>1.

[0018] Preferably, the step of invoking the corresponding backup control branch further includes: Based on the target flow rate currently being executed and the preset correction flow rate in the backup control branch, calculate a flow rate adjustment range; If the flow adjustment range exceeds a preset mutation threshold, the flow adjustment process will be decomposed into multiple sub-steps for execution. The number of sub-steps and the flow rate adjustment amount for each step are calculated using a model predictive control algorithm. The objective function of the model predictive control algorithm is to minimize the sum of squares of the deviations between the predicted and set values ​​of the bottom hole pressure in the prediction time domain. Its constraints include the stability threshold and the maximum power of the circulating pump. The clean water circulation flow rate is controlled to be gradually adjusted to the corrected flow rate value within multiple consecutive adjustment cycles at a flow rate change rate not exceeding the stability threshold.

[0019] Compared with the prior art, the beneficial effects of the present invention are: This invention obtains the first formation sequence and first mud-making potential parameters based on geological survey data before drilling, pre-generates a dynamic flow control sequence adapted to different formations, and then dynamically adjusts the clear water circulation flow rate during drilling. This achieves adaptive optimization of clear water drilling parameters in silty clay formations. It not only solves the problems of stuck drill and sediment accumulation caused by viscous mud in traditional fixed flow rate operations, as well as the energy waste and borehole collapse risks caused by excessive flow rate, making drilling operations more efficient and safer; but also, the dynamic flow rate adjustment strategy fits the mud-making characteristics of silty clay formations, eliminating the need for additional drilling fluid treatment agents and meeting environmental protection requirements. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the method framework structure of the present invention. Detailed Implementation

[0021] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0022] Please see Figure 1 This application provides an adaptive adjustment method for hydrogeological drilling parameters, applied to a drilling system for clear water drilling in silty clay formations, including: Before drilling begins, based on geological survey data, the first stratigraphic sequence of the drilling path is obtained. The first stratigraphic sequence contains multiple silty clay layers arranged in sequence and their corresponding first characteristic parameters. The first characteristic parameters include at least the first slurry-making potential parameter. In this embodiment, the first stratigraphic sequence is an ordered description of the stratigraphic distribution throughout the entire drilling route. It not only clarifies the order of the silty clay layers, but also marks basic information such as the thickness and burial depth of each layer. For example, "the first silty clay layer is buried at a depth of 0-5m with a thickness of 5m; the second silty clay layer is buried at a depth of 5-12m with a thickness of 7m."

[0023] The first characteristic parameter is a key indicator that characterizes the drilling adaptability of silty clay layers. Among them, the first mud-making potential parameter directly reflects the mud-making capacity of the formation. The larger the value, the faster the formation forms mud and the higher the mud viscosity after encountering water in clear water drilling.

[0024] The first pulping potential parameter is obtained based on the clay particle content, liquid limit, and plastic limit indices from geological exploration data, and is calculated using a preset quantitative model. The quantitative model is: First pulping potential parameter P = k1 × C + k2 × (WL - WP), where C is the mass fraction of clay particles, WL is the liquid limit, and WP is the plastic limit.

[0025] The calibration methods for coefficients k1 and k2 are as follows: By conducting indoor water drilling simulation experiments on typical silty clay samples in the target area, the actual mud-making rate and mud viscosity were measured, and the results were obtained by fitting using the multiple linear regression method.

[0026] As a preferred implementation method, in the silty clay strata of the North China Plain, the typical values ​​determined by fitting are k1=0.8 and k2=0.3. For different geological regions, it is recommended to collect representative samples for calibration before drilling to obtain the coefficients most suitable for local conditions.

[0027] In cases where geological exploration data lacks liquid limit or plastic limit indicators, supplementary testing can be conducted by selecting samples of similar strata from adjacent boreholes in the area. Localized empirical correlations can be established for estimation, such as using the regional empirical relationship between clay particle content and (liquid limit - plastic limit) for estimation.

[0028] Based on the first formation sequence and the first slurry potential parameters of each formation, a dynamic flow control sequence is pre-generated; the dynamic flow control sequence defines the target flow rate value that should be used for clear water circulation when drilling at different depths or in different formations. In this embodiment, the dynamic flow control sequence is a flow configuration table that corresponds one-to-one with the first formation sequence. It contains the core contents of drilling depth range, corresponding formation, first slurry potential parameter value, and target flow value. For example, "depth 0-5m, first silty clay layer, first slurry potential parameter 3.2, target flow value 30L / min; depth 5-12m, second silty clay layer, first slurry potential parameter 4.8, target flow value 45L / min".

[0029] The generation of the dynamic flow control sequence follows the principle that the larger the first slurry potential parameter, the higher the corresponding target flow rate, thereby carrying away the dispersed clay particles in time through a larger circulation flow rate and avoiding excessive slurry viscosity; while the smaller the first slurry potential parameter, the lower the corresponding target flow rate, reducing energy consumption and borehole wall erosion risk while meeting the sediment carrying requirements.

[0030] The specific steps for pre-generating a dynamic flow control sequence include: For each stratum in the first stratigraphic sequence, a basic target flow rate value is assigned to the stratum based on the first slurry potential parameter of that stratum; In this embodiment, the allocation logic of the basic target flow rate value is consistent with the quantitative correlation of the first slurry potential parameter. It still needs to use 30 L / min, corresponding to the benchmark value of 3.0 of the first slurry potential parameter, as the basic flow rate threshold. For every 10% increase in the deviation ratio, the basic target flow rate value increases by 8%, and for every 10% decrease in the deviation ratio, the basic target flow rate value decreases by 5%. For example, if the first slurry potential parameter of a certain formation is 4.5 (50% higher than the benchmark value), then its basic target flow rate value = 30 L / min × (1 + 50% × 0.8) = 42 L / min; If the first slurry potential parameter of a certain formation is 2.4 (20% less than the benchmark value), then its basic target flow rate value = 30L / min × (1 - 20% × 0.5) = 27L / min, ensuring that the basic target flow rate value can accurately match the slurry characteristics of a single formation.

[0031] It should be understood that this quantitatively correlated allocation method not only achieves basic matching between flow rate and pulping capacity, but also makes flow rate adjustment predictable through the correspondence between the fixed deviation ratio and the flow rate adjustment ratio, avoiding flow rate allocation deviations caused by human experience judgment. At the same time, it provides a unified and stable initial data foundation for subsequent curve optimization, improving the reliability of the optimization results.

