An intelligent control method for straightening process of high-pressure oil and gas well casing

Abstract

The present application belongs to the technical field of precision straightening of oil and gas well casing, and particularly relates to an intelligent control method for straightening process of high-pressure oil and gas well casing. Force sensors are installed on each loading ram; an incremental straightening force is applied to the convex side of the bent section of the casing, and a force-strain response curve is constructed in real time, and a curvature characteristic evolution sequence is constructed; according to the online form of the sequence, three deformation stages of elasticity, elastic-plastic transition and steady plastic flow are automatically identified, and the straightening force increment is asymmetrically adjusted in the elastic-plastic transition stage; at the same time, a plastic instability early warning index is constructed by the degree of cross-section ovalization, and when the index exceeds a preset safety threshold, the straightening force setting value is dynamically reduced and pressure is maintained during unloading; a comprehensive deviation index is calculated, and when the convergence condition is not met, the axial position of the next pass ram and the straightening force amplitude are updated. The dynamic attenuation coefficient is calculated according to the amplitude exceeding the threshold value, and the straightening force setting value is reduced, and after pressure maintenance, the casing is unloaded to avoid continuous loading when the casing approaches plastic cross-section instability.

Description

Technical Field

[0001] This invention belongs to the field of precision straightening processing technology for oil and gas well casing, and particularly relates to an intelligent control method for the straightening process of high-pressure oil and gas well casing. Background Technology

[0002] The straightness of the casing in high-pressure oil and gas wells directly affects wellbore integrity, downhole tool accessibility, and joint sealing reliability. Casing inevitably undergoes bending deformation during rolling, heat treatment, transportation, and storage, requiring straightening treatment to meet API standards. The 5CT standard requires that existing casing straightening mainly adopts a multi-point reverse bending pressure straightening process. The core of its straightening quality lies in the precise control of the straightening force and loading position. However, the existing control methods have significant shortcomings. The detection methods are outdated, relying on manual visual inspection or offline measurement, which can only obtain surface geometric information and cannot perceive the internal stress and strain state of the casing. The control model is coarse, relying on empirical formulas or fixed parameters calibrated offline, which cannot adapt to the differences in the actual yield strength and hardening characteristics of casing materials from different batches, resulting in frequent over-straightening and under-straightening. The control objective is singular, focusing only on straightness while ignoring residual stress and cross-sectional ellipticity, making it difficult to meet the comprehensive requirements of high-pressure oil and gas wells for casing crush resistance, fatigue life, and sealing performance. Furthermore, it lacks the ability to adaptively identify deformation stages and the automatic planning mechanism for multi-pass straightening paths. The pass parameters rely entirely on operational experience, resulting in low straightening efficiency and poor quality consistency. Therefore, an intelligent control method for casing straightening is needed that can perceive the bending state in real time, adaptively identify deformation stages, take into account multiple quality indicators, and automatically plan the straightening path. Summary of the Invention

[0003] This invention addresses the field of precision straightening and machining technology for oil and gas well casing, proposing a reasonable, simple, and theoretically sound intelligent control method for the straightening process of high-pressure oil and gas well casing.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] S1. Multiple detection sections are arranged axially on the outer surface of the sleeve to be straightened. Each detection section is uniformly arranged with strain gauges in the circumferential direction to form a circumferential strain detection array. Force sensors are installed on each loading head of the straightening equipment.

[0006] S2. Apply an increasing straightening force to the convex side of the bend section of the sleeve, and collect the force sensor signal and the strain gauge signal in real time. Use the strain value measured by the strain gauge on the convex side of the section with the most significant bend as the characteristic strain, and use the force value measured by the corresponding pressure head force sensor as the characteristic force. Construct a real-time response curve on the force-strain plane by sampling point.

[0007] S3. Perform online geometric curvature extraction on the real-time response curve. At each sampling point, trace back two historical sampling points with unequal sampling intervals from the response curve. Use the curvature of the circumcircle determined by these three points as the curvature feature value at the sampling point and arrange them according to the sampling time sequence to form a curvature feature evolution sequence.

[0008] S4. Based on the online morphology of the curvature feature evolution sequence, identify the sleeve deformation stage in real time and switch the control strategy. When the curvature feature value is lower than the detection dead zone threshold, it is determined to be the elastic stage. The straightening force is increased at a preset rate. When the curvature feature value exceeds the detection dead zone threshold, it is determined to enter the elastic-plastic transition stage. In this stage, if the curvature feature value is in an upward trend, the straightening force increment is rapidly reduced. If the curvature feature value is in a downward trend, the straightening force increment is slowly restored. When it is detected that the curvature feature value has exceeded the peak value and fallen back to below the convergence threshold, it is determined to enter the steady-state plastic flow stage. Switch to adjusting the straightening force at a preset target plastic strain rate until the feature strain reaches the target.

[0009] S5. During the entire process of straightening force loading from S2 to S4, the degree of cross-sectional ellipticity characterized by the difference in readings of each strain gauge in the circumferential direction within the same detection section is calculated in real time. Based on the degree of cross-sectional ellipticity and characteristic force, a plastic instability warning index is calculated. When the plastic instability warning index exceeds the preset safety threshold, it is determined that the sleeve is approaching plastic cross-sectional instability. The dynamic attenuation coefficient is calculated according to the threshold amplitude and the straightening force setting value is reduced. The reduced straightening force is used to maintain pressure and then unload. When the plastic instability warning index never exceeds the preset safety threshold, monitoring continues until the characteristic strain in S4 reaches the standard and then the process naturally proceeds to step S6.

