Pipe inner diameter deviation rounding method and system based on point cloud data

Abstract

The present application relates to the field of pipeline processing and intelligent control technology, and discloses a pipeline inner diameter deviation rounding method and system based on point cloud data; the method comprises the following steps: acquiring point cloud data before the current round of rounding action and constructing an actual deformation field, predicting the spatial residual response after unloading stabilization based on the actual deformation field, candidate rounding execution parameters and the current rounding response relationship, inversely solving the target rounding execution parameters and implementing rounding according to the spatial residual response; acquiring point cloud data again after unloading stabilization, determining the actual spatial residual response according to the difference between the actual deformation fields before and after rounding, and correcting the rounding response relationship by using the deviation between the actual spatial residual response and the predicted spatial residual response; when there is still an out-of-tolerance area, using the updated actual deformation field as the initial state of the next round, re-determining the subsequent rounding execution parameters based on the corrected rounding response relationship, until the inner diameter deviation and the roundness error meet the preset allowable range. The present application is used for closed-loop adaptive rounding of pipeline inner diameter deviation.

Description

Technical Field

[0001] This invention relates to the field of pipeline processing and intelligent control technology, specifically to a method and system for correcting pipeline inner diameter deviations based on point cloud data. Background Technology

[0002] During the manufacturing, welding, heat treatment, transportation, and installation of pipelines, factors such as material property fluctuations, residual stress, localized forces, and processing errors can easily cause elliptical, locally uneven, or other forms of internal diameter deviations in their cross-sections. For pipelines with high requirements for internal diameter and roundness, radial action needs to be applied to the out-of-tolerance locations using rounding equipment to ensure that the pipeline forms a spatial residual response that meets design requirements after unloading.

[0003] In existing pipe straightening processes, the pipe's dimensions or contour information is typically obtained using calipers, internal diameter measuring devices, cross-sectional profile detection devices, or 3D detection devices. The straightening position and straightening amount are then determined based on the detection results. Among these methods, using 3D point cloud data can obtain geometric information of multiple spatial locations on the pipe's inner surface. Compared to measuring the discrete dimensions of a small number of cross-sections, this is beneficial for characterizing the spatial distribution of pipe deformation in the axial and circumferential directions.

[0004] However, pipe straightening involves a coupled process of loading, plastic deformation, and unloading springback. Even if the same or similar geometric deviations are detected before straightening, the spatial residual response after unloading may differ due to variations in wall thickness, material properties, residual stress, and existing deformation states between different pipes and at different locations within the same pipe. Therefore, determining the straightening amount solely based on fixed empirical parameters according to the pre-straightening geometric deviation, or determining the straightening parameters solely based on a pre-established theoretical response relationship that remains unchanged during subsequent straightening, can easily lead to discrepancies between the predicted results and the actual spatial residual response after unloading springback, resulting in under-straightening or over-straightening.

[0005] Furthermore, when multiple areas to be straightened exist within the same pipeline, a straightening action at one location may cause deformation changes at adjacent locations or even in areas that already meet the allowable deviation. There is a certain spatial coupling relationship between the straightening actions. The actual spatial residual response generated by the previous straightening action will change the initial deformation state of the subsequent straightening action. If subsequent straightening actions still use the original state before straightening or the response relationship that has not been corrected by the actual straightening results for parameter determination, the prediction error may accumulate with multiple rounds of straightening. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for correcting pipe inner diameter deviation based on point cloud data. This solves the problems that existing correction methods, when determining correction parameters according to fixed empirical parameters or fixed theoretical response relationships, are difficult to reflect the differences in actual stiffness, plastic deformation, and unloading rebound at different locations of the pipe. Furthermore, the actual spatial residual response generated by the previous correction is not fully fed back to subsequent correction decisions during multiple correction rounds, which can easily lead to overcorrection, undercorrection, and cumulative prediction deviations.

[0007] The objective of this invention can be achieved through the following technical solutions: A pipe inner diameter deviation correction system based on point cloud data, the system comprising: The point cloud acquisition module is used to acquire the first point cloud data of the pipe to be straightened before straightening and the second point cloud data after at least one straightening operation is completed and the pipe is unloaded and stabilized. The deformation field construction module is used to construct the current actual deformation field and update the actual deformation field based on the first point cloud data and the second point cloud data, respectively. The response prediction module is used to obtain the predicted spatial residual response of the pipeline at multiple spatial locations after the implementation and unloading of the straightening action corresponding to the candidate straightening action parameters, based on the current actual deformation field, the candidate straightening circle execution parameters and the current straightening circle response relationship. The parameter inverse calculation module is used to determine the target circular correction execution parameters based on the current actual deformation field, the preset target pipeline shape, and the predicted spatial residual response. The circular correction execution module is used to perform circular correction according to the target circular correction execution parameters; The response correction module is used to obtain the actual spatial residual response based on the current actual deformation field and the updated actual deformation field, and to correct the current circularity correction response relationship based on the difference between the actual spatial residual response and the predicted spatial residual response. The iterative decision module is used to take the updated actual deformation field as the new current actual deformation field when there is still an out-of-tolerance region in the updated actual deformation field, and to re-predict the spatial residual response corresponding to the subsequent candidate circular correction execution parameters based on the corrected circular correction response relationship, and to redetermine the subsequent target circular correction execution parameters accordingly.

[0008] As a further aspect of the present invention: the deformation field construction module is configured to: filter, register and unify the point cloud data, determine the central axis of the pipeline, use the axial position along the central axis and the circumferential position around the central axis as spatial indices, determine the radial deviation of different spatial positions on the inner surface of the pipeline relative to the preset target pipeline shape, construct the current actual deformation field or update the actual deformation field according to the radial deviation, and determine the region where the radial deviation exceeds the preset allowable range as the region to be corrected.

[0009] As a further aspect of the present invention, the system further includes a coordinate mapping module, which is used to establish a spatial correspondence between the point cloud detection coordinate system, the pipeline coordinate system and the execution coordinate system of the straightening actuator, and to convert the area to be straightened in the actual deformation field into the corresponding action position and motion position of the straightening actuator according to the spatial correspondence, so as to establish a spatial correspondence between the area to be straightened determined by the point cloud data and the actual action position of the straightening actuator.

