Tank spinning forming thickness intelligent control method and system

Abstract

The invention provides an intelligent control method and system for the spinning forming thickness of a tank. According to the method, a spinning tank surface thickness fluctuation signal is collected, and a thickness gradient tensor is generated through time sequence decomposition; extracting a thickness fluctuation intensity value in the tensor and a time sequence lag amount of thickness change of adjacent points to calculate a phase deviation feature; based on the characteristics, the curvature of a laser scanning lens on an annular spinning roller is dynamically adjusted, and a three-dimensional wall thickness cloud picture for identifying the thickness abnormal area is output through focusing and scanning of the tank body; a spinning roller motion track is obtained, thickness change-track space deviation is calculated by associating the thickness abnormal area with the phase deviation characteristics, and a track compensation vector is generated; and according to the vector, a spinning roller feeding path is dynamically optimized, and wall thickness fluctuation caused by material springback is inhibited. Dynamic wave suppression of the wall thickness of the spinning tank is achieved, thickness fluctuation is compressed, and the rejection rate is reduced.

Description

Technical Field

[0001] This application relates to the field of intelligent thickness control technology, and in particular to an intelligent thickness control method and system for tank spinning forming. Background Technology

[0002] Spin forming of tank bodies is a core process for manufacturing hollow rotating parts such as pressure vessels and storage tanks. The uniformity of its wall thickness directly determines the sealing strength and pressure resistance reliability of the product. In actual production, it is necessary to address multi-source dynamic disturbances: First, nonlinear fluctuations in material thickness lead to a mismatch between the preset trajectory and the actual deformation trajectory, easily causing localized excessive thinning or material accumulation; second, complex geometric structures require real-time adjustment of the roller pressure distribution to avoid micro-cracks or bulging deformation caused by stress concentration; third, thermo-coupling effects exacerbate material creep, causing wall thickness control to deviate from the theoretical thinning rate. Traditional position control modes rely on static trajectory planning and cannot perceive dynamic changes in the forming state. There is an urgent need for an intelligent thickness control method that integrates real-time thickness sensing, multi-parameter coupling optimization, and online compensation control to improve material utilization and product qualification rate while ensuring forming accuracy.

[0003] The current targeted solution is a closed-loop control system based on a laser profilometer and a decision module. This solution uses a high-precision laser profilometer to scan the contour of the tank forming area in real time, extracting geometric features such as the radius of action of the spinning wheel and the width of the flange. The decision module combines a deep neural network prediction model and a particle swarm optimization algorithm to dynamically output the optimal spinning conditions parameters. The control mechanism adjusts the spinning wheel posture and the mandrel speed in real time according to the optimized parameters to actively suppress the wall thickness reduction rate. This system replaces fixed parameter control with an online sensing-optimization-execution closed-loop architecture, significantly improving thickness consistency. However, its core deficiency lies in the lack of real-time diagnostic capability for micro-forming defects, and the reliance on a preset model for multi-physics coupling optimization makes it difficult to adapt to material property drift and sudden mechanical disturbances. Summary of the Invention

[0004] This application provides a method and system for intelligent control of the thickness of tank spinning, which solves the problems of uncontrolled wall thickness fluctuation caused by material springback, inaccurate phase offset compensation, and uncontrollable spinning scrap rate in the prior art.

[0005] In a first aspect, this application provides a method for intelligent control of the thickness of a tank body during spinning, including: The thickness fluctuation signal on the surface of the tank during the spinning process is collected, and the thickness fluctuation signal is separated by time-series decomposition technology to generate a thickness gradient tensor; Obtain the thickness fluctuation intensity value in the thickness gradient tensor, and obtain the temporal lag of thickness change between adjacent scan points to calculate phase offset characteristics; Based on the thickness fluctuation intensity value and phase shift characteristics, the curvature of the laser scanning lens arranged in a ring on the tank is dynamically adjusted, and the surface of the tank is focused and scanned according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map, which identifies areas with abnormal thickness. The rotation trajectory of the tank is obtained, and based on the thickness anomaly region and the phase offset feature, the spatial deviation between the thickness change and the rotation trajectory is correlated to generate a trajectory compensation vector for correcting the rotation trajectory. Based on the trajectory compensation vector, the rotary feed path is dynamically optimized to suppress wall thickness fluctuations caused by material springback and achieve intelligent control of tank thickness.

[0006] Optionally, the thickness fluctuation signal on the surface of the tank during the spinning process is acquired, and the thickness fluctuation signal is separated by time-series decomposition technology to generate a thickness gradient tensor, including: Thickness measurements are continuously collected at various locations on the surface of the tank by thickness sensors arranged in a ring on the spinning forming device. At the same time, the spatial coordinates and sampling timestamps of each measurement point are recorded. The thickness measurements, spatial coordinates and sampling timestamps are combined to form a thickness fluctuation signal dataset. The thickness fluctuation signal dataset is divided into multiple data segments according to a fixed time window. Each data segment contains thickness measurement values ​​at consecutive time points and their corresponding spatial coordinates. The thickness variation between adjacent measurement points is calculated based on the thickness measurement value of the data segment, and the spacing value of the measurement points is obtained. Based on the thickness change and the distance between measurement points, a thickness change trend value is generated, and the thickness change trend value is arranged in the order of the time window of the data segment so that each time window corresponds to a set of thickness change trend values. Integrate the set of thickness change trend values ​​from all time windows to construct the thickness gradient tensor.

[0007] Optionally, the thickness fluctuation intensity value in the thickness gradient tensor is obtained, and the temporal lag of the thickness change between adjacent scan points is obtained to calculate the phase shift characteristics, including: Extract the average value of the thickness change trend value in the thickness gradient tensor, and calculate the difference between the thickness change trend value at each spatial location point and the average value as the thickness fluctuation intensity value; In the thickness gradient tensor, adjacent spatial locations are located, and the time difference between the thickness change trend values ​​of adjacent spatial locations exceeding a preset threshold is recorded as the time lag. The sampling time interval of the thickness fluctuation signal is obtained, and a phase offset value is generated based on the time lag and the sampling time interval. The phase offset values ​​of all adjacent spatial locations are integrated, and the phase offset values ​​are arranged according to their spatial positional relationship to form a phase offset feature.