[0010] S6. After unloading the single-pass straightening, a comprehensive deviation index is constructed from the maximum residual straightness deviation, maximum residual stress, and maximum residual ellipticity after unloading. If the comprehensive deviation index meets the convergence condition, the straightening is completed. If it does not meet the condition, the axial position of the pressure head in the next pass follows the shift towards the peak position of the residual curvature. The straightening force amplitude in the next pass increases with the comprehensive deviation index and the force increment gain coefficient according to the straightening force amplitude of the current pass. Then, return to step S2.

[0011] Preferably, the formula for calculating the curvature characteristic value in S3 is:

[0012] ,

[0013] ,

[0014] in, For the current sampling point The curvature characteristic value at that point, Historical sampling points At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... On the force-strain plane, the current sampling point Compared with historical sampling points The Euclidean distance between them On the force-strain plane, the current point Historical Points The Euclidean distance between them Historical sampling points on the force-strain plane Compared with historical sampling points The Euclidean distance between them Sampling points on the force-strain plane With sampling points The Euclidean distance between them For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force is measured by the pressure head force sensor.

[0015] Preferably, the increment of the upward posture corrective force in S4 is:

[0016] ,

[0017] in, For the current sampling point The increase in corrective force at the point, This represents the increase in straightening force at the beginning of the elastoplastic transition phase. The curvature compression coefficient controls the compressive strength of the curvature with respect to the increase in straightening force. For the current sampling point The curvature characteristic value at the point, and the incremental straightening force in the descent state are:

[0018] ,

[0019] in, This represents the increase in straightening force at the current sampling point k. This represents the increase in straightening force at the beginning of the elastoplastic transition phase. The curvature compression coefficient controls the compressive strength of the curvature with respect to the increase in straightening force. For the current sampling point The curvature characteristic value at that point, The peak recovery coefficient controls the recovery force of the straightening force increment after exceeding the peak curvature. This represents the historical maximum value of the local curvature characteristic value from the starting point of the elastoplastic transition stage to the current sampling point k.

[0020] Preferably, the formula for calculating the degree of ellipticity of the cross-section in S5 is:

[0021] ,

[0022] ,

[0023] in, The degree of ellipticity of the cross section at the current sampling point k. This represents the total number of strain gauges installed on the test section. This refers to the index of the strain gauge number within the same testing section. For the first in the same detection section The readings of each strain gauge at sampling point k. For the first The angular coordinates of each strain gauge in the circumferential direction of the cross section, and the calculation formula for the plastic instability early warning index are as follows:

[0024] ,

[0025] ,

[0026] in, For the current sampling point Plastic instability early warning indicators at the location, For the current sampling point The degree of ellipticity of the cross section at that location, Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points The degree of ellipticity of the cross section at that location, Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point Compared with historical sampling points The Euclidean distance on the elliptic-force plane For the current sampling point Compared with historical sampling points Euclidean distance on the elliptic-force plane Historical sampling points Compared with historical sampling points The Euclidean distance on the elliptic-force plane Sampling points Plastic instability early warning indicators at the location, Sampling points Plastic instability early warning indicators at the location, For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At the location corresponding to the characteristic force measured by the pressure head force sensor, the formula for calculating the dynamic attenuation coefficient is as follows:

[0027] ,

[0028] in, For the current sampling point The dynamic attenuation coefficient at that location, This represents the lower limit of the attenuation coefficient, i.e., the minimum value that the dynamic attenuation coefficient is allowed to take. The attenuation compression coefficient, For the current sampling point Plastic instability early warning indicators at the location, The preset safety threshold is the upper limit allowed for the elliptic-force curvature characteristic value.

[0029] Preferably, the formula for calculating the comprehensive deviation index in S6 is as follows:

[0030] ,

[0031] in, To form a comprehensive deviation index, This represents the maximum residual straightness deviation along the entire length of the casing after unloading. The target straightness tolerance is the maximum straightness deviation allowed by the process requirements. This represents the maximum residual stress on the casing surface after unloading. Allowable residual stress, which is the maximum permissible residual stress that does not affect the casing's resistance to crushing and fatigue life. This refers to the maximum residual ellipticity among the various test sections after unloading, i.e., the degree to which the straightened section deviates from a perfect circle. The allowable ellipticity rate, i.e., the maximum permissible degree of ellipticity of the cross-section that does not affect the sealing performance of the sleeve joint, is defined by the following formula for updating the axial position of the next pressure head:

[0032] ,

[0033] in, This is the position for the next axial loading of the pressure head. This indicates the current axial loading position of the pressure head. The axial coordinates represent the location of the peak residual curvature. To determine the position-following saturation coefficient, the update formula for the amplitude of the next straightening force is:

[0034] ,

[0035] in, The amplitude of the straightening force for the next pass. The straightening force amplitude per pass. For force increment gain coefficient, For force-increasing saturation coefficient, This is the current track deviation index.

[0036] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0037] 1. During the elastoplastic transition phase, the increase in straightening force is rapidly reduced by the curvature compression coefficient when the curvature characteristic value is rising, and slowly recovered by the peak recovery coefficient when the curvature characteristic value is falling, so that the increase in straightening force is kept to a minimum near the peak of the curvature characteristic value, thus achieving precise adjustment of the straightening force.

[0038] 2. This invention uses the degree of cross-sectional ellipticization and the plastic instability warning index to determine in real time whether the casing is approaching plastic cross-sectional instability. When the plastic instability warning index exceeds the preset safety threshold, the dynamic attenuation coefficient is calculated according to the threshold amplitude and the straightening force setting value is reduced. The reduced straightening force is used to maintain pressure and then unload, so as to avoid the casing from continuing to be loaded when it is approaching plastic cross-sectional instability.