[0032] Based on the order of the first formation sequence, the basic target flow rate values ​​of adjacent formations are smoothed and optimized to generate a curve of target flow rate value changing with drilling depth. The integral value of this curve on the drilling depth axis is minimized, and the flow rate change rate does not exceed a preset stability threshold, thereby generating the final target flow rate value. In this embodiment, the minimum integral value of the curve on the drilling depth axis means that the area enclosed by the curve formed by the change of the target flow rate with depth and the depth axis is the smallest throughout the entire drilling stroke. The most important purpose is to minimize the total energy consumption and avoid unnecessary flow waste while meeting the formation slurry adaptation requirements.

[0033] The specific calculation method for the rate of change of flow rate is (the difference in target flow rate between adjacent depth points) ÷ (the difference in depth between adjacent depth points). The preset stability threshold is experimentally calibrated to be 5L / min·m, that is, for every 1m of drilling, the change in the target flow rate value shall not exceed 5L / min, in order to prevent sudden changes in flow rate from impacting the circulation system inside the borehole and to ensure drilling stability.

[0034] The step of generating a curve showing the target flow rate as a function of drilling depth, with the integral value of this curve minimized on the drilling depth axis, is achieved as follows: Specifically, the flow curve of the entire drilling process is modeled as a polygon composed of multiple flow values ​​of formation nodes. Formation nodes include the starting depth point, ending depth point and key feature points within each silty clay layer (in this embodiment, key feature points within the layer are set at intervals of 2m). The coordinates of each node are represented by "depth value (horizontal axis) - target flow value (vertical axis)". Subsequently, the vertex coordinates of the polygon are optimized using an iterative algorithm. The iterative algorithm employs a constrained gradient descent method, and the specific steps are as follows: Initialization: The basic target flow value of each layer node is used as the initial value of the ordinate of each vertex. The initial iteration step size is set to δ = 0.5L / min, the convergence threshold is set to ε = 0.1L·m, and the maximum number of iterations is set to N_max = 200.

[0035] Iterative process: In each iteration, the ordinate (i.e., the target flow value) of each vertex is optimized sequentially according to vertex order. For the current vertex i, the partial derivative (numerical gradient) of the objective function (total polygon area) with respect to the flow value F_i is calculated, which is approximately the change in total area ΔS when F_i increases by δ.

[0036] Coordinate Update and Constraint Check: Update F_i according to the gradient direction: F_i' = F_i - η * ΔS, where η is the learning rate, initially set to 1.0. After the update, immediately check whether the slopes of all affected adjacent edges (i.e., edges (i-1,i) and (i,i+1)) satisfy the stability threshold constraint. If any constraint is violated, the update is canceled, and the process is retried with a step size of η / 2 until the constraint is satisfied or the step size is less than 0.1.

[0037] Convergence criterion: After completing one full vertex traversal, calculate the change in the total area of ​​the polygon. If the absolute value of the change is less than ε, or the number of iterations reaches N_max, stop the iteration and output the current flow value of each vertex as the final target flow value; otherwise, proceed to the next iteration.

[0038] To facilitate understanding, an example is provided for explanation. For instance, the basic target flow rates of two adjacent strata are 30 L / min (stratum thickness 5m, burial depth 0-5m) and 42 L / min (stratum thickness 7m, burial depth 5-12m), respectively. The initial polygon vertices are (0,30), (5,30), (5,42), and (12,42). The absolute value of the slope of the line connecting the adjacent vertices (5,30) and (5,42) is infinite, exceeding the constraint. After iterative optimization, the vertices were adjusted to (0,30), (4.5,30), (5.5,42), and (12,42). The slope of the line connecting the adjacent vertices (4.5,30) and (5.5,42) is (42-30)÷(5.5-4.5)=12L / min・m, which still exceeds the constraint. Continue iteratively adjusting the vertex (5.5, 42) to (5.5, 36). At this point, the slope of the line is (36-30) ÷ (5.5-4.5) = 6L / min·m, which still does not meet the requirements. Adjust it again to (5.5, 35), with a slope of 5L / min·m, which meets the constraints. The total area of ​​the polygon is also reduced compared to the initial state, finally forming an optimized curve that meets the requirements.

[0039] The innovations brought about by this optimization process are: First, through polygon modeling and iterative optimization, the flow curve not only fits the actual slurry production needs of the formation, but also minimizes energy consumption. According to actual measurement comparison, compared with the traditional fixed flow mode, the total energy consumption can be reduced. Secondly, the design of the slope constraint between adjacent vertices not only avoids the impact of sudden flow changes on the circulation inside the hole, but also reduces the frequent start-stop and power changes of the circulating pump, which can extend the service life of the pump body and reduce equipment maintenance costs. Third, the setting of key feature points within the layer enables the flow rate adjustment to accurately adapt to the possible local differences in slurry production capacity within the formation, avoiding local sediment accumulation or borehole wall erosion caused by uneven parameters within the layer, improving the verticality and integrity of the borehole wall, and laying a better foundation for subsequent geological sampling or engineering construction.

[0040] It should be noted that the specific method for determining the target flow rate is as follows: First, a basic flow rate threshold is set, that is, when the first pulping potential parameter is the baseline value (the baseline value is set to 3.0 in this embodiment), the target flow rate is 30L / min; Then, adjust the flow rate according to the deviation ratio between the first pulping potential parameter and the benchmark value. For every 10% increase in the deviation ratio, the target flow rate value increases by 8% accordingly, and for every 10% decrease in the deviation ratio, the target flow rate value decreases by 5% accordingly, to ensure that the flow rate adjustment is both adapted to the pulping characteristics and avoids excessive fluctuations.

[0041] The generation of the dynamic flow control sequence can be automatically completed by the control module of the drilling operation system. The control module has built-in the above calculation logic, parameter calibration coefficients and gradient descent iterative algorithm. After inputting the first formation sequence and the corresponding first slurry potential parameters, it can output the complete dynamic flow control sequence in real time. At the same time, it supports manual fine-tuning within ±10% according to the actual geological conditions on site.

[0042] During drilling operations, the control system dynamically adjusts the clean water circulation flow rate to the corresponding target flow rate value according to the dynamic flow control sequence.

[0043] In this embodiment, the flow control unit of the operating system is linked with the drilling depth detection unit. The depth detection unit collects the current drilling depth in real time with an accuracy of ±0.1m and feeds the depth data back to the control module in real time.

[0044] The control module matches the corresponding interval in the dynamic flow control sequence based on the depth data, generates a flow adjustment command and sends it to the flow control unit. The flow control unit adjusts the flow by adjusting the output power of the circulating pump. The flow adjustment accuracy can reach ±1L / min, ensuring that the deviation between the actual flow and the target flow value is controlled within 5%.