[0008] Optionally, based on the thickness fluctuation intensity value and phase shift characteristics, the curvature of the laser scanning lens arranged in a ring on the tank is dynamically adjusted, and the surface of the tank is focused and scanned according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map. The three-dimensional wall thickness cloud map identifies areas of abnormal thickness, including: The thickness fluctuation intensity value is combined with the phase shift value in the phase shift feature to generate the curvature adjustment value of the laser scanning lens; Adjust the hydraulic regulator pressure of the laser scanning lens according to the curvature adjustment value so that the radius of curvature of the laser scanning lens is changed to the target value; The adjusted laser scanning lens performs a focused scan on the surface of the tank. During the focused scan, the laser beam emitted by the laser scanning lens is incident perpendicularly on the surface of the tank and then reflected and captured by the receiver. Calculate the time difference between the laser beam's emission and its capture by the receiver, and convert the time difference into the thickness value of each scanning point on the surface of the tank; The thickness values ​​of all scan points are integrated and arranged according to spatial coordinates to form a three-dimensional wall thickness cloud map. The areas of continuous thickness anomalies in the three-dimensional wall thickness cloud map with thickness values ​​lower than a preset thickness threshold are marked as thickness anomaly areas.

[0009] Optionally, the thickness values ​​of all scanned points are integrated and arranged according to spatial coordinates to form a three-dimensional wall thickness cloud map. Regions in the three-dimensional wall thickness cloud map with continuous thickness anomalies whose thickness values ​​are lower than a preset thickness threshold are marked as thickness anomaly regions, including: Extract the thickness values ​​of all scanning points and their corresponding spatial coordinates, and mark the thickness value of each scanning point on the corresponding spatial coordinates, so that the spatial coordinates and the thickness values ​​form a corresponding mapping relationship. Arrange all spatial coordinate points with labeled thickness values ​​according to the grid order of the spatial coordinates to construct a three-dimensional lattice structure, and integrate all three-dimensional lattice structures to form a three-dimensional wall thickness cloud map. In the three-dimensional wall thickness cloud map, thickness anomalies with thickness values ​​lower than a preset thickness threshold are identified, and consecutive adjacent thickness anomalies are connected to form a closed region as a thickness anomaly region.

[0010] Optionally, the trajectory of the rotating wheel in the tank is obtained, and based on the thickness anomaly region and the phase shift characteristics, the spatial deviation between the thickness change and the trajectory of the rotating wheel is correlated to generate a trajectory compensation vector for correcting the trajectory of the rotating wheel, including: A spatial coordinate sequence of the wheel's motion trajectory is obtained by a wheel trajectory recording device. The spatial coordinate sequence includes the three-dimensional position coordinates of the wheel's center point at each time point. The spatial coordinates of each abnormal point within the thickness abnormality area are located, and the spatial coordinates of the thickness abnormal points are matched with the spatial coordinate sequence of the wheel motion trajectory to obtain the matching trajectory points. Calculate the spatial straight-line distance between the thickness anomaly point and the matching trajectory point, and combine the spatial straight-line distance with the phase offset value to generate a spatial deviation correction component; The offset direction of the wheel motion trajectory is determined based on the spatial deviation correction component, and the offset direction is determined by the direction of the line connecting the thickness anomaly point to the matching trajectory point; The spatial deviation correction components and offset directions corresponding to all thickness anomalies are integrated to form a trajectory compensation vector describing the trajectory of the spinning wheel.

[0011] Optionally, based on the trajectory compensation vector, the rotary feed path is dynamically optimized to suppress wall thickness fluctuations caused by material springback and achieve intelligent control of the tank thickness, including: The spatial deviation correction component and offset direction in the trajectory compensation vector are analyzed, and the spatial deviation correction component is combined with the preset material springback compensation coefficient to generate the path correction amount; The adjustment direction of the rotary feed path is determined based on the offset direction, and the adjustment direction is consistent with the offset direction; Obtain the spatial coordinate sequence of the current feed path of the rotary wheel, and superimpose the path correction amount onto the spatial coordinate sequence according to the adjustment direction to form the updated feed path coordinates; The updated feed path coordinates are converted into control signals for the rotary drive, and the control signals are executed by the rotary drive to make the rotary wheel move along the updated feed path coordinates to suppress wall thickness fluctuations caused by material springback.

[0012] Secondly, this application provides an intelligent control system for the thickness of tank spinning, comprising: The separation module is used to collect the thickness fluctuation signal on the surface of the tank during the spinning process, and to separate the thickness fluctuation signal through time-series decomposition technology to generate a thickness gradient tensor. The calculation module is used to obtain the thickness fluctuation intensity value in the thickness gradient tensor and obtain the temporal lag of the thickness change between adjacent scan points to calculate the phase shift characteristics. The adjustment module is used to dynamically adjust the curvature of the laser scanning lens that is arranged in a ring on the rotating wheel of the tank according to the thickness fluctuation intensity value and phase offset characteristics, and to perform focused scanning on the surface of the tank according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map, which identifies areas with abnormal thickness. The generation module is used to obtain the rotation trajectory of the tank's wheel, and based on the thickness anomaly region and the phase offset feature, associate the spatial deviation between the thickness change and the rotation trajectory to generate a trajectory compensation vector for correcting the rotation trajectory. The optimization module is used to dynamically optimize the rotary feed path based on the trajectory compensation vector, so as to suppress the wall thickness fluctuation caused by material springback and realize intelligent control of the tank thickness.

[0013] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement the intelligent control method for the thickness of tank spinning as described in the first aspect above.

[0014] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements a method for intelligent control of the thickness of a tank body during spinning as described in the first aspect.