[0010] As a further aspect of the present invention: the straightening response relationship is used to characterize the correspondence between candidate straightening execution parameters and the spatial residual responses generated at multiple spatial positions of the pipeline after the corresponding straightening action is implemented and unloaded and stabilized; the candidate straightening execution parameters include at least one of the straightening action position, jacking displacement, jacking pressure, and holding time; the straightening response relationship is related to the current actual deformation state of the pipeline, so that the same candidate straightening execution parameter corresponds to different predicted spatial residual responses under different current actual deformation states.

[0011] As a further aspect of the present invention: the response prediction module is configured to determine the predicted spatial residual response of each candidate rounding action to its action area and multiple spatial locations outside the action area for one or more candidate rounding action positions, so as to characterize the spatial coupling effect of a candidate rounding action on other rounding areas to be rounded and / or areas that have met the allowable deviation; the parameter inverse calculation module, based on the spatial coupling effect, with the goal of meeting the preset requirements for the overall inner diameter deviation and / or roundness error of the pipe after rounding, jointly determines one or more rounding action positions and at least one of the corresponding jacking displacement, jacking pressure and holding time under the constraints of at least one of the allowable pressure, allowable displacement and allowable deformation range of the rounding actuator.

[0012] As a further aspect of the present invention: the response correction module is configured to: after one or a group of rounding actions are completed and the unloading stability is reached, register the point cloud data before and after the rounding action to a unified spatial coordinate reference, determine the actual spatial residual response after unloading of the rounding action to multiple spatial positions of the pipeline based on the radial deviation change of the corresponding spatial position before and after the rounding action, and compare the actual spatial residual response after unloading with the predicted spatial residual response obtained for the rounding action to obtain the spatial response prediction deviation.

[0013] As a further aspect of the present invention: the response correction module is further configured to correct the local residual response parameters corresponding to the position, direction and amount range of the current rounding action based on the spatial response prediction deviation. The local residual response parameters include at least one of the parameters characterizing the local stiffness of the pipeline, the unloading rebound characteristics and the correspondence between the rounding action and the spatial residual response, and form a rounding response relationship applicable to the current actual deformation state of the pipeline based on the corrected local residual response parameters.

[0014] As a further aspect of the present invention: the system further includes an execution data acquisition module, used to acquire at least one of the following execution data during the rounding execution process: actual pushing pressure, actual pushing displacement, action time, and loading sequence of the rounding execution mechanism, and to establish a correspondence between the execution data of this rounding action, the actual deformation field before rounding, and the actual spatial residual response after unloading and stabilization; the response correction module corrects the local residual response parameters corresponding to this rounding action based on the execution data and the actual spatial residual response.

[0015] As a further aspect of the present invention: the iterative decision module is configured to: after each rounding action or a set of rounding actions is completed, redetermine the current actual deformation state of the pipeline and the remaining area to be rounded based on the second point cloud data after unloading and stabilization; use the current actual deformation state redetermined after the previous rounding action as the initial state for the prediction of the next rounding response, instead of continuing to use the actual deformation state determined before the previous rounding action; and based on the current actual deformation state and the corrected rounding response relationship, re-obtain the prediction spatial residual response corresponding to the subsequent candidate rounding actions; and determine the position, action sequence, and rounding execution parameters of the subsequent rounding actions based on the re-obtained prediction spatial residual response, until at least one of the pipeline inner diameter deviation and roundness error meets the preset allowable range.

[0016] A method for correcting pipe inner diameter deviation based on point cloud data, the method comprising the following steps: S1: Construction of actual deformation field: Obtain the first point cloud data of the pipe to be straightened before straightening, and construct the current actual deformation field of the pipe based on the first point cloud data; S2: Spatial residual response prediction: Based on the current actual deformation field, candidate straightening circle execution parameters and current straightening circle response relationship, obtain the predicted spatial residual response of the pipeline at multiple spatial locations after the straightening circle action corresponding to the candidate straightening circle execution parameters is implemented and unloaded and stabilized. S3: Target rounding execution parameters inverse calculation and rounding: Based on the current actual deformation field, the preset target pipeline shape and the predicted spatial residual response, the target rounding execution parameters are determined inversely, and the pipeline is rounded according to the target rounding execution parameters; S4: Update the actual deformation field construction: After the rounding action is completed and the unloading stable state is reached, the second point cloud data is obtained, and the actual deformation field is constructed and updated based on the second point cloud data; S5: Determination of Actual Spatial Residual Response and Correction of Circular Correction Response: Register the first point cloud data and the second point cloud data to the same spatial coordinate reference. Construct the current actual deformation field and the updated actual deformation field based on the registered data. Obtain the actual spatial residual response of the pipeline at multiple spatial locations after unloading stabilization based on the radial deviation changes of the corresponding spatial positions in the current actual deformation field and the updated actual deformation field. Compare the actual spatial residual response with the predicted spatial residual response and correct the circular correction response relationship corresponding to the current circular correction action based on the difference between the two. S6: State recursive iterative roundness correction: When the updated actual deformation field still has an out-of-tolerance region, the updated actual deformation field is used as the new initial state for the next roundness correction response prediction. Based on the new initial state and the corrected roundness correction response relationship, the prediction space residual response corresponding to the subsequent candidate roundness correction execution parameters is obtained again, and the subsequent target roundness correction execution parameters are re-determined. S3-S5 are repeated until at least one of the pipe inner diameter deviation and roundness error meets the preset allowable range.

[0017] The beneficial effects of this invention are: (1) The present invention utilizes point cloud data to construct the actual spatial deformation field of the inner surface of the pipe, so that the determination of the rounding parameters is not only based on the local diameter measurement results, but also considers the spatial distribution of the inner diameter deviation along the axial and circumferential directions of the pipe.

[0018] (2) After the rounding action is completed and the unloading is stable, the present invention obtains the actual spatial residual response based on the change of the actual deformation field before and after the rounding action, and uses the difference between the actual spatial residual response and the predicted spatial residual response to correct the rounding response relationship, so that the subsequent spatial residual response prediction can adapt to the actual rounding response characteristics exhibited by the current pipeline and its local area.

[0019] (3) After each round of rounding is completed, the updated actual deformation field is used as the initial state for the next round of rounding response prediction, and the remaining rounding area to be rounded and the subsequent rounding execution parameters are re-determined, so that the multi-round rounding decision can reflect the actual plastic deformation and unloading springback results of the previous rounding, and reduce the cumulative error caused by continuously using the original state prediction.