[0039] 3. When the comprehensive deviation index does not meet the convergence condition, the axial position of the next pass indenter shifts to the position of the peak value of the residual curvature, and the amplitude of the next pass straightening force increases with the comprehensive deviation index and the force increment gain coefficient according to the current pass straightening force amplitude, thus realizing the automatic update of the axial position of the next pass indenter and the amplitude of the next pass straightening force. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a schematic diagram of an intelligent control method for the straightening process of high-pressure oil and gas well casing. Detailed Implementation

[0042] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described below in conjunction with the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0043] Numerous specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways than those described herein, and therefore the invention is not limited to the specific embodiments disclosed in the following specification.

[0044] Examples, such as Figure 1As shown, multiple detection sections are arranged axially on the outer surface of the sleeve to be straightened. Strain gauges are uniformly arranged circumferentially on each detection section, forming a circumferential strain detection array. Force sensors are installed on each loading head of the straightening equipment. An increasing straightening force is applied to the convex side of the bent section of the sleeve. The force sensor signals and the signals of each strain gauge are collected synchronously in real time. The strain value measured by the strain gauge on the convex side of the section with the most significant bend is used as the characteristic strain, and the force value measured by the force sensor on the corresponding pressure head is used as the characteristic force. A real-time response curve is constructed on the force-strain plane for each sampling point. Online geometric curvature extraction is performed on the real-time response curve: at each sampling point, the geometric curvature is extracted from the response curve backward. By tracing back two historical sampling points with unequal sampling intervals, the curvature of the circumscribed circle determined by these three points is used as the curvature characteristic value at that sampling point. These curvature characteristic evolution sequences are formed by arranging them according to the sampling time sequence. Multiple detection sections are axially arranged on the outer surface of the casing, with strain gauges uniformly distributed circumferentially on each section. A force sensor is installed on the loading head to simultaneously sense the casing's bending deformation field and the applied straightening force, achieving real-time dual-channel sensing of force and deformation. This addresses the shortcomings of traditional methods that can only acquire surface geometric information and cannot achieve online closed-loop control. The strain value of the strain gauge on the convex side of the section with the most significant bending is used as the characteristic strain, and the corresponding force of the loading head is used as the characteristic force. This process is then applied in force-strain... Constructing a real-time response curve on the variable plane is intended to abstract the complex bending of the entire casing into a single-degree-of-freedom force-strain relationship, transforming straightening control into an analysis problem of a single curve. This eliminates the need for prior knowledge of material parameters such as yield strength and hardening modulus, naturally adapting to the performance differences of casings from different furnace batches. Online geometric curvature extraction is performed on this response curve. At each sampling point, two historical sampling points with unequal sampling intervals are traced back, and the curvature of the three circumscribed circles is used as the curvature characteristic value of that sampling point, arranged chronologically into an evolution sequence. This is because the force-strain curve is nearly straight in the elastic segment and exhibits significant bending in the elastoplastic transition segment. In the steady-state plastic segment, the curvature tends to be linear again. The degree of curvature of the curve can uniquely identify the deformation stage. The curvature of the three-point circumcircle is a purely geometric quantity, which has a natural smoothing effect on measurement noise and does not require numerical differentiation. The unequal interval of the three-point configuration ensures the stable calculation of the circumcircle, thus obtaining the curvature characteristic evolution sequence. Its low-value stability, rise, peak, fall, and low-value stability directly correspond to the physical process of elasticity, elastoplastic transition, and steady-state plasticity. This solves the core problem of online real-time determination of the sleeve deformation stage without relying on material parameters and human experience, and provides an objective basis for subsequent switching control strategies and precise adjustment of straightening force. The formula for calculating the curvature characteristic value is:

[0045] ,

[0046] ,

[0047] in, For the current sampling point The curvature characteristic value at that point, Historical sampling points At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... On the force-strain plane, the current sampling point Compared with historical sampling points The Euclidean distance between them On the force-strain plane, the current point Historical Points The Euclidean distance between them Historical sampling points on the force-strain plane Compared with historical sampling points The Euclidean distance between them Sampling points on the force-strain plane With sampling points The Euclidean distance between them For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At the corresponding point, the characteristic force is measured by the indenter force sensor. Using the characteristic strain measured by the strain gauge on the convex side of the most significantly bent section and the characteristic force measured by the corresponding indenter force sensor as coordinate elements on the force-strain plane, this physically bypasses the dependence on prior parameters such as material yield strength and hardening modulus, allowing curvature calculation to be entirely based on real-time acquired force and strain data. A three-point configuration is formed using the current sampling point and two historical sampling points with unequal sampling intervals. The second-order geometric feature of the extracted curve, i.e., curvature, is compiled using the minimum information that three points can carry. Two points can only determine a straight line and cannot reflect bending; only three points can uniquely determine the circumcircle. Therefore, the three-point configuration is the minimum information for curvature extraction. A scheme with fewer than three points is mathematically invalid, while more than three points introduce fitting assumptions and computational delays. The cross product, twice the area of ​​the triangle enclosed by the three points, is used as the numerator to linearly quantify the degree of collinearity deviation of the three points. When the three points are collinear, this value is zero; for the straight section of the stress-strain curve, the more non-collinear the three points, the larger the value; for the curved section of the curve, only multiplication and subtraction are used. This method measures the degree of collinearity without introducing transcendental functions. It uses the product of the Euclidean distances of the three sides as the denominator and leverages the geometric relationship that the radius of the circumcircle equals the product of the three sides divided by four times the area of ​​the triangle to achieve scale normalization. This ensures that the curvature characteristic value depends only on the size of the circumcircle determined by the three points, regardless of their proximity. This avoids the situation where three points far apart on a straight section and three points close together on a curved section have the same curvature characteristic value due to their similar areas. The method uses the square root of the sum of the squares of the strain difference and the squares of the force difference to determine... The Euclidean distance, satisfying translational and rotational invariance in a dimensional plane with strain on the horizontal axis and force on the vertical axis, ensures that the measurement of the geometric interval between two points does not drift with the selection of the coordinate origin and the conversion of coordinate axis units, thus keeping the curvature characteristic value stable under any unit system. The unequally spaced three-point configuration ensures that the triangle does not degenerate, avoiding the problem of the circumcircle radius approaching infinity and the instability of curvature calculation when three equally spaced points approach collinearity in the straight section of the curve, thus ensuring stable low values ​​rather than singular values ​​in the elastic stage. Each element undertakes the functions of coordinate foundation, information compilation, collinearity measurement, scale normalization, and geometric invariance, respectively. Through recursive calculation point by point, the curvature characteristic value of each sampling point is output at each sampling time and arranged into a curvature characteristic evolution sequence. This solves the problems of traditional tangent modulus differential calculation being sensitive to measurement noise, requiring prior material parameters, and being unable to stably quantify the curvature of the force-strain curve online.

[0048] Based on the online morphology of the curvature feature evolution sequence, the sleeve deformation stage is identified in real time, and the control strategy is switched accordingly. When the curvature feature value is lower than the detection dead zone threshold, it is determined to be in the elastic stage, and the straightening force is increased at a preset rate. After the curvature feature value exceeds the detection dead zone threshold, it is determined to enter the elastoplastic transition stage. In this stage, if the curvature feature value is rising, the straightening force increment is rapidly reduced; if the curvature feature value is falling, the straightening force increment is slowly restored. When it is detected that the curvature feature value has exceeded the peak value and fallen back below the convergence threshold, it is determined to enter the steady-state plastic flow stage, and the straightening force is switched to be adjusted at a preset target plastic strain rate until the feature strain reaches the target. During the bending and straightening process of the sleeve... The process sequentially progresses through three stages: elasticity, elastoplastic transition, and steady-state plastic flow. In the elastic stage, stress and strain exhibit a linear relationship, and deformation fully recovers after unloading. During the elastoplastic transition, stiffness rapidly decreases from elastic to plastic values, and even a small increase in force triggers a significant increase in strain. In the steady-state plastic stage, stiffness remains stable at a low value, and plastic deformation accumulates steadily. These three stages have completely different requirements for loading methods and must be controlled separately. The force-strain curve is nearly linear in the elastic segment, bends in the elastoplastic transition segment, and tends to be linear again in the steady-state plastic segment. The evolution of the curvature characteristic value corresponds one-to-one with these three stages, providing the most direct and objective geometric criterion for stage identification. Therefore, a curvature characteristic value below the detection dead zone threshold is used to determine the elasticity of the flow. The elastic phase involves a preset rate-increasing straightening force. The dead zone threshold is used to shield the small fluctuations in the curvature characteristic value near zero in the elastic phase due to noise, preventing false triggering of phase switching. The transition to the elastoplastic transition phase is determined when the curvature characteristic value exceeds the dead zone threshold. During the upward phase, the material is approaching its yield point, and its stiffness is rapidly decreasing. If the straightening force increment is not quickly reduced, the same force step size will cause increasingly larger strain increments, exceeding the yield point and causing overshoot. During the downward phase, the material has already passed the yield point, and the stiffness decay is becoming more gradual. If a small step size is maintained, the loading efficiency will be too low. Therefore, the upward phase is rapidly reduced, and the downward phase is slowly restored. This asymmetry stems from the material's tangential stiffness near the yield point. The irreversible physical property of a significantly faster attenuation rate than the recovery rate after peak; when the curvature characteristic value exceeds the peak value and falls back below the convergence threshold, it is determined to enter the steady-state plastic flow stage. The straightening force is adjusted according to the preset target plastic strain rate to ensure uniform accumulation of plastic deformation and precise control of the total deformation, avoiding excessively rapid plastic deformation that could exacerbate cross-sectional ellipticization or excessive residual stress, until the characteristic strain reaches the target. This achieves objective and automatic identification of deformation stages and precise matching of control strategies for each stage with the mechanical properties of the material, solving the problems of stage determination relying on fixed strain thresholds or operational experience, and mismatch between loading strategies and actual material states in traditional processes, fundamentally suppressing overshoot and undercorrection near the yield critical point. The increment of the upward posture straightening force is:

[0049] ,

[0050] in, For the current sampling point The increase in corrective force at the point, This represents the increase in straightening force at the beginning of the elastoplastic transition phase. The curvature compression coefficient controls the compressive strength of the curvature with respect to the increase in straightening force. For the current sampling point The curvature characteristic value at the point, and the incremental straightening force in the descent state are:

[0051] ,

[0052] in, This represents the increase in straightening force at the current sampling point k. This represents the increase in straightening force at the beginning of the elastoplastic transition phase. The curvature compression coefficient controls the compressive strength of the curvature with respect to the increase in straightening force. For the current sampling point The curvature characteristic value at that point, The peak recovery coefficient controls the recovery force of the straightening force increment after exceeding the peak curvature. This represents the historical maximum value of the local curvature characteristic from the starting point of the elastoplastic transition stage to the current sampling point k. The product structure of the compressibility factor and the restoration factor realizes the transformation from continuous curvature characteristics to the adjustment of the straightening force increment. The adjustment direction is jointly determined by the physical mechanism of the material tangential stiffness near the yield point and the product coupling structure of the two factors: in the upward posture, because the current curvature characteristic value is constantly refreshed to the historical maximum value, the restoration factor is always 1, and only the compressibility factor participates in the action, and the straightening force increment decreases rapidly with the increase of curvature; in the downward posture, because the current curvature characteristic value is lower than the historical maximum value, the restoration factor is gradually activated, partially offsetting the suppression of the compressibility factor, and the straightening force increment recovers slowly. Therefore, the upward posture corresponds to rapid reduction, and the downward posture corresponds to slow recovery. The compression factor is expressed as the reciprocal of the product of the curvature compression coefficient and the current curvature characteristic value. When the current curvature characteristic value is close to zero, the factor takes a value close to 1 and the step size is not compressed, so that the loading rhythm at the stage switching point is smoothly connected with the elastic stage. When the current curvature characteristic value increases, the denominator increases smoothly and the step size decreases smoothly without abrupt changes. When the current curvature characteristic value is very large, the step size approaches zero but is always positive, which naturally has a lower limit protection and overcomes the problems of simple inverse proportion being unbounded near zero and linear decreasing having no lower limit. The recovery factor is expressed as the product of the peak recovery coefficient and the difference between the current curvature characteristic value and the historical maximum value. The historical maximum value only increases and never decreases, recording the highest position ever reached by the curvature. This provides a reference benchmark for determining whether the current curvature is in an upward or downward phase, eliminating the need for explicit upward or downward judgment in the control logic. The ratio of the current curvature characteristic value to the historical maximum value measures the position of the current curvature relative to the historical peak. It is always 1 in the upward phase and less than 1 after surpassing the peak, continuously decreasing. The recovery factor is automatically activated accordingly, its changes naturally lagging behind the fluctuations of the curvature itself, eliminating the need for post-peak determination. The curvature compression coefficient independently determines the compression force of the step size at the same curvature level, while the peak recovery coefficient independently determines the recovery speed of the step size in the downward phase. Their separation allows for separate calibration of compression strength and recovery force, without mutual constraint, adapting to differentiated settings for different steel grades of casing. A single expression continuously covers the entire process of rising and falling, and the adjustment logic is simple and transparent. It is completely consistent with the physical causal chain of the deformation stage of the aforementioned curvature feature evolution sequence identification. It provides clear and unambiguous step size adjustment instructions for the execution of straightening force, and solves the problems of needing explicit switching judgment, step size jump at the switching point, and manual splicing of two independent exponents for asymmetric dependence of segmented adjustment functions.Throughout the straightening force loading process from S2 to S4, the degree of cross-sectional ellipticity, characterized by the difference in readings of strain gauges in the circumferential direction within the same detection section, is calculated in real time. Based on the degree of cross-sectional ellipticity and characteristic force, a plastic instability early warning index is constructed. When the plastic instability early warning index exceeds the preset safety threshold, it is determined that the sleeve is approaching plastic cross-sectional instability. The dynamic attenuation coefficient is calculated according to the threshold amplitude, and the straightening force setting value is reduced. The reduced straightening force is used to maintain pressure before unloading. When the plastic instability early warning index never exceeds the preset safety threshold, monitoring continues until the characteristic strain in S4 reaches the standard, and then the process naturally proceeds to step S6. The formula for calculating the degree of cross-sectional ellipticity is:

[0053] ,

[0054] ,

[0055] in, The degree of ellipticity of the cross section at the current sampling point k. This represents the total number of strain gauges installed on the test section. This refers to the index of the strain gauge number within the same testing section. For the first in the same detection section The readings of each strain gauge at sampling point k. For the first The angular coordinates of each strain gauge in the circumferential direction of the cross section, and the calculation formula for the plastic instability early warning index are as follows:

[0056] ,

[0057] ,

[0058] in, For the current sampling point Plastic instability early warning indicators at the location, For the current sampling point The degree of ellipticity of the cross section at that location, Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points The degree of ellipticity of the cross section at that location, Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point Compared with historical sampling points The Euclidean distance on the elliptic-force plane For the current sampling point Compared with historical sampling points Euclidean distance on the elliptic-force plane Historical sampling points Compared with historical sampling points The Euclidean distance on the elliptic-force plane Sampling points Plastic instability early warning indicators at the location, Sampling points Plastic instability early warning indicators at the location, For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At the location corresponding to the characteristic force measured by the pressure head force sensor, the formula for calculating the dynamic attenuation coefficient is as follows:

[0059] ,

[0060] in, For the current sampling point The dynamic attenuation coefficient at that location, This represents the lower limit of the attenuation coefficient, i.e., the minimum value that the dynamic attenuation coefficient is allowed to take. The attenuation compression coefficient, For the current sampling point Plastic instability early warning indicators at the location, A preset safety threshold is set, which is the upper limit allowed for the ellipticity-force curvature characteristic value. In the formula for the degree of ellipticity of the cross section, the reading of the j-th strain gauge at sampling point k within the same detection cross section provides the original observation of the cross section deformation state. Its circumferential angular coordinates indicate the orientation of each strain gauge on the circumference of the cross section. The double angle cosine and sine are used as the second-order Fourier projection coefficients because the geometric essence of ellipticity is a second harmonic with a 180-degree period, producing two protrusions and two depressions within a 360-degree circumferential range. A single angle corresponds to pure bending, while a double angle corresponds to ellipticity. All strain gauge readings are multiplied by the double angle cosine and sine respectively. Summing the second-order Fourier cosine and sine components of the circumferential strain distribution yields both zero in pure bending, only producing significant non-zero values ​​when the cross-section changes from a circle to an ellipse. Taking the square root of the sum of their squares gives the second-order Fourier amplitude, representing the intensity of the pure elliptic component and unaffected by the orientation angle of the ellipse's major axis. Dividing this by the sum of the absolute values ​​of all strain gauge readings—the zeroth-order component representing the overall strain level—gives the degree of ellipticity of the cross-section. Physically, this represents the proportion of ellipticity in the total strain; it is zero when the cross-section is a perfect circle, and increases with increasing ellipticity. The higher the severity value, the more naturally it filters out first-order bending components and higher-order harmonic interference. In the formula for the early warning index of plastic instability, the degree of ellipticity of the cross section at the current sampling point is used as the vertical axis, and the characteristic force measured by the corresponding indenter force sensor is used as the horizontal axis. Together with the degree of ellipticity of the cross section and the characteristic force at two historical sampling points, they form a three-point configuration on the elliptic-force plane. Twice the area of ​​the triangle enclosed by the three points quantifies the degree of deviation of the three points from the collinear state. The Euclidean distance between any two of the three points is defined by the square root of the sum of the square of the difference in the degree of ellipticity of the cross section and the square of the difference in the characteristic force. The difference between two cross-dimensional directions is integrated into a single geometric interval. The product of the Euclidean distances of the three sides is used as the normalized denominator so that the warning index depends only on the size of the circumcircle determined by the three points and is independent of the absolute distance between the three points. The curvature of the circumcircle of the three points is the plastic instability warning index. When the ellipticization increases slowly with the force, the curve is close to a straight line and the value is close to zero. When the casing approaches the plastic section instability, the bending stiffness of the section drops sharply. The same force increment causes a larger ellipticization increment, the curve rises sharply, and the value shows a step jump, thus providing an early warning of section instability.In the dynamic attenuation coefficient formula, the preset safety threshold defines the upper limit of the allowable elliptic-force curvature characteristic value. The original over-threshold amplitude is obtained by subtracting the preset safety threshold from the absolute value of the plastic instability warning index. It is negative when not exceeding the limit and positive after exceeding the limit. The larger of the original over-threshold amplitude and zero is taken as the effective over-threshold amplitude. The negative value when not exceeding the limit is forcibly truncated to zero to ensure that attenuation is not triggered when not exceeding the limit. The attenuation compression coefficient controls the strength of the mapping from the effective over-threshold amplitude to the attenuation ratio. One plus the reciprocal of the product of the attenuation compression coefficient and the effective over-threshold amplitude constitutes the compression factor. When not exceeding the limit, it takes a value of 1 and there is no attenuation. After exceeding the limit, it decreases smoothly as the over-threshold amplitude increases. The deeper the over-threshold amplitude, the more severe the attenuation, and it is always positive. The lower limit of the attenuation coefficient is... The minimum allowable value of the dynamic attenuation coefficient is specified to ensure that the straightening force at most decreases to this lower limit rather than zero under the most severe instability risk, maintaining controllable clamping and pressure holding capabilities for the casing. The larger of the lower limit of the attenuation coefficient and the compressibility factor is taken as the dynamic attenuation coefficient, which is applied to the current straightening force setting value. The attenuated straightening force setting value is equal to the dynamic attenuation coefficient multiplied by the original setting value. Thus, by using second-order Fourier decomposition to extract pure elliptic components, using the curvature of the three-point circumscribed circle to measure the geometrical abrupt change caused by the attenuation of cross-sectional stiffness, and using a structure with compressibility factor and double truncation, a protection logic of no attenuation if not exceeding the limit, smooth attenuation if exceeding the limit, and attenuation with a lower limit is achieved, providing a complete chain of advanced early warning and graded protection for plastic cross-section instability.

[0061] After unloading during a single pass of straightening, a comprehensive deviation index is constructed from the maximum residual straightness deviation, maximum residual stress, and maximum residual ellipticity. If the comprehensive deviation index meets the convergence condition, straightening is complete; otherwise, the axial position of the indenter in the next pass shifts towards the peak position of the residual curvature, and the straightening force amplitude in the next pass increases with the comprehensive deviation index and the force increment gain coefficient according to the current pass's straightening force amplitude. The process then returns to step S2. The formula for calculating the comprehensive deviation index is:

[0062] ,

[0063] in, To form a comprehensive deviation index, This represents the maximum residual straightness deviation along the entire length of the casing after unloading. The target straightness tolerance is the maximum straightness deviation allowed by the process requirements. This represents the maximum residual stress on the casing surface after unloading. Allowable residual stress, which is the maximum permissible residual stress that does not affect the casing's resistance to crushing and fatigue life. This refers to the maximum residual ellipticity among the various test sections after unloading, i.e., the degree to which the straightened section deviates from a perfect circle. The allowable ellipticity rate, i.e., the maximum permissible degree of ellipticity of the cross-section that does not affect the sealing performance of the sleeve joint, is defined by the following formula for updating the axial position of the next pressure head:

[0064] ,

[0065] in, This is the position for the next axial loading of the pressure head. This indicates the current axial loading position of the pressure head. The axial coordinates represent the location of the peak residual curvature. To determine the position-following saturation coefficient, the update formula for the amplitude of the next straightening force is:

[0066] ,

[0067] in, The amplitude of the straightening force for the next pass. The straightening force amplitude per pass. For force increment gain coefficient, For force-increasing saturation coefficient, This is the comprehensive deviation index for the current test run. In the formula for the comprehensive deviation index, the ratio of the maximum residual straightness deviation of the entire casing length after unloading to the target straightness tolerance, the ratio of the maximum residual stress on the casing surface after unloading to the allowable residual stress, and the ratio of the maximum residual ellipticity in each test section after unloading to the allowable ellipticity constitute three excess ratios. The comprehensive deviation index is obtained by dividing the sum of the fourth powers of these three excess ratios by the sum of the squares of the three excess ratios. Its physical meaning is the degree to which excess energy concentrates on the worst-case indicator. When all three exceed the limit equally, it is approximately equal to the square of the average excess. When a single indicator severely exceeds the limit, it is dominated by that indicator and rapidly amplified, realizing the shift from the worst-case indicator... Automatic quality assessment, no manual weight setting required; in the formula for updating the axial position of the pressure head in the next pass, the position deviation is obtained by subtracting the axial loading position of the current pressure head from the axial coordinate of the peak position of the residual curvature. The sign indicates the direction of the deviation. The absolute value of the position deviation is multiplied by the position following saturation coefficient, then divided by one and the product is added to obtain the saturation factor. When the deviation is large, the saturation factor approaches 1, and the movement is approximately equal to the total deviation, achieving basic alignment in one step. When the deviation is small, the saturation factor is approximately equal to the product of the position following saturation coefficient and the deviation, and the movement is approximately equal to the product of the position following saturation coefficient and the square of the deviation. Fine approximation is achieved by multiplying the deviation by a saturation factor to obtain the actual movement. The axial loading position of the pressure head in the current pass is then added to the actual movement to obtain the axial loading position of the pressure head in the next pass. This method combines the characteristics of rapid approximation of large deviations with fine adjustment of small deviations. In the formula for updating the straightening force amplitude in the next pass, the force increment saturation factor is obtained by multiplying the current pass's comprehensive deviation index by the force increment saturation coefficient, dividing by one, and adding this product. When the comprehensive deviation index is large, the saturation factor approaches 1, capping the force increment to prevent over-correction caused by excessive force application. When the comprehensive deviation index is small, the saturation factor is approximately equal to the product of the force increment saturation coefficient and the comprehensive deviation index, and the force increment line... The force increment is gradually increased by multiplying the force increment gain coefficient by the force increment saturation factor to obtain the relative force increment. The force increment gain coefficient specifies the maximum growth ratio of the straightening force amplitude relative to the current amplitude when the force increment saturation factor reaches its upper limit, and limits the absolute upper limit of the force increment. The straightening force amplitude of the next pass is obtained by multiplying the current pass straightening force amplitude by an additional force relative increment. The straightening force remains unchanged when the comprehensive deviation index is zero, and increases when the comprehensive deviation index increases. These three factors together constitute a complete logic for adaptive planning of pass parameters, which is automatically dominated by the worst index for evaluation, quickly aligns large deviations, finely approximates small deviations, and gradually increases the force increment with a cap.

[0068] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments for application in other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A smart control method for the straightening process of high-pressure oil and gas well casing, characterized in that, Includes the following steps: S1. Multiple detection sections are arranged axially on the outer surface of the sleeve to be straightened. Each detection section is uniformly arranged with strain gauges in the circumferential direction to form a circumferential strain detection array. Force sensors are installed on each loading head of the straightening equipment. S2. Apply an increasing straightening force to the convex side of the bend section of the sleeve, and collect the force sensor signal and the strain gauge signal in real time. The strain value measured by the strain gauge on the convex side of the section with the most significant bend is the characteristic strain, and the force value measured by the corresponding pressure head force sensor is the characteristic force. Construct a real-time response curve on the force-strain plane by sampling point. S3. Perform online geometric curvature extraction on the real-time response curve. At each sampling point, trace back two historical sampling points with unequal sampling intervals from the response curve. Use the curvature of the circumcircle determined by these three points as the curvature feature value at the sampling point and arrange them according to the sampling time sequence to form a curvature feature evolution sequence. S4. Based on the online morphology of the curvature feature evolution sequence, identify the sleeve deformation stage in real time and switch the control strategy. When the curvature feature value is lower than the detection dead zone threshold, it is determined to be the elastic stage. The straightening force is increased at a preset rate. When the curvature feature value exceeds the detection dead zone threshold, it is determined to enter the elastic-plastic transition stage. In this stage, if the curvature feature value is in an upward trend, the straightening force increment is rapidly reduced. If the curvature feature value is in a downward trend, the straightening force increment is slowly restored. When it is detected that the curvature feature value has exceeded the peak value and fallen back to below the convergence threshold, it is determined to enter the steady-state plastic flow stage. Switch to adjusting the straightening force at a preset target plastic strain rate until the feature strain reaches the target. S5. During the entire process of straightening force loading from S2 to S4, the degree of cross-sectional ellipticity characterized by the difference in readings of each strain gauge in the circumferential direction within the same detection section is calculated in real time. Based on the degree of cross-sectional ellipticity and characteristic force, a plastic instability warning index is calculated. When the plastic instability warning index exceeds the preset safety threshold, it is determined that the sleeve is approaching plastic cross-sectional instability. The dynamic attenuation coefficient is calculated according to the threshold amplitude and the straightening force setting value is reduced. The reduced straightening force is used to maintain pressure and then unload. When the plastic instability warning index never exceeds the preset safety threshold, monitoring continues until the characteristic strain in S4 reaches the standard and then the process naturally proceeds to step S6. S6. After unloading the single-pass straightening, a comprehensive deviation index is constructed from the maximum residual straightness deviation, maximum residual stress, and maximum residual ellipticity after unloading. If the comprehensive deviation index meets the convergence condition, the straightening is completed. If it does not meet the condition, the axial position of the pressure head in the next pass follows the shift towards the peak position of the residual curvature. The straightening force amplitude in the next pass increases with the comprehensive deviation index and the force increment gain coefficient according to the straightening force amplitude of the current pass. Then, return to step S2.