[0015] In this application, the technical solution achieves precise control of the spinning forming process through multi-source data fusion and intelligent trajectory optimization. Specifically, the thickness gradient tensor based on time-series decomposition significantly improves the characterization accuracy of material flow properties; the dynamic adjustment mechanism of the laser scanning lens ensures high-resolution reconstruction of the three-dimensional wall thickness cloud map; and the correlation analysis between thickness anomalies and the spinning wheel trajectory enables intelligent tracing of process deviations. This method overcomes the limitations of traditional experience-based processing, effectively suppressing material springback through real-time optimization of the trajectory compensation vector, significantly improving the uniformity of tank wall thickness, and providing a data-driven intelligent solution for precision spinning forming.

[0016] Furthermore, precise control of the spinning process is achieved through high-precision thickness monitoring and intelligent gradient analysis. Specifically, the thickness fluctuation dataset collected synchronously by multiple sensors significantly improves the spatiotemporal resolution of material deformation characteristics; dynamic calculation of thickness change trend values ​​effectively quantifies the gradient characteristics of material flow; and intelligent analysis of time lag and phase shift characteristics enables precise tracing of process deviations. This method overcomes the limitations of traditional empirical control by constructing a data-driven closed-loop control system through the collaborative analysis of thickness gradient tensors and phase characteristics. This significantly improves the wall thickness uniformity and process stability of tank spinning, providing an intelligent solution for precision metal forming.

[0017] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart of a method for intelligent control of tank body thickness during spinning according to this application is shown; Figure 2 This application provides a schematic diagram of the structure of an intelligent control system for the thickness of a tank spinning process. Figure 3 A schematic diagram of the structure of a computing device provided in this application is shown. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0021] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a chronological order, nor do they limit "first" and "second" to different types.

[0022] Researchers have discovered a fundamental bottleneck in tank spinning wall thickness control technology: while a closed-loop scheme based on a laser profilometer can optimize macroscopic thickness uniformity, its lack of awareness of microscopic defects and adaptive disturbance response leads to instability in accuracy. Specifically, the localized thickening effect caused by material springback is smoothed out during profilometry scanning, and sudden mechanical disturbances cause lag in wheel trajectory compensation, exacerbating the risk of microcracks in stress concentration areas. This contradiction stems from the blind spot in the dynamic decoupling of thickness fluctuation phase characteristics and motion trajectory deviation, necessitating the construction of a closed-loop control architecture for the spatiotemporal evolution of thickness gradient and trajectory compensation.

[0023] To address the aforementioned challenges, this invention proposes an intelligent control method for the thickness of tank spinning forming. Its innovation lies in overcoming the limitations of static contour monitoring by driving lens tuning and trajectory vector fusion through wave phase characteristics. Specifically: The method involves real-time acquisition of thickness fluctuation signals during the spinning process, followed by time-series decomposition to generate a thickness gradient tensor, quantifying the fluctuation intensity and phase shift characteristics; dynamic adjustment of the curvature of the annular laser scanning lens to achieve adaptive focusing scanning of the curved tank surface, outputting a high-resolution three-dimensional wall thickness cloud map; correlation of the thickness anomaly zone with the spatial deviation of the spinning wheel trajectory through phase shift to generate a trajectory compensation vector; and real-time optimization of the spinning wheel feed path to precisely suppress material springback. This method overturns traditional control paradigms: gradient tensor phase analysis achieves millisecond-level dynamic capture of microscopic springback thickening for the first time; the dynamic focusing mechanism overcomes the scanning distortion problem caused by complex curvature; and the compensation vector, by fusing thickness-trajectory spatiotemporal deviations, forms a closed-loop control chain of "wave perception - cloud map reconstruction - deviation decoupling - trajectory self-healing," providing a full-link optimization paradigm for spinning forming, from material transient response to precise mechanical trajectory control.

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] Figure 1 This application provides a flowchart of a method for intelligent control of tank body spin forming thickness, as shown in the following embodiment. Figure 1 As shown, the method includes: 101. Collect the thickness fluctuation signal on the surface of the tank during the spinning process, and separate the thickness fluctuation signal using time-series decomposition technology to generate a thickness gradient tensor.

[0026] Optionally, step 101 may specifically include the following steps: 1011. Thickness measurements at various locations on the surface of the tank are continuously collected by thickness sensors arranged in a ring on the spinning forming device. At the same time, the spatial coordinates and sampling timestamps of each measurement point are recorded. The thickness measurements, spatial coordinates and sampling timestamps are combined to form a thickness fluctuation signal dataset.

[0027] 1012. The thickness fluctuation signal dataset is divided into multiple data segments according to a fixed time window, and the data segments contain thickness measurement values ​​at continuous time points and their corresponding spatial coordinates.

[0028] 1013. Calculate the thickness change between adjacent measurement points based on the thickness measurement value of the data segment, and obtain the spacing value of the measurement points.

[0029] 1014. Generate a thickness change trend value based on the thickness change amount and the measurement point spacing value, and arrange the thickness change trend values ​​in the order of the time window of the data segment so that each time window corresponds to a set of thickness change trend values.

[0030] 1015. Integrate the set of thickness change trend values ​​from all time windows to construct the thickness gradient tensor.

[0031] In the above scheme, spinning refers to a process of forming by rotating and pressing. The tank body refers to the formed container. Thickness fluctuation signal refers to the signal indicating thickness change. Temporal decomposition technique refers to a method of decomposing signals according to time. Thickness gradient tensor refers to a tensor reflecting the thickness gradient. Thickness sensor refers to a sensor that measures thickness. Thickness measurement value refers to the measured value of thickness. Spatial coordinates refer to the coordinates of the location. Sampling timestamp refers to the time point of sampling. Thickness fluctuation signal dataset refers to the set of data on thickness fluctuations. Data segment refers to a segmented set of data. Thickness change amount refers to the numerical change in thickness. Measurement point spacing refers to the distance between measurement points. Thickness change trend value refers to the trend of thickness change. Time window refers to the time range of the data segment. Thickness change trend value set refers to the set of thickness change trends.