[0045] It should be understood that the circulation pump has a maximum output power corresponding to the maximum value of the clean water circulation flow rate, which serves as a hardware constraint for flow rate adjustment.

[0046] When drilling crosses the interface of different silty clay layers, the control module adopts a gradual flow rate adjustment strategy: within a depth range of 0.5m above and below the interface, the target flow rate value is gradually transitioned from the set value of the upper stratum to the set value of the lower stratum to avoid the impact of sudden flow rate changes on the circulation stability in the borehole. For example, when transitioning from 30L / min in the first layer to 45L / min in the second layer, the flow rate is increased at a rate of 3L / min·m to ensure a smooth transition of the mud performance in the borehole.

[0047] If the flow rate fails to reach the target value due to sudden events such as abnormal borehole pressure or sudden changes in sediment volume during drilling, the control module will automatically issue an alarm signal, adjust the flow rate to the emergency flow rate corresponding to the formation (15% higher than the target flow rate), and prompt the operator to check for abnormalities. After the fault is eliminated, the system will resume operation according to the dynamic flow control sequence.

[0048] This invention obtains the first formation sequence and first mud-making potential parameters based on geological survey data before drilling, pre-generates a dynamic flow control sequence adapted to different formations, and then dynamically adjusts the clear water circulation flow rate during drilling. This achieves adaptive optimization of clear water drilling parameters in silty clay formations. It not only solves the problems of stuck drill and sediment accumulation caused by viscous mud in traditional fixed flow rate operations, as well as the energy waste and borehole collapse risks caused by excessive flow rate, making drilling operations more efficient and safer; but also, the dynamic flow rate adjustment strategy fits the mud-making characteristics of silty clay formations, eliminating the need for additional drilling fluid treatment agents and meeting environmental protection requirements.

[0049] In clear water drilling in silty clay formations, the initial slurry-making potential parameters of adjacent formations often differ, resulting in different basic target flow rates. If the flow rate curve is only optimized through preliminary polygon modeling without designing a specific transition mechanism for the formation interface, problems may still arise when drilling crosses the interface.

[0050] Because the stratigraphic interface is a region of abrupt change in geological properties, with significant differences in clay particle content, liquid limit, and plastic limit between adjacent strata, abrupt changes in mud-making capacity can easily lead to sudden changes in the viscosity and fluidity of the drilling mud inside the borehole. If the flow rate jumps directly from the target value of the current stratum to the target value of the next stratum, it will cause turbulence in the circulating flow field inside the borehole, resulting in the accumulation of sediment at the interface and increasing the risk of stuck drill bit.

[0051] The bonding strength of the borehole wall at the interface is usually weak. Pressure fluctuations inside the borehole caused by sudden changes in flow rate may damage the stability of the borehole wall, especially in deep drilling or loose silty clay formations, which may easily lead to borehole wall collapse.

[0052] As one embodiment of the present invention, the step of smoothly transitioning and optimizing the basic target flow values ​​of adjacent strata based on the sequential order of the first stratigraphic sequence further includes: Before drilling operations begin, a flow transition interval is pre-defined for each pair of adjacent formations; The length of the flow transition interval is set according to the absolute value of the difference between the first slurry potential parameters of two adjacent strata. The larger the difference, the longer the preset transition interval. In this embodiment, the preset flow transition interval needs to be completed before the drilling operation begins, and a separate flow transition interval is set for each pair of adjacent strata. The flow transition interval refers to the depth range located near the interface between two adjacent strata, used to achieve a smooth and gradual change in flow. Its starting depth is the interface depth minus half the length of the transition interval, and its ending depth is the interface depth plus half the length of the transition interval, ensuring that the transition intervals are symmetrically distributed around the interface.

[0053] The length of the flow transition interval is set based on the absolute value of the difference between the first slurry-making potential parameters of two adjacent strata, specifically through quantification rules: First, set the basic transition length L0 = 1.0m, which corresponds to the absolute value of the first pulping potential parameter difference ΔP = 0.5 (16.7% of the parameter baseline value 3.0). When ΔP < 0.5, the length of the transition interval is fixed at L0 = 1.0m, which is suitable for scenarios where the difference in slurry production capacity between the two formations is small, thus avoiding an increase in energy consumption due to an excessively long transition interval. When 0.5 ≤ ΔP ≤ 2.0, the transition interval length L = L0 + (ΔP - 0.5) × 0.8m, that is, for every 0.1 increase in the difference, the length increases by 0.08m. For example, when ΔP = 1.0, L = 1.0 + (1.0 - 0.5) × 0.8 = 1.4m, and when ΔP = 2.0, L = 1.0 + (2.0 - 0.5) × 0.8 = 2.2m. When ΔP > 2.0, the upper limit of the transition interval length is set to 3.0m to avoid the transition interval being too long and affecting drilling efficiency. At the same time, a smooth transition is ensured by optimizing the flow rate change rate.

[0054] To facilitate understanding, let's illustrate with an example. We assume that the first slurry-making potential parameter of the first stratum (0-5m burial depth) is 3.2, and the first slurry-making potential parameter of the second stratum (5-12m burial depth) is 4.8. ΔP = |4.8-3.2| = 1.6, which is in the range of 0.5-2.0. The length of the transition interval L = 1.0 + (1.6-0.5) × 0.8 = 1.88m, the starting depth = 5 - 1.88 / 2 = 4.06m, and the ending depth = 5 + 1.88 / 2 = 5.94m. That is, the transition interval is 4.06-5.94m.

[0055] During drilling, when the drill bit enters the flow transition zone, the clean water circulation flow rate is controlled to gradually change from the target flow rate of the current stratum to the target flow rate of the next stratum before drilling through the interface between the two strata, with a flow rate change rate not exceeding the stability threshold.

[0056] In this embodiment, during the drilling process, the depth detection unit collects the current drilling depth in real time. When the drill bit is detected to have entered the flow transition interval (i.e., the current depth is greater than or equal to the starting depth of the transition interval), the control module starts the flow gradual change control logic.

[0057] First, calculate the total amplitude of the flow rate change. Total amplitude = target flow rate value of the next formation - target flow rate value of the current formation. If the total amplitude is positive, the flow rate shows an increasing gradual change; if it is negative, the flow rate shows a decreasing gradual change. Then, based on the transition interval length and the stability threshold, determine the minimum required interval length for the gradual change. The minimum required length = |total amplitude| ÷ stability threshold. In this embodiment, the stability threshold = 5 L / min·m. If the total amplitude = 12 L / min, then the minimum required length = 12 ÷ 5 = 2.4 m. If the preset transition interval length is greater than or equal to the minimum required length, the flow rate will gradually change at a uniform speed according to the preset interval length. The flow rate change rate = total amplitude ÷ transition interval length, ensuring that the change rate is less than or equal to the stability threshold. The final determination logic for the length of the transition interval is as follows: After calculating the preset transition interval length L_preset and the minimum required length L_min, the larger of the two values ​​is taken as the final transition interval length L_final, i.e., L_final=max(L_preset,L_min).