[0020] (4) When determining the target correction circle execution parameters, the present invention considers the influence of the spatial residual response generated by the correction circle action on the spatial position outside the action area, which helps to reduce the possibility that local correction circle will cause other areas to be corrected or areas that have already met the allowable deviation to exceed the tolerance again. Attached Figure Description

[0021] The invention will now be further described with reference to the accompanying drawings.

[0022] Figure 1 This is a system module framework diagram of one embodiment of the present invention; Figure 2 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 As shown, in one embodiment, a method and system for correcting pipe inner diameter deviation based on point cloud data are provided, the system comprising: The point cloud acquisition module is used to acquire the first point cloud data of the pipe to be straightened before straightening and the second point cloud data after at least one straightening operation is completed and the pipe is unloaded and stabilized. The deformation field construction module is used to construct the current actual deformation field and update the actual deformation field based on the first point cloud data and the second point cloud data, respectively. The response prediction module is used to obtain the predicted spatial residual response of the pipeline at multiple spatial locations after the implementation and unloading of the straightening action corresponding to the candidate straightening action parameters, based on the current actual deformation field, the candidate straightening circle execution parameters and the current straightening circle response relationship. The parameter inverse calculation module is used to determine the target circular correction execution parameters based on the current actual deformation field, the preset target pipeline shape, and the predicted spatial residual response. The circular correction execution module is used to perform circular correction according to the target circular correction execution parameters; The response correction module is used to obtain the actual spatial residual response based on the current actual deformation field and the updated actual deformation field, and to correct the current circularity correction response relationship based on the difference between the actual spatial residual response and the predicted spatial residual response. The iterative decision module is used to take the updated actual deformation field as the new current actual deformation field when there is still an out-of-tolerance region in the updated actual deformation field, and to re-predict the spatial residual response corresponding to the subsequent candidate circular correction execution parameters based on the corrected circular correction response relationship, and to redetermine the subsequent target circular correction execution parameters accordingly.

[0025] Specifically, before the rounding, the point cloud acquisition module acquires the first point cloud data of the pipe to be rounded, and the deformation field construction module constructs the current actual deformation field based on the spatial deviation between the actual shape of the pipe represented by the first point cloud data and the preset target pipe shape. This field is used to characterize the actual deviation state of the pipe at different spatial positions before rounding.

[0026] The response prediction module uses the current actual deformation field as the prediction starting state. Based on the candidate straightening circle execution parameters and the current straightening circle response relationship, it predicts the pipeline deformation state after the corresponding straightening circle action is implemented and the pipeline stabilizes after unloading. It then determines the prediction space residual response based on the change in the predicted deformation state relative to the current actual deformation field. The current straightening circle response relationship characterizes the residual change in the pipeline's unloading stable state relative to its state before straightening circle action after implementing the corresponding candidate straightening circle action under a given actual deformation state.

[0027] To clarify the relationship between the current actual deformation field, the spatial residual response, and the deformation state after straightening, let the current actual deformation field before the k-th round (k≥1) straightening action be D. k D k (z,θ) represents the radial deviation of the actual radial distance at the axial position z and the circumferential position θ from the target radial distance corresponding to the preset target pipe shape; the candidate circular correction execution parameter is denoted as P. k The current circular correction response relationship is denoted as F. k The predicted spatial residual response corresponding to the candidate correction circle execution parameters is: ; The predicted spatial residual response represents the radial deviation change relative to the current actual deformation field after the circular correction action is performed and the deformation stabilizes. Therefore, the predicted actual deformation field after the candidate circular correction action satisfies: ; As one feasible implementation, the circularity response relationship is pre-established and dynamically updated through the following steps: (I) Construction of Offline Simulation Database: Based on the nominal geometric parameters (including outer diameter, wall thickness, and length) and material parameters (including elastic modulus, Poisson's ratio, yield strength, and hardening index) of the pipe to be straightened, an elastoplastic finite element model for pipe straightening is established in finite element analysis software. In the model, multiple candidate straightening action positions are uniformly selected along the axial and circumferential directions of the pipe, and multiple horizontal jacking displacement candidate values ​​are set for each candidate position (e.g., values ​​within the range of 0~5mm with a step size of 0.5mm). The loading and unloading process is simulated one by one, and the radial residual deformation of all spatial nodes on the inner surface of the pipe is recorded after each simulation unloading stabilizes. All simulation results constitute the initial straightening response relationship database.

[0028] (II) Online Proxy Model: To meet the real-time requirements of online prediction, based on the aforementioned simulation database, a proxy model is trained using Kriging interpolation or support vector regression methods. This proxy model takes the feature vector of the current actual deformation field of the pipeline and the candidate straightening execution parameters (operation position, jacking displacement) as input, and the predicted spatial residual response as output. This proxy model is the current straightening response relationship F that can be quickly invoked during the straightening process. k .

[0029] Inverse parameter calculation module to predict actual deformation field The optimization objective function is constructed with the goal of minimizing the radial deviation at each spatial location. As one implementation, the predicted squared radial deviations at each spatial location are weighted and summed, with the minimum weighted sum of squares serving as the optimization objective. A first weight can be assigned to the region to be corrected, and a second weight can be assigned to regions that currently meet the preset allowable range but may be affected by the spatial coupling of the correction action. This simultaneously considers the elimination of deviation in the region to be corrected and the impact of the correction action on other spatial locations. When roundness error is also considered, the predicted roundness error of each axial section determined based on the predicted actual deformation field is added to the optimization objective, or the constraint condition is that the predicted roundness error does not exceed a preset allowable roundness value.

[0030] Under the aforementioned optimization objective, and within at least one constraint of the allowable jacking pressure, allowable jacking displacement, allowable holding time, and allowable deformation range of the pipeline, the parameter inverse calculation module iteratively optimizes the candidate pipeline straightening execution parameters. As one implementation method, a sequential quadratic programming algorithm is used to iteratively solve for the candidate pipeline straightening execution parameters, determining the candidate parameters that satisfy the preset optimization conditions and do not violate the constraints, and identifying these as the target pipeline straightening execution parameters. The pipeline straightening execution module then performs pipeline straightening according to the target pipeline straightening execution parameters.