[0032] In this embodiment, the system first continuously collects thickness measurements at various locations on the tank surface using thickness sensors arranged in a ring around the spinning forming device. Multiple thickness sensors (such as ultrasonic thickness probes) are installed around the spinning forming device, evenly distributed along the circumference of the tank to ensure coverage of all critical areas on the tank surface. When the spinning spindle rotates the billet, each sensor acquires the thickness measurement value (reflecting the local wall thickness of the tank) at a fixed sampling frequency in real time at the contact point. The coordinate recording module synchronously records the spatial coordinates (radial, axial, and circumferential positions) of each measurement point in three-dimensional space, as well as a precise sampling timestamp. The data integration unit binds the thickness values ​​with the corresponding spatial location and time information to form a thickness fluctuation signal dataset, providing raw input for subsequent time-series analysis.

[0033] Subsequently, the system divides the thickness fluctuation signal dataset into multiple data segments according to a fixed time window: the signal segmentation engine cuts the continuous thickness fluctuation signal dataset into multiple data segments based on a preset sampling period. Each data segment contains the continuous thickness measurements of all measurement points within that time window and their corresponding spatial coordinates. The segmentation process uses timestamp alignment technology to ensure that the time within the data segments is continuous and non-overlapping, forming a discretized dataset based on time windows, which facilitates the analysis of thickness fluctuation characteristics across different time periods.

[0034] Next, the system calculates the thickness change between adjacent measurement points and obtains the distance between measurement points based on the thickness measurements of the data segment. The change calculator iterates through the thickness data in each data segment, extracts the thickness measurements of two adjacent measurement points in spatial coordinate order, and generates the thickness change (quantifying the thickness difference between adjacent positions) through an arithmetic operation of subtracting the previous point value from the later point value. The distance calculation module calculates the actual physical distance between two points as the distance value based on the spatial coordinates of adjacent points (such as the three-dimensional Euclidean distance calculation formula). This process uses spatial interpolation technology to ensure that adjacent points are clearly defined and avoids distance calculation errors caused by coordinate jumps.

[0035] Then, the system generates thickness change trend values ​​based on the thickness change and the distance between measurement points, and arranges the set according to time windows: the trend generator divides each thickness change by its corresponding distance to obtain the thickness change rate per unit distance, i.e., the thickness change trend value (reflecting the local gradient intensity of the tank surface thickness). The sequence arranger arranges the thickness change trend values ​​in all data segments in ascending order according to the time window sequence, generating a set of thickness change trend values ​​corresponding to each window. This process uses time-series indexing technology to ensure a strict correspondence between trend values ​​and time windows, forming a thickness gradient evolution sequence in the time dimension.

[0036] Finally, the system integrates the thickness change trend values ​​from all time windows to construct a thickness gradient tensor: the tensor construction engine extracts the thickness change trend values ​​from all time windows and stacks them into a three-dimensional matrix structure according to the time window index. The tensor formatting module defines the dimension of this matrix as "number of time windows × number of measurement points per window × 1" (the third dimension is the single-channel trend value), and outputs a thickness gradient tensor. This tensor fully encapsulates the gradient features of the tank surface thickness in spatial distribution and temporal evolution, and can be directly input into deep learning models or quality control algorithms for anomaly detection.

[0037] In practical applications, in the scenario of intelligent thickness control during the spinning process of fire extinguisher canisters, on the fire extinguisher canister spinning production line, the system first continuously collects thickness measurements at various positions on the rotating canister surface using thickness sensors (such as non-contact laser thickness gauges) arranged in a ring on the spinning device. Simultaneously, it records the spatial coordinates of each measurement point (the canister's rotation angle and axial displacement position fed back by the encoder) and the sampling timestamp (millisecond-level time stamp). These three are combined to form a thickness fluctuation signal dataset (this dataset stores the dynamic correlation information of space-time-thickness values ​​in a three-dimensional matrix) (step 1011). Subsequently, the system divides the continuous thickness fluctuation signal dataset into multiple data segments according to a fixed time window. Each data segment contains the thickness measurement values ​​at consecutive time points within that window and their corresponding spatial coordinates (step 1012). Next, for each data segment, the system calculates the thickness change between adjacent measurement points based on the thickness measurement values ​​(such as the thickness difference between adjacent laser measurement points along the canister's circumference) and obtains the distance between measurement points (calculated by spatial coordinate difference between axial and circumferential distances) (step 1013). Based on the thickness change and the distance between measurement points, the system generates a thickness change trend value characterizing the local deformation strength (this value quantifies the rate of thickness change per unit distance), and arranges all thickness change trend values ​​in the order of the time window of the data segment, so that each time window corresponds to a set of thickness change trend values ​​(the set contains the distribution of trend values ​​of all measurement areas under the current window) (step 1014). Finally, the system integrates the set of thickness change trend values ​​of all time windows, and constructs a thickness gradient tensor of the three-dimensional structure with spatial coordinates as row indexes, time windows as column indexes, and trend values ​​as tensor elements (this tensor dynamically maps the thickness evolution law of the tank material flow throughout the spinning process) (step 1015). This thickness gradient tensor is input into the intelligent control system to predict the risk of material accumulation in real time: when the thickness change trend value of a specific area is detected to be continuously high, the rotary feed speed is automatically adjusted to balance the deformation and avoid strength failure caused by local thinning of the tank.

[0038] The scheme described in step 101 above enables multi-dimensional dynamic monitoring and gradient analysis of tank wall thickness fluctuations during the spinning process. Through real-time data acquisition from a ring-shaped thickness sensor array, an innovative dataset of thickness fluctuation signals, including spatial coordinates and timestamps, is constructed. This technology employs a combination of time window segmentation and adjacent point difference calculation to transform the raw thickness data into a gradient tensor reflecting local variation trends, overcoming the limitations of traditional single-point thickness measurement. The innovative spatial-temporal dual analysis framework achieves dynamic quantitative characterization of thickness distribution features, providing a high-precision data foundation for process optimization.

[0039] 102. Obtain the thickness fluctuation intensity value in the thickness gradient tensor, and obtain the temporal lag of the thickness change between adjacent scan points to calculate the phase shift characteristics.