[0058] The starting depth of the transition interval remains at the original preset value (interface depth - L_preset / 2). The ending depth is adjusted to the starting depth + L_final.

[0059] This rule ensures that regardless of the magnitude of the formation parameter differences, the rate of change in flow rate will never exceed the stability threshold, prioritizing the stability of the drilling process. Simultaneously, it also accommodates situations where the preset interval is already sufficiently long, eliminating the need for additional extensions and thus maintaining drilling efficiency.

[0060] In specific control, the control module adjusts the flow rate in 0.1m increments, calculates the target flow rate value for each increment, and calculates the current flow rate value as: current formation target flow rate value + (total amplitude ÷ transition interval length) × (current depth - transition interval start depth). This flow rate value is then sent to the flow control unit to achieve continuous and gradual flow rate changes.

[0061] For example, assuming the current formation target flow rate is 30 L / min, the next formation target flow rate is 42 L / min, the total range is 12 L / min, the adjusted transition interval length is 2.4 m (4.06-6.46 m), and the flow rate change rate is 12 ÷ 2.4 = 5 L / min·m, which meets the stability threshold requirement. When the drill bit reaches a depth of 4.16 m, the flow rate is 30 + 5 × (4.16 - 4.06) = 30.5 L / min; when it reaches a depth of 5.06 m, the flow rate is 30 + 5 × (5.06 - 4.06) = 35 L / min, until the interface is drilled through (5 m depth). Then, the flow rate continues to change gradually at this rate, precisely reaching 42 L / min at a depth of 6.46 m, completing a smooth transition.

[0062] The innovation of this application lies in avoiding the accumulation of sediment at the interface. The smooth and gradual change of flow rate keeps the flow field in the hole stable and will not cause eddies or dead zones due to sudden changes in flow rate. It effectively carries away sediment generated at the interface due to changes in slurry production capacity, reducing the risk of stuck drill bit. It can also protect the stability of the orifice wall because the gradual change in flow rate within the transition zone avoids sudden changes in pressure inside the orifice, reduces the impact of pressure fluctuations on the weak orifice wall at the interface, and thus improves the stability of the orifice wall. The length of the transition interval can be dynamically adjusted according to the actual difference in stratigraphic parameters. Even if there are slight deviations between the preliminary geological survey data and the actual situation on site, the transition mechanism can buffer these deviations, improve the robustness of the scheme, and effectively adapt to geological survey deviations.

[0063] It should be understood that in clear water drilling in silty clay formations, the initial slurry-making potential parameters of adjacent formations often differ, and the corresponding basic target flow rates will also be different. If the flow rate curve is only optimized through preliminary polygon modeling without designing a specific transition mechanism for the formation interface, problems may still occur when drilling crosses the interface.

[0064] Formation interfaces are regions of abrupt geological change, where adjacent formations exhibit significant differences in clay particle content, liquid and plastic limits, and other indicators. These abrupt changes in mud-making capacity can easily lead to sudden variations in the viscosity and fluidity of the drilling mud within the borehole. If the flow rate jumps directly from the target value of the current formation to the target value of the next formation, it will cause turbulence in the circulating flow field within the borehole, resulting in the accumulation of sediment at the interface and increasing the risk of stuck drill bit.

[0065] The bonding strength of the borehole wall at the interface is usually weak. Pressure fluctuations inside the borehole caused by sudden changes in flow rate may damage the stability of the borehole wall, especially in deep drilling or loose silty clay formations, which can easily lead to borehole wall collapse.

[0066] As one embodiment of the present invention, the method further includes: For at least one critical stratum node in the dynamic flow control sequence, at least one backup control branch is pre-generated; The steps for generating backup control branches include: calculating the probability of working conditions risk based on the first characteristic parameters of key formation nodes using the historical drilling database. For abnormal operating conditions where the probability exceeds a preset risk threshold, a corresponding corrective flow control strategy is preset.

[0067] Specifically, the selection criteria for key strata nodes must first be clarified. Key strata nodes refer to specific depths or strata locations with high failure risks and significant impacts on drilling safety during the drilling process. These include, but are not limited to, the starting and ending depths of each silty clay layer, the midpoint of the strata interface transition section, the abrupt change point of the first slurry potential parameter (the absolute value of the parameter difference is ≥1.0), and key depth points of deep silty clay layers with a burial depth exceeding 30m (one is set every 10m) to ensure coverage of all high-risk operation areas.

[0068] After key geological nodes are identified, operational risk probability calculations are performed based on the historical drilling database. The historical drilling database contains drilling data from the past 5 years in silty clay formations in the region and adjacent areas, covering formation characteristic parameters (clay particle content, liquid limit, plastic limit, first mud-making potential parameters, etc.), operational data (flow rate, drilling speed, borehole pressure, etc.), fault records (stuck drill bit, borehole collapse, abnormal mud viscosity, etc.), and handling solutions, with a total of no less than 100 sets of valid drilling cases.

[0069] Furthermore, the calculation of the probability of working condition risk adopts a statistical analysis method based on feature parameter matching. First, the first feature parameter of the target key stratum node is extracted (the core is the first slurry potential parameter, with clay particle content and liquid limit-plastic limit difference as auxiliary references). Then, cases with similarity ≥ 85% with the target parameter are screened in the historical drilling database. The similarity calculation method is: similarity = 1 - Σ|target parameter value - case parameter value| ÷ Σtarget parameter value, to ensure that the screened cases have reference value. Subsequently, the number of occurrences of various abnormal working conditions in the screened cases was counted. Abnormal working conditions include excessive mud viscosity (≥30mPa・s), abnormal increase in borehole pressure (≥1.2 times the normal pressure), excessive sediment accumulation thickness (≥5cm), and stuck drill failure. The risk probability of each abnormal working condition = the number of occurrences of that abnormal working condition ÷ the total number of screened cases.

[0070] The preset risk threshold is 30% after experimental verification and engineering practice. That is, when the risk probability of a certain abnormal working condition exceeds 30%, it is judged as a high-risk abnormal working condition that needs to be focused on prevention and control, and a corresponding modified flow control strategy is preset for it.