2. The intelligent control method for straightening high-pressure oil and gas well casing according to claim 1, characterized in that, The formula for calculating the curvature characteristic value in S3 is as follows: ， ， in, For the current sampling point The curvature characteristic value at that point, Historical sampling points At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... On the force-strain plane, the current sampling point Compared with historical sampling points The Euclidean distance between them On the force-strain plane, the current point Historical Points The Euclidean distance between them Historical sampling points on the force-strain plane Compared with historical sampling points The Euclidean distance between them Sampling points on the force-strain plane With sampling points The Euclidean distance between them For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At the point where bending is most pronounced, the characteristic strain measured by strain gauges on the convex side of the cross-section is... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force is measured by the pressure head force sensor.

3. The intelligent control method for straightening high-pressure oil and gas well casing according to claim 1, characterized in that, The increment of the ascending posture corrective force in S4 is: ， in, For the current sampling point The increase in corrective force at the point, This represents the increase in straightening force at the beginning of the elastoplastic transition phase. The curvature compression coefficient controls the compressive strength of the curvature with respect to the increase in straightening force. For the current sampling point The curvature characteristic value at the point, and the incremental straightening force in the descent state are: ， in, This represents the increase in straightening force at the current sampling point k. This represents the increase in straightening force at the beginning of the elastoplastic transition phase. The curvature compression coefficient controls the compressive strength of the curvature with respect to the increase in straightening force. For the current sampling point The curvature characteristic value at that point, The peak recovery coefficient controls the recovery force of the straightening force increment after exceeding the peak curvature. This represents the historical maximum value of the local curvature characteristic value from the starting point of the elastoplastic transition stage to the current sampling point k.

4. The intelligent control method for straightening high-pressure oil and gas well casing according to claim 1, characterized in that, The formula for calculating the degree of ellipticity of the cross section in S5 is as follows: ， ， in, The degree of ellipticity of the cross section at the current sampling point k. This represents the total number of strain gauges installed on the test section. This refers to the index of the strain gauge number within the same testing section. For the first in the same detection section The readings of each strain gauge at sampling point k. For the first The angular coordinates of each strain gauge in the circumferential direction of the cross section, and the calculation formula for the plastic instability early warning index are as follows: ， ， in, For the current sampling point Plastic instability early warning indicators at the location, For the current sampling point The degree of ellipticity of the cross section at that location, Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... Historical sampling points The degree of ellipticity of the cross section at that location, Historical sampling points At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point Compared with historical sampling points The Euclidean distance on the elliptic-force plane For the current sampling point Compared with historical sampling points Euclidean distance on the elliptic-force plane Historical sampling points Compared with historical sampling points The Euclidean distance on the elliptic-force plane Sampling points Plastic instability early warning indicators at the location, Sampling points Plastic instability early warning indicators at the location, For the current sampling point At this location, the characteristic force measured by the pressure head force sensor corresponds to... For the current sampling point At the location corresponding to the characteristic force measured by the pressure head force sensor, the formula for calculating the dynamic attenuation coefficient is as follows: ， in, For the current sampling point The dynamic attenuation coefficient at that location, This represents the lower limit of the attenuation coefficient, i.e., the minimum value that the dynamic attenuation coefficient is allowed to take. The attenuation compression coefficient, For the current sampling point Plastic instability early warning indicators at the location, The preset safety threshold is the upper limit allowed for the elliptic-force curvature characteristic value.

5. The intelligent control method for straightening high-pressure oil and gas well casing according to claim 1, characterized in that, The formula for calculating the comprehensive deviation index in S6 is as follows: ， in, To form a comprehensive deviation index, This represents the maximum residual straightness deviation along the entire length of the casing after unloading. The target straightness tolerance is the maximum straightness deviation allowed by the process requirements. This represents the maximum residual stress on the casing surface after unloading. Allowable residual stress, which is the maximum permissible residual stress that does not affect the casing's resistance to crushing and fatigue life. This refers to the maximum residual ellipticity among the various test sections after unloading, i.e., the degree to which the straightened section deviates from a perfect circle. The allowable ellipticity rate, i.e., the maximum permissible degree of ellipticity of the cross-section that does not affect the sealing performance of the sleeve joint, is defined by the following formula for updating the axial position of the next pressure head: ， in, This is the position for the next axial loading of the pressure head. This indicates the current axial loading position of the pressure head. The axial coordinates represent the location of the peak residual curvature. To determine the position-following saturation coefficient, the update formula for the amplitude of the next straightening force is: ， in, This is the amplitude of the straightening force for the next pass. The straightening force amplitude per pass. For force increment gain coefficient, For force-increasing saturation coefficient, This is the current track deviation index.

Images