[0031] In this invention, 'reaching a stable unloading state' or 'after unloading stabilization' refers to a state where the following conditions are met: After the push head of the rounding actuator is completely detached from the inner wall of the pipe and all external loads are removed, the radial rebound displacement of multiple key sections on the inner surface of the pipe is continuously monitored by displacement sensors and / or point cloud acquisition devices. When the rate of change of the radial rebound displacement is less than 0.01 mm / min within 10 consecutive seconds, it is determined that the unloading stabilization state has been reached. Only then can the point cloud acquisition module trigger the acquisition of the second point cloud data.

[0032] After the current round (kth round) has completed the circular correction and unloaded and stabilized, the actual deformation field D is updated based on the second point cloud data. k+1 The actual spatial residual response generated by the current wheel's circular correction action is: ; The spatial response prediction bias is: ; The response correction module is based on E k The distribution of different axial and circumferential positions is used to correct the current rounding response relationship.

[0033] As one feasible implementation, the response correction module employs a Bayesian regularization method to perform online correction of local parameters in the circular correction response relationship. Specifically, the actual spatial residual response R obtained from this circular correction is... actual,k As a new observation sample, along with the corresponding actual deformation field D before circular correction k and actual execution parameter P k The data is then input into the initially constructed surrogate model. Using a Bayesian framework, the posterior distribution of the surrogate model's hyperparameters is updated based on the new samples. Where θ is the hyperparameter vector of the surrogate model, and P(θ) is the prior distribution. The updated posterior distribution, representing the likelihood function, is used to generate predictions for the (k+1)th round. The hyperparameters include, but are not limited to, parameters characterizing the local stiffness of the pipeline and its unloading rebound characteristics.

[0034] It is important to emphasize that the core timing logic of this invention is "correction first, iteration later": after each round of circular correction is completed and the second point cloud data is acquired, the system first executes the response correction step, using the actual spatial residual response obtained from this circular correction to update the circular correction response relationship; only after the correction is completed does the iterative decision module determine and update the actual deformation field D. k+1 Does a deviation still exist that exceeds the preset allowable range? If so, the updated actual deformation field is used as the current actual deformation field for the next round of circular correction (i.e., let D...). k+1 Become the new D k The system then calls upon the corrected circular response relationship to re-predict the subsequent residual response and inversely calculate the parameters. This strict timing of "correcting first and then predicting" ensures that the system error can converge round by round, which is a key feature that distinguishes it from existing open-loop or semi-open-loop schemes.

[0035] The above technical solution can correct the straightening response relationship based on the spatial residual response after the actual straightening of the pipeline, and determine the subsequent straightening parameters based on the updated actual deformation state. This reduces over-straightening or under-straightening caused by the deviation between theoretical prediction and actual springback and plastic deformation response, and improves the accuracy of pipeline inner diameter deviation straightening and adaptability to different pipeline actual conditions.

[0036] In one embodiment of the present invention, the deformation field construction module is configured to: filter, register, and unify the coordinates of the point cloud data, determine the central axis of the pipeline, use the axial position along the central axis and the circumferential position around the central axis as spatial indices, determine the radial deviation of different spatial positions on the inner surface of the pipeline relative to the preset target pipeline shape, construct the current actual deformation field or update the actual deformation field based on the radial deviation, and determine the region where the radial deviation exceeds the preset allowable range as the region to be corrected.

[0037] Specifically, the deformation field construction module filters the acquired point cloud data to remove outliers and obvious anomalies. For point cloud data obtained from multiple scanning positions, point cloud registration unifies each point cloud to the same spatial coordinate reference, resulting in point cloud data of the pipe's inner surface. The pipe's central axis is fitted based on the processed point cloud data. Using the axial position along the central axis and the circumferential position around the central axis as spatial position indices, the actual radial distance from the pipe's inner surface to the central axis at each spatial position is determined. This is compared with the target radial distance of a preset target pipe shape at the corresponding spatial position. The radial deviation between the two represents the actual deformation state at that spatial position, and the radial deviations at each spatial position constitute the actual deformation field. The first and second point cloud data are processed using the same spatial position indices to obtain the current actual deformation field and the updated actual deformation field. The radial deviations at each spatial position are compared with a preset allowable range. When the radial deviation exceeds the preset allowable range, the corresponding position is determined as an out-of-tolerance position, and the area to be corrected is determined based on the continuously distributed out-of-tolerance positions in space.

[0038] Through the above technical solution, this embodiment uses the axial and circumferential positions of the pipe as spatial position indices. Based on the deviation between the actual radial distance on the inner surface of the pipe and the radial distance corresponding to the preset target pipe shape, the actual deformation field is constructed, and the area to be corrected is determined accordingly. This allows the distribution of the pipe's inner diameter deviation at different spatial positions to be characterized, providing deformation data with spatial position and deviation information for subsequent rounding processing. This avoids the problem that relying solely on local diameter measurement results makes it difficult to reflect the deformation state of the pipe at different positions.

[0039] As one embodiment of the present invention, the system further includes a coordinate mapping module, which is used to establish a spatial correspondence between the point cloud detection coordinate system, the pipeline coordinate system and the execution coordinate system of the straightening actuator, and to convert the area to be straightened in the actual deformation field into the corresponding action position and motion position of the straightening actuator according to the spatial correspondence, so as to establish a spatial correspondence between the area to be straightened determined by the point cloud data and the actual action position of the straightening actuator.

[0040] Specifically, the coordinate mapping module sets multiple common reference points or calibration components with multiple known feature points located at different spatial positions within the point cloud detection range and the working range of the straightening actuator. The position of the common reference points or calibration components in the point cloud detection coordinate system is acquired by the point cloud acquisition device, and their corresponding position in the pipeline coordinate system is determined based on the preset positional relationship between the common reference points or calibration components and the pipeline positioning reference. The corresponding position of the common reference points or calibration components in the execution coordinate system of the straightening actuator is determined based on the zero-point position, installation position, and motion relationship of the straightening actuator, thereby establishing a spatial correspondence between the point cloud detection coordinate system, the pipeline coordinate system, and the execution coordinate system of the straightening actuator. After determining the area to be straightened, the position of the area to be straightened in the pipeline coordinate system is converted into a target action position in the execution coordinate system of the straightening actuator based on the spatial correspondence, and the corresponding target motion position is determined based on the positional relationship between each motion axis of the straightening actuator and the execution coordinate system of the straightening actuator.