[0071] The revised flow control strategy formulates differentiated flow adjustment schemes based on the causes and impacts of different abnormal operating conditions, specifically as follows: For working conditions where mud viscosity exceeds the standard (risk probability > 30%), the core cause is that the mud-making capacity exceeds expectations, resulting in excessively rapid dispersion of clay particles. The corrected flow control strategy is to increase the flow rate by 15%-25% based on the original target flow rate. The specific increase ratio is determined according to the risk probability. For every 10% increase in risk probability, the increase ratio increases by 5%, but not exceeding 25%. At the same time, the flow rate change rate is kept below the stability threshold (5L / min·m). By increasing the flow rate, the clay particles are carried away more quickly, thus reducing the mud viscosity. For cases of abnormally high pressure inside the borehole (risk probability > 30%), it is mostly due to excessive flow leading to increased circulation resistance inside the borehole. The corrected flow control strategy is to reduce the flow rate by 10%-15% based on the original target flow rate and adopt a "step-down flow reduction" mode, reducing the total reduction by 20% for every 0.5m depth to avoid sudden flow drop causing sediment accumulation. For cases where sediment accumulation exceeds the standard (risk probability > 30%), the corrected flow control strategy is to adopt "pulse flow regulation", that is, based on the original target flow value, alternate between 1.2 times the flow rate for 1 minute and the original flow rate for 2 minutes every 3 minutes. The flow field disturbance is enhanced by the flow pulse to promote sediment suspension and carrying. The pulse period can be finely adjusted by the control module according to the actual sediment situation. For stuck drill situations (risk probability > 30%), the corrected flow control strategy is as follows: immediately adjust the flow rate to 1.3 times the original target flow rate, while simultaneously reducing the drilling speed by 50%. If the stuck drill risk is resolved after 3 minutes, gradually restore the original flow rate and drilling speed. If the risk is not resolved, continue to increase the flow rate to 1.5 times the original target value (not exceeding the equipment's maximum flow rate) and issue an alarm signal to prompt the operator to assist in handling the situation.

[0072] Each high-risk abnormal condition corresponds to an independent backup control branch. The backup control branch includes triggering conditions (critical formation node + abnormal condition judgment index, such as mud viscosity ≥30mPa・s), correction flow rate calculation rules, flow rate adjustment mode, and termination conditions (the original dynamic flow control sequence is restored after the abnormal condition is resolved). All backup control branches and the original dynamic flow control sequence are stored together in the control module of the operating system. The control module collects depth data, mud viscosity data, and borehole pressure data in real time during the drilling process. When the drill bit enters a critical formation node and meets the triggering conditions of a certain type of abnormal condition, it automatically switches to the corresponding backup control branch and executes the correction flow control strategy; when the abnormal condition is resolved (the index returns to normal within 3 consecutive adjustment steps), it automatically switches back to the original dynamic flow control sequence.

[0073] As a specific implementation method, if a key geological node is a deep silty clay layer with a burial depth of 35m, and its first mud-making potential parameter is 5.2, 30 cases with similarity ≥85% are selected from the historical drilling database. Among them, the mud viscosity exceeds the standard 12 times, and the risk probability = 12 ÷ 30 = 40% > 30%, so a modified flow control strategy needs to be preset.

[0074] The original target flow rate for this node was 50 L / min. Due to a risk probability exceeding the threshold of 40% (10% higher than the threshold), the corrected flow rate was 50 L / min × (1 + 15% + 5%) = 60 L / min. The flow rate change rate was (60 - 50) ÷ 1.0 m = 10 L / min·m (exceeding the stability threshold). Therefore, a gradual flow lifting mode was adopted, with a transition depth of 1.0 m. The flow rate gradually increased from 50 L / min to 60 L / min at a rate of 1 L / min every 0.1 m, ensuring the change rate met the requirements. When drilling reached a depth of 35 m, if the mud viscosity was detected to be 32 mPa·s, the control module automatically triggered the backup control branch, executing the gradual flow lifting strategy until the mud viscosity dropped below 30 mPa·s, at which point the original flow control was restored.

[0075] The innovation of this application lies in the fact that this step predicts high-risk working conditions in advance and presets backup control branches, avoiding the delayed response after abnormal working conditions occur in traditional operations, improving the continuity of drilling operations, and reducing downtime losses caused by failures.

[0076] The differentiated flow control strategy is precisely designed to address the causes of different abnormal operating conditions. This not only solves the problem that a single flow adjustment cannot adapt to multiple faults, but also avoids energy waste or secondary faults caused by blindly adjusting the flow.

[0077] As one embodiment of the present invention, in the step of calculating the probability of working conditions based on the historical drilling database, if the key stratum node has no match in the historical database or the matching data is insufficient, the following steps are performed: Based on the first feature parameters of key formation nodes, similarity retrieval is performed in the drilling knowledge graph, which constructs the correlation between formation features, drilling parameters and working condition risks. In this embodiment, the drilling knowledge graph is a pre-constructed structured relational database. The core construction logic is a ternary relation of "formation characteristics - drilling parameters - working condition risks". Formation characteristics include core indicators such as the first mud-making potential parameter, clay particle content, liquid limit, plastic limit, and water content. Drilling parameters include target flow rate, drilling speed, and circulating pressure. Working condition risks include the definition, characteristics, and impact range of various abnormal working conditions. The data sources of the knowledge graph cover nearly 10 years of drilling projects in silty clay formations across the country, industry technical standards, and academic research results, ensuring the comprehensiveness and authority of the relational relationships.

[0078] Specifically, similarity retrieval uses the first slurry potential parameter as the core retrieval dimension, and clay particle content and liquid limit-plastic limit difference as auxiliary retrieval dimensions. The retrieval algorithm uses the cosine similarity calculation method to calculate the similarity between the feature vector of the current key stratigraphic node and the feature vector of all historical stratigraphic nodes in the knowledge graph. The weight allocation of the feature vector is as follows: the first slurry potential parameter accounts for 60%, clay particle content accounts for 25%, and liquid limit-plastic limit difference accounts for 15%, ensuring that the core parameter plays a dominant role in the similarity results.

[0079] Obtain the top N historical stratigraphic nodes with the highest similarity to the current key stratigraphic node features, along with their associated abnormal working condition records. Specifically, the value of N is experimentally verified to be set to 50, ensuring a sufficient sample size to support probability inference while avoiding increased computational complexity due to excessive samples. If the total number of matched historical stratigraphic nodes in the knowledge graph is less than 50, all matched nodes are selected (at least 10 are guaranteed; if less than 10, supplemented with risk probability benchmarks recommended by industry standards).