[0040] Optionally, step 102 may specifically include the following steps: 1021. Extract the average value of the thickness change trend value in the thickness gradient tensor, and calculate the difference between the thickness change trend value at each spatial location point and the average value as the thickness fluctuation intensity value.

[0041] 1022. Locate adjacent spatial points in the thickness gradient tensor and record the time difference between the thickness change trend values ​​of adjacent spatial points exceeding a preset threshold as the time lag.

[0042] 1023. Obtain the sampling time interval of the thickness fluctuation signal, and generate a phase offset value based on the time lag and the sampling time interval.

[0043] 1024 Integrate the phase offset values ​​of all adjacent spatial locations and arrange the phase offset values ​​according to their spatial relationship to form a phase offset feature.

[0044] In the above scheme, thickness fluctuation intensity refers to the intensity of thickness fluctuation. Temporal lag refers to the time delay. Phase offset characteristic refers to the characteristic of phase offset. Average value refers to the average of the values. Difference refers to the difference between two values. Spatial location point refers to a location point in space. Preset threshold refers to the critical value for judging anomalies. Time point difference refers to the difference between time points. Sampling time interval refers to the sampling time interval. Phase offset value refers to the phase offset. Spatial positional relationship refers to the correlation between spatial positions.

[0045] In this embodiment, the system first extracts the average value of the thickness change trend value from the thickness gradient tensor: the data parsing engine traverses the thickness gradient tensor generated in step 101 (containing a set of thickness change trend values ​​arranged in time window order), and performs an arithmetic average of the thickness change trend values ​​of each spatial location point across all time windows to generate a global average value. The fluctuation intensity calculator then subtracts the thickness change trend value of the current time window from this global average value for each spatial location point to obtain a thickness fluctuation intensity value that quantifies the local thickness fluctuation amplitude (positive values ​​indicate fluctuations higher than the average, and negative values ​​indicate fluctuations lower than the average), forming a fluctuation intensity dataset corresponding one-to-one with each spatial location.

[0046] Subsequently, the system locates adjacent spatial points in the thickness gradient tensor and records the temporal lag: the neighborhood localization module identifies pairs of adjacent spatial points in the thickness gradient tensor based on spatial coordinate topological relationships (such as mesh structure or point cloud adjacency relationships). A threshold comparator monitors the thickness change trend of each pair of adjacent points in real time. When the value exceeds a preset fluctuation threshold (such as the material yield threshold), a timestamp recorder marks the exact time when the two points reach the threshold. The lag calculation unit subtracts the timestamps of the two points and outputs a temporal lag describing the fluctuation propagation delay, providing a time difference basis for phase analysis.

[0047] Next, the system acquires the sampling time interval of the thickness fluctuation signal and generates a phase offset value: the sampling configurator extracts the sampling time interval of the thickness fluctuation signal from the original acquisition parameters in step 101. The phase converter divides the timing lag obtained in step 1022 by the sampling time interval, converting it into a dimensionless delay step in units of sampling period. The normalization processor further multiplies this step number by a fixed angle coefficient, finally outputting a normalized phase offset value, realizing the mapping from time domain delay to angular phase difference.

[0048] Finally, the system integrates the phase offset values ​​of all adjacent spatial locations to form a phase offset feature: the spatial integration engine collects the phase offset values ​​of all adjacent spatial locations and sorts them according to their spatial coordinate relationships (such as distribution along the tank axis or circumference). The topology mapper reassembles the sorted phase offset values ​​according to the original spatial grid structure to construct a two-dimensional or three-dimensional phase offset feature matrix (e.g., a phase distribution heatmap with circumferential position as the horizontal axis and axial position as the vertical axis). This feature matrix encapsulates the asynchronous propagation law of thickness fluctuations in space and can be directly input into the subsequent servo compensation system.

[0049] In practical applications, in the scenario of analyzing thickness fluctuations during the force-spinning process of fire extinguisher tanks, the system, based on the thickness gradient tensor constructed in step 101 (which integrates the thickness variation trend values ​​at various spatial locations on the tank surface), first extracts the average value of the thickness variation trend values ​​in the thickness gradient tensor (calculated globally through the spatial and temporal dimensions of the tensor elements), and calculates the difference between the thickness variation trend value at each spatial location and this average value (this difference quantifies the degree to which the thickness fluctuation in a local area deviates from the overall average level), as the thickness fluctuation intensity value (this value highlights the risk area of ​​local thickening or thinning caused by abnormal material flow); then, it locates adjacent spatial locations in the thickness gradient tensor (e.g., equally spaced along the tank generatrix direction). The system uses a laser measuring point group to record the time difference between adjacent spatial locations where the thickness change trend exceeds a preset threshold, serving as a time lag (this value characterizes the propagation delay effect of material deformation fluctuations). Next, it acquires the sampling time interval of the thickness fluctuation signal (determined by the pulse period of the spinning machine's spindle encoder) and generates a phase offset value based on the time lag and sampling time interval (e.g., dividing the time lag by the sampling time interval to convert it into the phase angle difference of material deformation fluctuations). Finally, it integrates the phase offset values ​​of all adjacent spatial locations and arranges them according to their spatial relationship (constructing a two-dimensional matrix with the tank's axial height as the ordinate and the circumferential angle as the abscissa) to form a phase offset feature (this feature maps the spatial distribution law of material flow lag during spinning). This phase offset feature is input into the spindle trajectory adaptive control system. When a high phase offset value is detected in a specific area (e.g., the phase offset value in the rounded corner transition area at the bottom of the tank is significantly higher than that in the generatrix area), the spindle feed speed is dynamically adjusted to balance the material flow and avoid strength failure caused by localized thinning.

[0050] The scheme described in step 102 above achieves accurate analysis of the phase characteristics of thickness fluctuation features. Based on statistical analysis of the thickness gradient tensor, an innovative method for calculating fluctuation intensity and hysteresis was designed. This technique accurately captures the phase propagation characteristics of thickness changes through temporal difference analysis of adjacent points. An innovative method for constructing spatial phase shift features transforms abstract temporal hysteresis into quantifiable spatial distribution parameters, providing a new analytical dimension for understanding the spatiotemporal evolution of material flow. This technical approach, which correlates time delay with spatial location, significantly improves the comprehensiveness of thickness fluctuation characteristic analysis.