[0041] Through the above technical solution, this embodiment establishes a spatial correspondence between the point cloud detection coordinate system, the pipeline coordinate system, and the execution coordinate system of the straightening actuator. The area to be straightened determined by the point cloud data is converted into the target action position and movement position corresponding to the straightening actuator. This allows the detected pipeline deformation position to correspond to the actual action position of the straightening actuator, reducing the positional deviation caused by manual positioning during the conversion of the detection result to the equipment execution position, and providing a spatial positional basis for the straightening actuator to accurately locate the area to be straightened.

[0042] As one embodiment of the present invention, the straightening response relationship is used to characterize the correspondence between candidate straightening execution parameters and the spatial residual response generated at multiple spatial positions of the pipeline after the corresponding straightening action is implemented and unloaded and stabilized; the candidate straightening execution parameters include at least one of the straightening action position, jacking displacement, jacking pressure, and holding time; the straightening response relationship is related to the current actual deformation state of the pipeline, so that the same candidate straightening execution parameter corresponds to different predicted spatial residual responses under different current actual deformation states.

[0043] Specifically, the straightening response relationship simultaneously incorporates candidate straightening execution parameters and the current actual deformation state of the pipeline. The current actual deformation state is characterized by the radial deviation of multiple spatial locations in the current actual deformation field. The candidate straightening execution parameters include the straightening action position and at least one of the following: jacking displacement, jacking pressure, and holding time. When determining the predicted spatial residual response corresponding to the candidate straightening execution parameters, the current actual deformation state is used as the initial state before the straightening action. The straightening action conditions are determined according to the candidate straightening execution parameters. After calculating the loading and unloading deformation process of the pipeline under the straightening action conditions, the spatial residual response of multiple spatial locations of the pipeline relative to the initial state is obtained. The spatial residual response characterizes the predicted spatial residual response of the corresponding candidate straightening execution parameter under the current actual deformation state. For different current actual deformation states, calculations are performed using the corresponding actual deformation state as the initial state, so that the predicted spatial residual response corresponding to the same candidate straightening execution parameter changes with the different current actual deformation states.

[0044] Through the above technical solution, this embodiment uses the current actual deformation state of the pipeline as the input condition for the rounding response relationship, so that the predicted spatial residual response corresponding to the candidate rounding execution parameters can be adjusted with the change of the current actual deformation state of the pipeline. This reduces the spatial residual response prediction deviation caused by the difference in the actual deformation state of the pipeline when using a fixed response relationship, and makes the predicted spatial residual response more consistent with the rounding response characteristics of the pipeline under the current actual deformation state, providing a more accurate response prediction basis for determining the target rounding execution parameters.

[0045] In one embodiment of the present invention, the response prediction module is configured to determine the predicted spatial residual response of each candidate rounding action to its action area and multiple spatial locations outside the action area for one or more candidate rounding action positions, so as to characterize the spatial coupling effect of a candidate rounding action on other rounding areas to be rounded and / or areas that have met the allowable deviation; the parameter inverse calculation module, based on the spatial coupling effect, with the goal of meeting the preset requirements for the overall inner diameter deviation and / or roundness error of the pipe after rounding, jointly determines one or more rounding action positions and at least one of the corresponding jacking displacement, jacking pressure and holding time under the constraints of at least one of the allowable pressure, allowable displacement and allowable deformation range of the rounding actuator.

[0046] Specifically, the response prediction module, for one or more candidate straightening action locations, determines the predicted spatial residual response of the action area and multiple spatial locations outside the action area after the corresponding straightening action is implemented and stabilized. Based on this, it determines the impact of the candidate straightening action on other areas to be straightened and areas that have met the allowable deviation. The parameter inversion module generates candidate combinations of different straightening action locations and corresponding execution parameters within the allowable parameter range of the straightening actuator. When a candidate combination includes multiple straightening action locations, these multiple straightening action locations and their corresponding execution parameters are used together as candidate execution conditions to determine the overall predicted deformation result after the candidate combination is implemented and stabilized, thus characterizing the comprehensive deformation impact on different spatial locations of the pipeline after multiple straightening actions are implemented together. The allowable pressure, allowable displacement, and allowable pipeline deformation range are predetermined based on the equipment parameters of the straightening actuator and the pipeline design requirements. Based on the overall predicted deformation results corresponding to each candidate combination, determine whether the overall inner diameter deviation and / or roundness error of the pipeline meets the preset requirements, and exclude candidate combinations whose jacking pressure, jacking displacement, or pipeline deformation exceeds the corresponding allowable range, as well as candidate combinations in the predicted rounding results that cause the area that originally met the allowable deviation to exceed the allowable range again; determine one or more target rounding action positions and at least one of their corresponding jacking displacement, jacking pressure, and holding time from the remaining candidate combinations, so that the determination of the target rounding execution parameters simultaneously considers the spatial coupling effects generated by different rounding actions.

[0047] Through the above technical solution, this embodiment considers the spatial residual response influence of each candidate rounding action on the action area and other spatial positions when determining the rounding action position and execution parameters. It also performs joint screening and determination of multiple candidate rounding actions within the equipment capacity and pipeline allowable deformation range. This can reduce the problem of local over-correction, under-correction or qualified areas exceeding the tolerance again caused by not considering spatial coupling when determining each rounding action individually, and improve the rounding accuracy of the overall pipeline inner diameter deviation and roundness.

[0048] In one embodiment of the present invention, the response correction module is configured to: after one or a group of rounding actions are completed and the unloading stabilization state is reached, register the point cloud data before and after the rounding action to a unified spatial coordinate reference, determine the actual spatial residual response after unloading of the rounding action to multiple spatial positions of the pipeline based on the radial deviation change of the corresponding spatial position before and after the rounding action, and compare the actual spatial residual response after unloading with the predicted spatial residual response obtained for the rounding action to obtain the spatial response prediction deviation.

[0049] Specifically, after one or more rounding operations are completed and fully unloaded, point cloud data after rounding is acquired once the pipeline reaches a stable unloading state. The point cloud data before and after the rounding operation are registered to a unified spatial coordinate reference, ensuring that the corresponding spatial positions before and after rounding are under the same coordinate reference. The radial deviation of each corresponding spatial position before and after rounding is determined, and the actual spatial residual response after unloading at the corresponding spatial position is determined based on the change in the radial deviation. The actual spatial residual response after unloading corresponding to this rounding operation is composed of the actual spatial residual responses at multiple spatial positions. The actual spatial residual response after unloading is compared with the predicted spatial residual response obtained before the implementation of the same rounding operation according to the corresponding spatial position, and the spatial response prediction deviation is determined based on the difference between the two.