[0080] Based on the frequency of abnormal operating conditions associated with N historical stratigraphic nodes, and combined with the feature similarity between the current node and each historical node, a weighted calculation is performed to deduce the probability of operating condition risk for the current key stratigraphic node.

[0081] In this embodiment, the specific calculation logic is as follows: First, the frequency of occurrence of various abnormal working conditions in each historical stratigraphic node is calculated. The frequency of occurrence = the number of times a certain type of abnormal working condition occurs in the drilling project corresponding to the historical node ÷ the drilling time (hours) of the project at that node, and then uniformly converted into a quantitative indicator of "times / hour". Next, calculate the weight coefficient for each historical node. The weight coefficient = the feature similarity between the historical node and the current node ÷ the sum of the feature similarities of all N historical nodes, ensuring that the sum of all weight coefficients is 1. Finally, the risk probability of a certain type of abnormal working condition at the current node is calculated using the following formula: Risk probability = Σ (frequency of this type of abnormal working condition at a certain historical node × weight coefficient of that historical node) × K, where K is the scenario correction coefficient, which is set according to the differences in the depth, ambient temperature, equipment model, etc. of the current drilling project, and the value range is 0.8-1.2. For deep drilling (depth > 50m), it is 1.2; for shallow drilling (depth < 20m), it is 0.8; and for conventional depth, it is 1.0, to ensure that the probability calculation fits the actual scenario.

[0082] After the risk probability simulation is completed, the subsequent process is consistent with the historical database for sufficient scenarios: for abnormal working conditions that exceed the preset risk threshold, a corresponding corrective flow control strategy is preset, an independent backup control branch is generated and stored in the control module, and it is automatically triggered or switched according to real-time data during the drilling process.

[0083] The innovation of this application lies in the fact that, based on similarity retrieval and weighted calculation of drilling knowledge graph, it breaks through the geographical and project limitations of a single historical database. Even in the case of first drilling or in scenarios with insufficient data, it can still achieve accurate prediction of working condition risks, filling the technical gap of traditional methods in scenarios with missing data.

[0084] The introduction of eigenvector weighted calculation and scenario correction coefficients enables risk probability calculation to take into account both the correlation of historical data and the individual differences of the current project, providing a more reliable basis for the design of backup control branches.

[0085] As one embodiment of the present invention, the method further includes: During drilling operations, a second real-time working condition parameter is acquired to characterize the current borehole wall condition and cuttings carrying status. The second real-time working condition parameter includes at least drill pipe lifting resistance, return water consistency, and return sand volume. Real-time acquisition of drill pipe lifting resistance is achieved through the following methods: A pressure sensor is installed on the drilling rig lifting system, and the peak hydraulic pressure at the initial moment of the uniform lifting action of the drill rod at the borehole opening when it is stationary is collected as a characterization value of the drill rod lifting resistance. In this embodiment, the sampling time is the initial moment when the drill rod at the borehole is in a stationary state and is being lifted at a constant speed. The peak hydraulic pressure collected at this time is directly used as the characterization value of the drill rod lifting resistance. The sampling frequency is consistent with the drilling sampling cycle, set to 1 time / minute. The measurement accuracy of the pressure sensor is ±0.1MPa to ensure that the characterization value can accurately reflect the change of frictional resistance between the drill rod and the borehole wall and mud cake. The consistency of the return water is collected by an online viscometer installed on the return water pipeline. The measurement range is 1-100 mPa·s, and the accuracy is ±1 mPa·s. The viscosity of the circulating return water is monitored in real time, which indirectly reflects the mud's ability to carry rock cuttings and the mud-making status in the borehole.

[0086] The amount of returned sand is collected by a weighing sand meter at the end of the return water pipe. The metering range is 0-5 kg / min and the accuracy is ±0.01 kg. At the same time, a particle size analyzer is used to obtain the particle size distribution data of the returned sand. The acquisition frequency is 1 time / minute to ensure a comprehensive understanding of the quantity and particle size characteristics of the rock cuttings carried.

[0087] The second real-time operating condition parameters are compared with the expected state based on the dynamic flow control sequence; In this embodiment, the specific comparison logic is as follows: First, the types of abnormal working conditions are clearly defined, including at least the excessive mud cake development condition and the borehole wall instability condition. The judgment conditions for both types of conditions are defined by quantitative indicators to ensure the accuracy of the judgment: The criteria for judging the excessive mud cake condition are that two indicators are met simultaneously: First, the drill pipe lifting resistance continues to increase, that is, within three consecutive sampling periods, the characteristic value of the drill pipe lifting resistance shows a gradual increasing trend, and the total increase is ≥10%; Second, the ratio of return water flow to return sand volume shows a monotonically increasing trend over M consecutive sampling periods, where M is set to 5 and the return water flow is the actual flow rate of the current clean water circulation. The increase in this ratio means that the return water demand corresponding to the unit return sand volume has increased, which indirectly reflects the decrease in return water carrying efficiency caused by the thickening of mud cake.

[0088] The judgment condition for borehole wall instability is that two indicators are met simultaneously: First, within a single sampling period, the increase in the real-time measured value of the return water consistency relative to the previous sampling period exceeds 20% of its normal operating condition baseline value, where the normal operating condition baseline value is the average value of the return water consistency of the previous 10 sampling periods within this depth range; Second, in the particle size distribution of the current return sand volume, the percentage of the volume of particles larger than the normal cutting rock cutting particle size exceeds K times its historical average level, where K>1. In this embodiment, K is set to 1.5. The normal cutting rock cutting particle size is the upper limit of the rock cutting particle size preset according to the first slurry potential parameter (e.g., when the first slurry potential parameter ≥4.0, the normal particle size upper limit is 5mm), and the historical average level is the statistical average value of the previous 20 sampling periods of this key stratum node.

[0089] When the comparison result triggers one of the abnormal operating conditions, the corresponding backup control branch is invoked to overwrite the pre-generated target flow value.

[0090] In this embodiment, the control module automatically calls the corresponding backup control branch, and uses the corrected flow control strategy in the branch to override the pre-generated target flow value, as follows: The control module has a built-in mapping table between abnormal operating conditions and backup control branches. When an excessive mud cake development condition is identified, the preset corrected flow control strategy for that condition is invoked; when a borehole wall instability condition is identified, the corresponding borehole wall instability correction strategy is invoked.