[0051] 103. Based on the thickness fluctuation intensity value and phase offset characteristics, dynamically adjust the curvature of the laser scanning lens arranged in a ring on the tank, and perform focused scanning on the surface of the tank according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map, wherein the three-dimensional wall thickness cloud map identifies areas with abnormal thickness.

[0052] Optionally, step 103 may specifically include the following steps: 1031. The thickness fluctuation intensity value is combined with the phase offset value in the phase offset feature to generate the curvature adjustment value of the laser scanning lens.

[0053] 1032. Adjust the hydraulic regulator pressure of the laser scanning lens according to the curvature adjustment value of the laser scanning lens so that the radius of curvature of the laser scanning lens is changed to the target value.

[0054] 1033. The adjusted laser scanning lens performs a focused scan on the surface of the tank. During the focused scan, the laser beam emitted by the laser scanning lens is incident perpendicularly on the surface of the tank and then reflected and captured by the receiver.

[0055] 1034. Calculate the time difference between the emission of the laser beam and its capture by the receiver, and convert the time difference into the thickness value of each scanning point on the surface of the tank.

[0056] 1035. Integrate the thickness values ​​of all scan points and arrange the thickness values ​​according to spatial coordinates to form a three-dimensional wall thickness cloud map, and mark the areas of continuous thickness anomalies in the three-dimensional wall thickness cloud map with thickness values ​​lower than a preset thickness threshold as thickness anomaly areas.

[0057] Specifically, step 1035 may include the following processes: extracting the thickness values ​​of all scan points and their corresponding spatial coordinates, and marking the thickness value of each scan point on the corresponding spatial coordinates, so that the spatial coordinates and the thickness values ​​form a corresponding mapping relationship; arranging all spatial coordinate points with marked thickness values ​​according to the grid order of the spatial coordinates to construct a three-dimensional point matrix structure, and integrating all three-dimensional point matrix structures to form a three-dimensional wall thickness cloud map; identifying thickness anomaly points with thickness values ​​lower than a preset thickness threshold in the three-dimensional wall thickness cloud map, and connecting consecutive adjacent thickness anomaly points to form a closed region as a thickness anomaly region.

[0058] In the above scheme, the laser scanning lens refers to the lens used for laser scanning. The 3D wall thickness cloud map refers to a graphic reflecting the wall thickness distribution. The thickness anomaly region refers to the region with abnormal thickness. The laser scanning lens curvature adjustment value refers to the numerical value used to adjust the lens curvature. The hydraulic regulator pressure value refers to the numerical value used to adjust the pressure. The target value refers to the set target value. Focused scanning refers to the scanning process of laser focusing. The laser beam refers to the laser beam. The receiver refers to the device that receives the laser. The time difference refers to the time difference between laser emission and reception. The scanning point refers to the location point of the laser scan. The thickness value refers to the numerical value of the thickness. Spatial coordinates refer to the coordinates of the location. The preset thickness threshold refers to the critical value for judging thickness anomalies. The thickness anomaly point refers to the location point of the thickness anomaly. The closed region refers to a closed spatial region. The 3D lattice structure refers to the 3D lattice structure. The grid order refers to the arrangement order of the grid. The mapping relationship refers to the correspondence between coordinates and numerical values.

[0059] In this embodiment, the system first combines the thickness fluctuation intensity value with the phase shift value in the phase shift feature: the parameter fusion engine extracts the thickness fluctuation intensity value (reflecting the local thickness change amplitude of the tank) generated in step 102 and the phase shift value (describing the hysteresis angle of thickness fluctuation propagation in space) from the phase shift feature, and fuses the two types of parameters according to a preset ratio using a weighted superposition algorithm. The curvature calculation module matches the target curvature parameter in the curvature mapping table (pre-stored curvature parameters corresponding to different combinations of thickness fluctuation and phase shift) based on the fusion result, generating a laser scanning lens curvature adjustment value (scalar value) for controlling lens deformation, providing input commands for hydraulic adjustment.

[0060] Subsequently, the system adjusts the hydraulic regulator pressure based on the curvature adjustment value of the laser scanning lens: the pressure converter receives the curvature adjustment value of the laser scanning lens and calculates the target pressure value through a linear proportional relationship (a positive correlation between the curvature adjustment value and the hydraulic pressure). The hydraulic actuator drives the piston displacement of the hydraulic regulator, changing the volume of hydraulic oil in the sealed cavity, causing the flexible film of the liquid lens to deform under the pressure difference. The curvature feedback sensor monitors the film's curvature radius in real time, and the pressure is fine-tuned through a closed-loop control algorithm until the actual curvature radius matches the target value, ensuring that the lens focal length matches the current scanning requirements.

[0061] Next, the system uses an adjusted laser scanning lens to perform a focused scan of the tank surface: the laser emission module projects a focused laser beam through the deformed laser scanning lens, and the optical calibration unit ensures that the beam is incident on the rotating tank surface at a perpendicular angle. The reflection capture system receives the laser signal reflected from the tank surface through a high-speed photoelectric sensor, and the synchronization controller coordinates the timing of laser pulse emission and receiver sampling to ensure that the reflection signal at each scanning point is accurately captured, forming a complete surface coverage scan trajectory.

[0062] Then, the system calculates the time difference between the laser beam's emission and capture and converts it into a thickness value: the time measurement circuit records the time difference between the laser pulse emission moment and the moment the receiver captures the reflected signal (i.e., the time of flight). Based on the principle of the constant speed of light, the thickness solver multiplies the time difference by the speed of light and divides it by two (to calculate the round-trip distance). Combining this with the laser incident angle and the theoretical thickness reference value of the tank, a geometric compensation algorithm is used to eliminate measurement errors, ultimately outputting the thickness value (the deviation between the actual thickness and the standard value) for each scanning point on the tank surface.