[0050] Through the above technical solution, this embodiment obtains the actual spatial residual response by comparing the radial deviation changes of the corresponding spatial positions before and after the rounding action is completed and completely unloaded to a stable state. This is then compared with the predicted spatial residual response corresponding to the rounding action, thereby determining the spatial response prediction deviation between the predicted result and the actual spatial residual response after the pipeline is unloaded. This avoids characterizing the final rounding effect by the instantaneous deformation under loading conditions, and provides data basis for subsequently correcting the rounding response relationship based on the actual spatial residual response characteristics of the pipeline.

[0051] As one embodiment of the present invention, the response correction module is further configured to correct the local residual response parameters corresponding to the position, direction and amount range of the current rounding action based on the spatial response prediction deviation. The local residual response parameters include at least one of the parameters characterizing the local stiffness of the pipeline, the unloading rebound characteristics and the correspondence between the rounding action and the spatial residual response, and form a rounding response relationship applicable to the current actual deformation state of the pipeline based on the corrected local residual response parameters.

[0052] Specifically, the response correction module determines the local residual response parameters corresponding to the current rounding action based on the location, direction, and magnitude of the action. The magnitude range can be pre-divided according to the jacking displacement or jacking pressure used to characterize the intensity of the rounding action. After obtaining the spatial response prediction deviation, the parameters are adjusted according to a preset step size within the preset value range of the corresponding local residual response parameters, or multiple candidate parameter values ​​are set, and the predicted spatial residual response corresponding to the current rounding action is re-determined using the adjusted parameters. The re-determined predicted spatial residual response is compared with the actual spatial residual response obtained after unloading, and the parameter value that reduces the spatial response prediction deviation is selected. The local residual response parameters corresponding to the corresponding action location, direction, and magnitude range are updated with the selected parameter values. Based on the updated local residual response parameters, a rounding response relationship applicable to the current actual deformation state of the pipeline is formed.

[0053] Through the above technical solution, this embodiment corrects the local residual response parameters corresponding to the specific circular correction action position, action direction and action amount range based on the spatial response prediction deviation, so that the circular correction response relationship can reflect the differences in local residual response under different positions and action conditions of the pipeline, thereby reducing the prediction deviation caused by using uniform or fixed response parameters and improving the accuracy of subsequent spatial residual response prediction and circular correction parameter determination.

[0054] As one embodiment of the present invention, the system further includes an execution data acquisition module, used to acquire at least one of the following execution data during the rounding execution process: actual pushing pressure, actual pushing displacement, action time, and loading sequence of the rounding execution mechanism, and to establish a correspondence between the execution data of this rounding action, the actual deformation field before rounding, and the actual spatial residual response after unloading and stabilization; the response correction module corrects the local residual response parameters corresponding to this rounding action based on the execution data and the actual spatial residual response.

[0055] Specifically, the execution data acquisition module collects at least one of the following execution data during the straightening process: actual jacking pressure, actual jacking displacement, application time, and loading sequence (the loading sequence refers to the order in which multiple straightening action positions are applied when a single straightening action involves multiple straightening action positions; and the timing of each jacking action when a single straightening action uses a graded loading method). The actual jacking pressure and actual jacking displacement are obtained through pressure and displacement detection elements respectively set in the straightening execution mechanism. The application time and loading sequence are recorded by the controller according to the execution process of the corresponding straightening action. When a single straightening action includes multiple consecutive straightening actions, the corresponding execution data is recorded according to the execution order of each action. A corresponding action identifier is set for each straightening action, and the execution data is associated with the actual deformation field before the straightening action and the actual spatial residual response obtained after unloading and stabilization. During the correction of the corresponding local residual response parameters, the actual execution data is used as the actual input condition for this straightening action, and the local residual response parameters corresponding to this straightening action are corrected in conjunction with the actual spatial residual response corresponding to the actual input condition.

[0056] Through the above technical solution, this embodiment collects at least one of the following execution data from the actual applied jacking pressure, jacking displacement, action time, and loading sequence of the straightening actuator, and establishes a correspondence between it and the actual deformation field before straightening and the actual spatial residual response after unloading. This allows the correction of local residual response parameters to be based on the actual execution conditions and the actual spatial residual response generated, thereby reducing the impact of the deviation between the target execution parameters and the actual execution state on the result of response parameter correction and reducing the prediction deviation of the straightening response relationship on the actual straightening process.

[0057] In one embodiment of the present invention, the iterative decision module is configured to: after each rounding action or a set of rounding actions is completed, redetermine the current actual deformation state of the pipeline and the remaining area to be rounded based on the second point cloud data after unloading and stabilization; use the current actual deformation state redetermined after the previous rounding action as the initial state for the prediction of the next rounding response, instead of continuing to use the actual deformation state determined before the previous rounding action; and obtain the prediction spatial residual response corresponding to the subsequent candidate rounding actions based on the current actual deformation state and the corrected rounding response relationship; and determine the position, action sequence and rounding execution parameters of the subsequent rounding actions based on the newly obtained prediction spatial residual response, until at least one of the pipeline inner diameter deviation and roundness error meets the preset allowable range.

[0058] Specifically, after each rounding action or set of rounding actions is completed and the system stabilizes, the iterative decision module redetermines the current actual deformation state of the pipeline based on the updated actual deformation field corresponding to the second point cloud data, and identifies the area still exceeding the preset allowable range as the remaining rounding area to be rounded. The updated actual deformation field is used as the initial state for the next rounding response prediction. For the remaining rounding area to be rounded, the predicted spatial residual response corresponding to subsequent candidate rounding actions is re-obtained, and the position, sequence, and execution parameters of subsequent rounding actions are re-determined based on the corrected rounding response relationship. After each rounding action is completed, the initial state and remaining rounding area to be rounded are updated in the above manner until at least one of the pipeline inner diameter deviation and roundness error meets the preset allowable range.

[0059] Through the above technical solution, this embodiment uses the newly obtained actual deformation state as the initial state for predicting the subsequent rounding response after each rounding action is completed and unloaded and stabilized. Based on this, the remaining rounding area to be rounded and the subsequent rounding action are re-determined. This enables the subsequent rounding decision to reflect the actual plastic deformation and springback results caused by the previous rounding, reduces the cumulative deviation caused by continuing to use the original deformation state for prediction, and improves the accuracy of determining the subsequent action position, action sequence and execution parameters in the multi-round rounding process.