[0091] The revised flow control strategy has a higher execution priority than the original dynamic flow control sequence. During execution, the flow rate change rate is kept below the stability threshold (5 L / min·m) to avoid secondary impacts on the borehole caused by sudden flow changes. If, during the execution of the backup control branch, the second real-time operating parameters recover to the expected normal state within 5 consecutive sampling cycles, the system automatically switches back to the original dynamic flow control sequence. If the parameters fail to recover after more than 10 sampling cycles, the control module issues a high-level alarm signal, prompting operators to check for other hidden faults. By replacing manual experience-based judgment with quantitative parameter monitoring, subjective misjudgments or omissions are avoided. The accuracy of identifying abnormal operating conditions such as excessive mud cake development and borehole wall instability is improved, allowing more time for timely handling.

[0092] The closed-loop logic of parameter comparison and automatic switching enables flow control to follow preset strategies while dynamically adjusting according to real-time working conditions, improving the adaptability of the entire drilling system to complex geological changes and further ensuring borehole integrity and drilling operation continuity.

[0093] In one embodiment of the present invention, the step of calling the corresponding backup control branch further includes: Calculate a flow adjustment range based on the target flow value currently being executed and the preset correction flow value in the backup control branch; If the flow adjustment exceeds a preset threshold for sudden change, the flow adjustment process will be broken down into multiple sub-steps. The number of sub-steps and the flow adjustment amount for each step are calculated using a model predictive control algorithm. The objective function of the model predictive control algorithm is to minimize the sum of squares of the deviations between the predicted and set values ​​of the bottom hole pressure in the prediction time domain. Its constraints include the stability threshold and the maximum power of the circulating pump. The clean water circulation flow rate is controlled within multiple consecutive adjustment cycles, and gradually adjusted to the corrected flow rate value at a rate of change not exceeding the stability threshold.

[0094] In this embodiment, the flow adjustment range is calculated as the absolute value of the difference between the current target flow value and the corrected flow value, i.e., adjustment range = |corrected flow value - current target flow value|.

[0095] The preset mutation threshold was set to 10 L / min after experimental calibration and engineering practice verification. This threshold is a critical value determined by comprehensively considering the stability of the pressure inside the orifice and the load-bearing capacity of the equipment. If the calculated flow adjustment range exceeds 10 L / min, the flow adjustment process needs to be decomposed into multiple sub-steps. If the adjustment range is ≤10 L / min, the flow rate is adjusted to the corrected value in one go at a rate not exceeding the stability threshold (5 L / min·m) according to the original logic, taking into account both adjustment efficiency and stability.

[0096] The number of sub-steps and the flow rate adjustment for each step are calculated using a model predictive control algorithm. The core optimization objective function of the model predictive control algorithm is to minimize the sum of squares of the deviations between the predicted and setpoint values ​​of the bottom hole pressure within the prediction time domain. The prediction time domain is set to 10 adjustment cycles (each adjustment cycle = 1 minute, consistent with the sampling cycle). The bottom hole pressure setpoint is the reference value of the bottom hole pressure under normal operating conditions in this depth range (preset based on the first slurry potential parameter, for example, when the first slurry potential parameter = 4.5, the setpoint = 0.8 MPa). By minimizing the sum of squares of deviations, the bottom hole pressure is ensured to remain stable within a reasonable range during the adjustment process.

[0097] The algorithm's constraints include two core elements: First, the rate of change in flow rate must not exceed the stability threshold (5L / min·m), meaning that the flow rate adjustment in each adjustment cycle must not exceed 5L / min (since the drilling depth advances by approximately 1m in each adjustment cycle, the corresponding flow rate change is ≤5L / min·m); Second, the adjusted real-time flow rate must not exceed the maximum flow rate corresponding to the maximum power of the circulating pump in the clean water circulation system (in this embodiment, the maximum flow rate is set to 60L / min), to avoid exceeding the equipment's operating limits.

[0098] The core of the model predictive control algorithm lies in: based on a simplified dynamic model of orifice bottom pressure, calculating the optimal flow adjustment sequence for multiple future cycles online. Its specific execution steps are as follows: First, the model is initialized: the system has a built-in orifice bottom pressure response model, which describes how flow rate changes affect orifice bottom pressure. Model parameters can be obtained through preliminary short-time step tests.

[0099] Then, the algorithm enters a rolling optimization loop: at the beginning of each control cycle, the algorithm performs the following operations: obtains the current measured value of the orifice bottom pressure and the flow rate; Based on this, the trajectory of the orifice bottom pressure change under different flow adjustment schemes is predicted over the next Np cycles. Then, an optimization problem is solved to find a set of flow adjustment values ​​that minimizes the deviation between the predicted pressure trajectory and the set value, while satisfying the constraints that the flow adjustment value in each cycle does not exceed the stability threshold and the total flow after adjustment does not exceed the upper limit of the equipment. Finally, only the first adjustment value in the solution is used as the output command for this cycle.

[0100] In the next cycle, the prediction starting point is refreshed with new pressure measurements, and the above rolling optimization process is repeated to achieve closed-loop control.

[0101] When adjusting from 30 L / min to 45 L / min, the algorithm does not pre-determine a fixed three-step process. A typical adjustment process is as follows: In the first cycle, the measured pressure is 0.78 MPa, and after optimization calculation, the commanded flow rate increases by 4 L / min; in the second cycle, the measured pressure is 0.795 MPa, and the algorithm calculates the optimal adjustment amount to be 5 L / min; in the third cycle, the pressure rises to 0.81 MPa, and to avoid overshoot, the algorithm commands only an increase of 3 L / min. This process continues in subsequent cycles until the flow rate reaches the target value and the pressure stabilizes near the set value. Therefore, this algorithm ensures both the smoothness of flow rate changes and actively maintains the stability of the orifice bottom pressure, fundamentally avoiding the risk of orifice wall instability caused by pressure fluctuations.

[0102] If a new abnormal operating condition occurs during the adjustment, the system will immediately pause the current optimization, quickly refresh the internal prediction model by integrating the latest data, and restart the rolling optimization based on the new model to ensure that the control strategy always matches the real-time state inside the orifice.

[0103] The innovation of this application lies in optimizing the sub-step decomposition scheme through model predictive control algorithm, so that the pressure at the bottom of the orifice remains stable near the set value during the flow adjustment process, avoiding secondary impact on the orifice wall due to large pressure fluctuations, and further reducing the risk of orifice wall collapse under abnormal working conditions.

[0104] The method of sub-step decomposition and gradual adjustment not only ensures the timeliness of flow adjustment (the total adjustment time is controlled within a reasonable range and does not affect the efficiency of handling abnormal operating conditions), but also avoids instantaneous load impact on equipment and extends the service life of key equipment such as circulating pumps.

[0105] During the adjustment process, the adjustment scheme can be dynamically recalculated based on real-time operating conditions, making the flow adjustment adaptive. Even if new changes occur in the operating conditions inside the orifice, it can quickly respond and adjust, further enhancing the ability to handle complex and abnormal operating conditions and improving the engineering practicality of the technical solution.