[0063] Finally, the system integrates the thickness values ​​to form a 3D wall thickness cloud map and marks abnormal areas: the point cloud construction engine binds the thickness values ​​of all scanned points to their spatial coordinates (the axial / circumferential positions of the tank recorded by the rotary encoder), fills the sampling gaps through a spatial grid interpolation algorithm, and generates a continuous 3D point matrix model. The anomaly detection module traverses the thickness values ​​in the point matrix, clusters adjacent points below a preset thickness threshold into continuous regions, and the marker renderer covers these thickness anomaly regions (such as material thinning or potential crack areas) with highlight blocks, outputting an interactively analyzeable 3D wall thickness cloud map.

[0064] In practical applications, during the real-time wall thickness monitoring of fire extinguisher tanks in the force-spinning process, the system generates thickness fluctuation intensity values ​​(characterizing the risk of local material accumulation or thinning) and phase shift characteristics (mapping the spatial distribution of material flow hysteresis) based on dynamic deformation data from the spinning process. First, these two values ​​are combined to generate a laser scanning lens curvature adjustment value (this value quantifies the dynamic compensation requirements of the tank surface deformation for laser focusing accuracy). Then, based on this adjustment value, the hydraulic regulator of the laser scanning lens is driven, adjusting its pressure to change the lens curvature radius to the target value (the hydraulic pressure change drives the lens surface deformation, adapting to the curvature changes of the high-speed rotating tank). Next, the system activates the adjusted laser scanning lens to perform a focused scan of the tank surface: the laser beam emitted by the lens is projected onto the rotating tank surface in a perpendicular manner, and the reflected beam is captured by a ring of highly sensitive receivers. By calculating the time difference between the laser beam's emission and capture by the receiver (this time difference is linearly related to the laser's round-trip distance), the time difference is converted into the thickness value at each scanning point on the tank surface (based on the principle of constant light speed and material refractive index calibration). Finally, the system extracts the thickness values ​​of all scan points and their corresponding spatial coordinates, marking the thickness value of each scan point on its corresponding spatial coordinates to form a mapping relationship. All spatial coordinate points with marked thickness values ​​are arranged according to the grid order of the spatial coordinates to construct a three-dimensional point matrix structure covering the entire area, and integrated into a three-dimensional wall thickness cloud map (the depth axis of the cloud map corresponds to the thickness value, and the planar axis corresponds to the latitude and longitude coordinates of the tank). Thickness anomalies with thickness values ​​below a preset thickness threshold are identified in the cloud map. Adjacent thickness anomalies are connected to form closed boundaries and marked as thickness anomaly regions (such as circumferential banded areas caused by excessive thinning by the spinning wheel). Data from these anomaly regions is fed back to the spinning control system in real time, triggering spinning wheel trajectory compensation to suppress uneven wall thickness.

[0065] The scheme described in step 103 above achieves adaptive optical scanning and 3D visualization for tank wall thickness detection. An innovative intelligent focusing scanning system is constructed through dynamic adjustment of lens curvature driven by wave characteristics. This technology employs a precise hydraulic pressure control mechanism to achieve real-time optimization of laser beam focusing parameters. Innovative time difference and thickness conversion algorithms, along with a 3D dot matrix modeling method, generate high-precision wall thickness distribution cloud maps and enable automatic identification of abnormal areas. This technical solution, which intelligently couples detection parameters with the optical system, significantly improves the accuracy and efficiency of wall thickness detection.

[0066] 104. Obtain the rotation trajectory of the tank's wheel, and based on the thickness anomaly region and the phase offset feature, correlate the spatial deviation between the thickness change and the rotation trajectory to generate a trajectory compensation vector for correcting the rotation trajectory.

[0067] Optionally, step 104 may specifically include the following steps: 1041. A spatial coordinate sequence of the motion trajectory of a rotating wheel is obtained by a rotating wheel trajectory recording device, wherein the spatial coordinate sequence includes the three-dimensional position coordinates of the center point of the rotating wheel at each time point.

[0068] 1042. Locate the spatial coordinates of each abnormal point within the thickness abnormality area, and perform spatial position matching between the spatial coordinates of the thickness abnormal points and the spatial coordinate sequence of the wheel motion trajectory to obtain the matching trajectory points.

[0069] 1043. Calculate the spatial straight-line distance between the thickness anomaly point and the matching trajectory point, and combine the spatial straight-line distance with the phase offset value to generate a spatial deviation correction component.

[0070] 1044. Determine the offset direction of the wheel motion trajectory based on the spatial deviation correction component, wherein the offset direction is determined by the direction of the line connecting the thickness anomaly point to the matching trajectory point.

[0071] 1045. Integrate the spatial deviation correction components and offset directions corresponding to all thickness anomaly points to form a trajectory compensation vector describing the trajectory of the rotating wheel.

[0072] In the above scheme, the wheel motion trajectory refers to the motion path of the wheel. The trajectory compensation vector refers to the vector that corrects the trajectory. The wheel trajectory recording device refers to the equipment that records the wheel trajectory. The spatial coordinate sequence refers to spatial coordinates arranged in order. The three-dimensional position coordinates refer to coordinates in three-dimensional space. An anomaly point refers to the location point of thickness anomaly. A matching trajectory point refers to the trajectory point that matches the anomaly point. The spatial straight-line distance refers to the straight-line distance between two points. The spatial deviation correction component refers to the component that corrects the spatial deviation. The offset direction refers to the direction of trajectory offset. The connecting line direction refers to the direction of the line connecting the two points.

[0073] In this embodiment, the system first acquires the spatial coordinate sequence of the wheel's motion trajectory through a wheel trajectory recording device. The trajectory recording device (such as an optical encoder or laser tracking system mounted on the wheel support) collects the three-dimensional position coordinates (axial, radial, and circumferential) of the wheel's center point at each sampling moment in real time. The coordinate integration module organizes this position data in timestamp order, forming a spatial coordinate sequence containing the wheel's motion trajectory at consecutive time points (e.g., time T1 corresponds to coordinates (x1, y1, z1), and T2 corresponds to (x2, y2, z2)). This sequence completely encapsulates the wheel's motion path during the tank forming process, providing a reference for subsequent spatial matching.