[0060] Please see Figure 2 As shown, a method for correcting pipe inner diameter deviation based on point cloud data includes the following steps: S1: Construction of actual deformation field: Obtain the first point cloud data of the pipe to be straightened before straightening, and construct the current actual deformation field of the pipe based on the first point cloud data; S2: Spatial residual response prediction: Based on the current actual deformation field, candidate straightening circle execution parameters and current straightening circle response relationship, obtain the predicted spatial residual response of the pipeline at multiple spatial locations after the straightening circle action corresponding to the candidate straightening circle execution parameters is implemented and unloaded and stabilized. S3: Target rounding execution parameters inverse calculation and rounding: Based on the current actual deformation field, the preset target pipeline shape and the predicted spatial residual response, the target rounding execution parameters are determined inversely, and the pipeline is rounded according to the target rounding execution parameters; S4: Update the actual deformation field construction: After the rounding action is completed and the unloading stable state is reached, the second point cloud data is obtained, and the actual deformation field is constructed and updated based on the second point cloud data; S5: Determination of Actual Spatial Residual Response and Correction of Circular Correction Response: Register the first point cloud data and the second point cloud data to the same spatial coordinate reference. Construct the current actual deformation field and the updated actual deformation field based on the registered data. Obtain the actual spatial residual response of the pipeline at multiple spatial locations after unloading stabilization based on the radial deviation changes of the corresponding spatial positions in the current actual deformation field and the updated actual deformation field. Compare the actual spatial residual response with the predicted spatial residual response and correct the circular correction response relationship corresponding to the current circular correction action based on the difference between the two. S6: State recursive iterative roundness correction: When the updated actual deformation field still has an out-of-tolerance region, the updated actual deformation field is used as the new initial state for the next roundness correction response prediction. Based on the new initial state and the corrected roundness correction response relationship, the prediction space residual response corresponding to the subsequent candidate roundness correction execution parameters is obtained again, and the subsequent target roundness correction execution parameters are re-determined. S3-S5 are repeated until at least one of the pipe inner diameter deviation and roundness error meets the preset allowable range.

[0061] To further illustrate the implementation process of this invention, a numerical example of a specific application scenario is given below: Assume the pipe to be straightened is a Q345 steel pipe with a nominal outer diameter of 500mm and a wall thickness of 10mm. The preset allowable inner diameter deviation range is ±2mm.

[0062] Before the first round of circularity correction, the point cloud acquisition module acquired the first point cloud data. The deformation field construction module calculated and found that there was an elliptical deformation at the axial position z=1000mm, with a maximum radial deviation D1 of -4.5mm (manifested as a local inward concavity), which exceeded the allowable range.

[0063] The response prediction module calls the initially established Kriging proxy model (initial rounding response relationship F1), and the parameter inverse calculation module optimizes through the sequential quadratic programming algorithm to inversely calculate the target candidate rounding execution parameters P1 in the current state: the action position z=1000mm, the pushing direction is aligned with the deepest part of the depression, and the target pushing displacement is 6.5mm (taking into account the material springback).

[0064] After the circularity straightening actuator performs the push and unloads to stabilize at P1, the point cloud acquisition module obtains the second point cloud data. The deformation field construction module calculates the updated radial deviation D2 as -1.0mm. Therefore, the first round actual spatial residual response R... actual,1 = -1.0 - (-4.5) = 3.5mm.

[0065] However, the prediction module originally predicted the rebound residual response R pred,1The expected deviation is 4.5 mm (i.e., zero). There is a spatial response prediction deviation E1 = 3.5 - 4.5 = -1.0 mm. This indicates that the actual stiffness of this local area is higher than the initial model estimate.

[0066] The response correction module is based on R actual,1 =3.5mm. The hyperparameter θ, which represents the local stiffness in the surrogate model F1, is updated posteriorly using Bayes' formula to obtain the corrected circularity response relationship F2.

[0067] At this point, since the updated deviation D2 (-1.0mm) is within the preset allowable range (±2mm), the iterative decision module determines that the area is within acceptable limits for rounding and does not need to proceed to the next round of pushing. If D2 is still -2.5mm (out of tolerance), the system will use D2 as a new starting point, continue to predict and execute the next round of parameters P2 using the corrected F2, thereby avoiding continuous undercorrection caused by inaccurate model parameters.

[0068] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A pipe inner diameter deviation correction system based on point cloud data, characterized in that, The system includes: The point cloud acquisition module is used to acquire the first point cloud data of the pipe to be straightened before straightening and the second point cloud data after at least one straightening operation is completed and the pipe is unloaded and stabilized. The deformation field construction module is used to construct the current actual deformation field and update the actual deformation field based on the first point cloud data and the second point cloud data, respectively. The response prediction module is used to obtain the predicted spatial residual response of the pipeline at multiple spatial locations after the implementation and unloading of the straightening action corresponding to the candidate straightening action parameters, based on the current actual deformation field, the candidate straightening circle execution parameters and the current straightening circle response relationship. The parameter inverse calculation module is used to determine the target circular correction execution parameters based on the current actual deformation field, the preset target pipeline shape, and the predicted spatial residual response. The circular correction execution module is used to perform circular correction according to the target circular correction execution parameters; The response correction module is used to obtain the actual spatial residual response based on the current actual deformation field and the updated actual deformation field, and to correct the current circularity correction response relationship based on the difference between the actual spatial residual response and the predicted spatial residual response. The iterative decision module is used to take the updated actual deformation field as the new current actual deformation field when there is still an out-of-tolerance region in the updated actual deformation field, and to re-predict the spatial residual response corresponding to the subsequent candidate circular correction execution parameters based on the corrected circular correction response relationship, and to redetermine the subsequent target circular correction execution parameters accordingly.

2. The pipe inner diameter deviation correction system based on point cloud data according to claim 1, characterized in that, The deformation field construction module is configured to: filter, register, and unify the coordinates of the point cloud data; determine the central axis of the pipeline; use the axial position along the central axis and the circumferential position around the central axis as spatial indices; determine the radial deviation of different spatial positions on the inner surface of the pipeline relative to the preset target pipeline shape; construct the current actual deformation field or update the actual deformation field based on the radial deviation; and determine the region where the radial deviation exceeds the preset allowable range as the region to be corrected.