[0106] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An adaptive adjustment method for hydrogeological drilling parameters, applied to a drilling system for clear water drilling in silty clay formations, characterized in that, The method includes: Before drilling begins, based on geological survey data, the first stratigraphic sequence of the drilling path is obtained. The first stratigraphic sequence contains multiple silty clay layers arranged in sequence and their corresponding first characteristic parameters. The first characteristic parameters include at least the first slurry-making potential parameter. Based on the first formation sequence and the first slurry-making potential parameters of each formation therein, a dynamic flow control sequence is pre-generated; the dynamic flow control sequence defines the target flow rate value that should be used for clear water circulation when drilling at different depths or in different formations; During drilling operations, the operating system is controlled to dynamically adjust the clean water circulation flow rate to the corresponding target flow rate value according to the dynamic flow control sequence.

2. The adaptive adjustment method for hydrogeological drilling parameters according to claim 1, characterized in that, The step of pre-generating a dynamic flow control sequence includes: For each stratum in the first stratigraphic sequence, a basic target flow rate value is assigned to the stratum based on the first slurry potential parameter of that stratum; Based on the chronological order of the first formation sequence, the basic target flow rate values ​​of adjacent formations are smoothed and optimized to generate a curve showing the change of the target flow rate value with drilling depth. This curve has the minimum integral value on the drilling depth axis and satisfies the condition that the flow rate change rate does not exceed a preset stability threshold, thereby generating the final target flow rate value.

3. The adaptive adjustment method for hydrogeological drilling parameters according to claim 2, characterized in that, The step of generating a curve showing the target flow rate as a function of drilling depth, with the integral value of this curve minimized on the drilling depth axis, is achieved in the following way: The flow rate curve for the entire drilling process is modeled as a polygon composed of flow rate values ​​from multiple formation nodes. The vertex coordinates of this polygon are optimized using an iterative algorithm to minimize the total area of ​​the polygon while satisfying the constraint that the absolute value of the slope of the line connecting all adjacent vertices is not greater than the stability threshold.

4. The adaptive adjustment method for hydrogeological drilling parameters according to claim 2, characterized in that, The step of optimizing the basic target flow values ​​of adjacent strata based on the sequential order of the first stratigraphic sequence further includes: Before drilling operations begin, a flow transition interval is pre-defined for each pair of adjacent formations; The length of the flow transition interval is set according to the absolute value of the difference between the first slurry potential parameters of two adjacent strata. The larger the difference, the longer the preset transition interval. During drilling, when the drill bit enters the flow transition zone, the clear water circulation flow rate is controlled to gradually change from the target flow rate value of the current formation to the target flow rate value of the next formation before drilling through the interface between the two formations, with a flow rate change rate not exceeding the stability threshold.

5. The adaptive adjustment method for hydrogeological drilling parameters according to claim 1, characterized in that, The method further includes: For at least one critical stratum node in the dynamic flow control sequence, at least one backup control branch is pre-generated; The step of generating the backup control branch includes: calculating the probability of working condition risk based on the first characteristic parameter of the key formation node using the historical drilling database; and setting a corresponding corrected flow control strategy for abnormal working conditions where the probability exceeds a preset risk threshold.

6. The adaptive adjustment method for hydrogeological drilling parameters according to claim 5, characterized in that, In the step of calculating the probability of working conditions based on the historical drilling database, if the key formation node has no match in the historical database or the matching data is insufficient, the following steps are performed: Based on the first feature parameters of the key formation nodes, a similarity search is performed in the drilling knowledge graph, which constructs the correlation between formation features, drilling parameters and working condition risks. Obtain the top N historical stratigraphic nodes with the highest similarity to the current key stratigraphic node features and their associated abnormal working condition records; Based on the frequency of abnormal operating conditions associated with the N historical stratigraphic nodes, and combined with the feature similarity between the current node and each historical node, a weighted calculation is performed to deduce the probability of operating condition risk of the current key stratigraphic node.

7. The adaptive adjustment method for hydrogeological drilling parameters according to claim 5, characterized in that, The method further includes: During drilling operations, a second real-time working condition parameter is acquired to characterize the current borehole wall condition and cuttings carrying status. The second real-time working condition parameter includes at least drill pipe lifting resistance, return water consistency, and return sand volume. The second real-time operating condition parameter is compared with the expected state based on the dynamic flow control sequence; When the comparison result triggers one of the abnormal operating conditions, the corresponding backup control branch is invoked to overwrite the pre-generated target flow value.

8. The adaptive adjustment method for hydrogeological drilling parameters according to claim 7, characterized in that, The real-time acquisition of drill pipe lifting resistance is achieved through the following methods: A pressure sensor is installed on the drilling rig lifting system, and the peak hydraulic pressure at the initial moment of the uniform lifting action of the drill rod at the borehole opening when it is stationary is collected as a characterization value of the lifting resistance of the drill rod.

9. The adaptive adjustment method for hydrogeological drilling parameters according to claim 7, characterized in that, The abnormal operating conditions include at least the conditions of excessive mud cake development and borehole wall instability; The criteria for judging the excessive mud cake condition are: the drill rod lifting resistance continues to increase, and at the same time, the ratio of return water flow to return sand volume shows a monotonically increasing trend within M consecutive sampling periods. The criteria for determining the instability of the borehole wall are as follows: within a single sampling period, the real-time measured value of the return water consistency increases by more than 20% of its normal operating condition baseline value relative to the previous sampling period, and the percentage of particles larger than the normal cutting rock chip size in the particle size distribution of the current return sand exceeds K times its historical average level, where K>1.

10. The adaptive adjustment method for hydrogeological drilling parameters according to claim 7, characterized in that, The step of invoking the corresponding backup control branch also includes: Based on the target flow rate currently being executed and the preset correction flow rate in the backup control branch, calculate a flow rate adjustment range; If the flow adjustment range exceeds a preset mutation threshold, the flow adjustment process will be decomposed into multiple sub-steps for execution. The number of sub-steps and the flow rate adjustment amount for each step are calculated using a model predictive control algorithm. The objective function of the model predictive control algorithm is to minimize the sum of squares of the deviations between the predicted and set values ​​of the bottom hole pressure in the prediction time domain. Its constraints include the stability threshold and the maximum power of the circulating pump. The clean water circulation flow rate is controlled to be gradually adjusted to the corrected flow rate value within multiple consecutive adjustment cycles at a flow rate change rate not exceeding the stability threshold.

Images