3. The pipe inner diameter deviation correction system based on point cloud data according to claim 1, characterized in that, The system also includes a coordinate mapping module, which is used to establish a spatial correspondence between the point cloud detection coordinate system, the pipeline coordinate system and the execution coordinate system of the straightening actuator, and to convert the area to be straightened in the actual deformation field into the corresponding action position and motion position of the straightening actuator according to the spatial correspondence, so as to establish a spatial correspondence between the area to be straightened determined by the point cloud data and the actual action position of the straightening actuator.

4. The pipe inner diameter deviation correction system based on point cloud data according to claim 1, characterized in that, The straightening response relationship is used to characterize the correspondence between the candidate straightening execution parameters and the spatial residual response generated at multiple spatial locations of the pipeline after the corresponding straightening action is implemented and the unloading is stable. The candidate straightening circle execution parameters include at least one of the following: the straightening circle action position, the jacking displacement, the jacking pressure, and the holding time; the straightening circle response relationship is related to the current actual deformation state of the pipeline, so that the same candidate straightening circle execution parameters correspond to different predicted spatial residual responses under different current actual deformation states.

5. The pipe inner diameter deviation correction system based on point cloud data according to claim 1, characterized in that, The response prediction module is configured to determine the predicted spatial residual response of each candidate rounding action to its action area and multiple spatial locations outside the action area for one or more candidate rounding action positions, so as to characterize the spatial coupling effect of a candidate rounding action on other rounding areas to be rounded and / or areas that have met the allowable deviation; the parameter inverse calculation module, based on the spatial coupling effect, with the goal of meeting the preset requirements for the overall inner diameter deviation and / or roundness error of the pipe after rounding, jointly determines one or more rounding action positions and at least one of the corresponding jacking displacement, jacking pressure and holding time under the constraints of at least one of the allowable pressure, allowable displacement and allowable deformation range of the rounding actuator.

6. The pipe inner diameter deviation correction system based on point cloud data according to claim 1, characterized in that, The response correction module is configured to: after one or a group of rounding actions are completed and the unloading stability is reached, register the point cloud data before and after the rounding action to a unified spatial coordinate reference, determine the actual spatial residual response after unloading of the rounding action to multiple spatial positions of the pipeline based on the radial deviation change of the corresponding spatial position before and after the rounding action, and compare the actual spatial residual response after unloading with the predicted spatial residual response obtained for the rounding action to obtain the spatial response prediction deviation.

7. The pipe inner diameter deviation correction system based on point cloud data according to claim 6, characterized in that, The response correction module is further configured to correct the local residual response parameters corresponding to the position, direction and amount range of the current straightening action based on the spatial response prediction deviation. The local residual response parameters include at least one of the parameters characterizing the local stiffness of the pipeline, the unloading rebound characteristics and the correspondence between the straightening action and the spatial residual response. Based on the corrected local residual response parameters, a straightening response relationship applicable to the current actual deformation state of the pipeline is formed.

8. The pipe inner diameter deviation correction system based on point cloud data according to claim 7, characterized in that, The system also includes an execution data acquisition module, which is used to acquire at least one of the execution data of the actual pushing pressure, actual pushing displacement, action time and loading sequence of the straightening actuator during the straightening process, and to establish the correspondence between the execution data of this straightening action, the actual deformation field before straightening, and the actual spatial residual response after unloading and stabilization. The response correction module corrects the local residual response parameters corresponding to this circular correction action based on the execution data and the actual spatial residual response.

9. The pipe inner diameter deviation correction system based on point cloud data according to claim 8, characterized in that, The iterative decision module is configured to: after each rounding action or a set of rounding actions, redetermine the current actual deformation state of the pipeline and the remaining area to be rounded based on the second point cloud data after unloading and stabilization; use the current actual deformation state redetermined after the previous rounding action as the initial state for the prediction of the next rounding response, instead of continuing to use the actual deformation state determined before the previous rounding action; and obtain the prediction space residual response corresponding to the subsequent candidate rounding actions based on the current actual deformation state and the corrected rounding response relationship; and determine the position, action sequence, and rounding execution parameters of the subsequent rounding actions based on the newly obtained prediction space residual response, until at least one of the pipeline inner diameter deviation and roundness error meets the preset allowable range.

10. A method for correcting pipe inner diameter deviation based on point cloud data, applicable to the pipe inner diameter deviation correction system based on point cloud data as described in any one of claims 1-9, characterized in that, The method includes the following steps: S1: Construction of actual deformation field: Obtain the first point cloud data of the pipe to be straightened before straightening, and construct the current actual deformation field of the pipe based on the first point cloud data; S2: Spatial residual response prediction: Based on the current actual deformation field, candidate straightening circle execution parameters and current straightening circle response relationship, obtain the predicted spatial residual response of the pipeline at multiple spatial locations after the straightening circle action corresponding to the candidate straightening circle execution parameters is implemented and unloaded and stabilized. S3: Target rounding execution parameters inverse calculation and rounding: Based on the current actual deformation field, the preset target pipeline shape and the predicted spatial residual response, the target rounding execution parameters are determined inversely, and the pipeline is rounded according to the target rounding execution parameters; S4: Update the actual deformation field construction: After the rounding action is completed and the unloading stable state is reached, the second point cloud data is obtained, and the actual deformation field is updated based on the second point cloud data. S5: Determination of Actual Spatial Residual Response and Correction of Circular Correction Response: Register the first point cloud data and the second point cloud data to the same spatial coordinate reference. Construct the current actual deformation field and the updated actual deformation field based on the registered data. Obtain the actual spatial residual response of the pipeline at multiple spatial locations after unloading stabilization based on the radial deviation changes of the corresponding spatial positions in the current actual deformation field and the updated actual deformation field. Compare the actual spatial residual response with the predicted spatial residual response and correct the circular correction response relationship corresponding to the current circular correction action based on the difference between the two. S6: State recursive iterative roundness correction: When the updated actual deformation field still has an out-of-tolerance region, the updated actual deformation field is used as the new initial state for the next roundness correction response prediction. Based on the new initial state and the corrected roundness correction response relationship, the prediction space residual response corresponding to the subsequent candidate roundness correction execution parameters is obtained again, and the subsequent target roundness correction execution parameters are re-determined. S3-S5 are repeated until at least one of the pipe inner diameter deviation and roundness error meets the preset allowable range.

Images