[0074] Subsequently, the system locates the spatial coordinates of each anomaly point within the thickness anomaly region and performs position matching: the anomaly point localization engine extracts the spatial coordinates of all discrete points within the thickness anomaly region from the 3D wall thickness contour map generated in step 103. The spatial mapper overlays these anomaly point coordinates onto the path grid composed of the spatial coordinate sequence of the wheel's motion trajectory, and assigns the nearest trajectory point to each anomaly point using a nearest neighbor matching algorithm, generating a one-to-one matching trajectory point. This process establishes a direct spatial association between the thickness defect and the wheel's position.

[0075] Next, the system calculates the spatial straight-line distance between thickness anomaly points and matching trajectory points and generates a spatial deviation correction component: the distance solver traverses each thickness anomaly point and its matching trajectory point, calculating the spatial straight-line distance between the two points using the three-dimensional Euclidean distance formula (quantizing the deviation amplitude between the actual position of the rotating wheel and the defect point). The component generator multiplies the distance value with the corresponding position phase offset value obtained in step 102 (describing the propagation hysteresis of thickness fluctuations) to generate a spatial deviation correction component (scalar value) that integrates spatial deviation and dynamic phase characteristics. This component simultaneously reflects the combined effect of static position offset and dynamic fluctuation delay.

[0076] Then, the system determines the offset direction of the wheel's motion trajectory based on the spatial deviation correction component: the direction vector calculation module extracts the line vector connecting the thickness anomaly point to the matching trajectory point (e.g., from point P). a Point to P t The vector V→(Δx,Δy,Δz) is normalized by the normalization processor to a unit direction vector, defined as the offset direction (e.g., V→(0.6,0,0.8) represents an axial offset of 60% and a circumferential offset of 80%). This direction indicates the ideal movement path the wheel needs to take from its current position to the defect point.

[0077] Finally, the system integrates all spatial deviation correction components and offset directions to form a trajectory compensation vector: the vector field construction engine collects the spatial deviation correction components (scalars) and offset directions (unit vectors) corresponding to all thickness anomaly points, and generates a correction vector (component value × direction vector) for each anomaly point through scalar-vector multiplication. The tensor integrator arranges these vectors according to a spatial grid structure to construct a trajectory compensation vector field covering the entire surface of the tank. This vector field can be directly input into the CNC system to guide the rotating wheel to move along the correction path to eliminate thickness anomalies.

[0078] In practical applications, in the scenario of thickness compensation control during the spinning of fire extinguisher canister heads, the system, based on a three-dimensional wall thickness cloud map (which identifies areas of abnormal thickness) and phase offset characteristics (characterizing the hysteresis distribution of material flow), first acquires the spatial coordinate sequence of the spinning wheel's motion trajectory through a spinning wheel trajectory recording device (such as an encoder and laser displacement sensor system integrated into the spinning machine spindle). This sequence accurately records the three-dimensional position coordinates of the spinning wheel's center point at each time point, forming a spatiotemporal mapping relationship. Subsequently, the spatial coordinates of each abnormal point within the thickness abnormal area are located (such as the coordinates of thinning points formed by material accumulation in the head transition area). The spatial coordinates of these thickness abnormal points are then spatially matched with the spatial coordinate sequence of the spinning wheel's motion trajectory (through spatiotemporal coordinate system projection alignment) to obtain the matched trajectory points (i.e., the actual trajectory of the spinning wheel). The system first calculates the location points corresponding to the thickness anomalies; then it calculates the spatial straight-line distance between the thickness anomalies and the matching trajectory points (this distance quantifies the geometric deviation between the actual path of the spinning wheel and the ideal forming position), and combines this distance with the phase offset value of the corresponding region (from the phase offset feature matrix) to generate a spatial deviation correction component (this component integrates the effects of geometric deviation and material hysteresis); then it determines the offset direction of the spinning wheel's motion trajectory based on the spatial deviation correction component (this direction is defined by the vector direction from the thickness anomaly point to the matching trajectory point, indicating the motion trend of the spinning wheel that needs to be compensated in the opposite direction); finally, it integrates the spatial deviation correction components and offset directions corresponding to all thickness anomalies to construct a global correction model in the form of a three-dimensional vector field, forming a trajectory compensation vector that describes the global control requirements of the spinning wheel's motion trajectory (this vector provides a basis for dynamic path compensation for the spinning machine's CNC system). This trajectory compensation vector drives the spinning wheel feed system in real time: when a significant increase in the amplitude of the trajectory compensation vector in the rounded corner area of ​​the end cap is detected, the radial feed of the spinning wheel is automatically adjusted to suppress local thinning caused by material flow hysteresis and ensure the uniformity of the wall thickness in the tank end cap area.

[0079] The scheme described in step 104 above achieves intelligent correlation analysis and compensation vector generation between the wheel trajectory and thickness anomalies. Through spatial position matching and deviation calculation, an innovative quantitative relationship model between process parameters and quality defects is established. The deviation correction component calculation method designed in this technology comprehensively considers the dual influence of spatial distance and phase characteristics. The innovative vector integration algorithm transforms discrete correction requirements into a systematic trajectory compensation scheme, providing clear direction and magnitude guidance for process adjustments. This data-driven compensation mechanism achieves a closed-loop connection from quality inspection to process correction.

[0080] 105. Based on the trajectory compensation vector, dynamically optimize the rotary feed path to suppress wall thickness fluctuations caused by material springback and achieve intelligent control of tank thickness.

[0081] Optionally, step 105 may specifically include the following steps: 1051. Analyze the spatial deviation correction component and offset direction in the trajectory compensation vector, and combine the spatial deviation correction component with the preset material springback compensation coefficient to generate the path correction amount.

[0082] 1052. Determine the adjustment direction of the rotary feed path according to the offset direction, wherein the adjustment direction is consistent with the offset direction.

[0083] 1053. Obtain the spatial coordinate sequence of the current feed path of the rotary wheel, and superimpose the path correction amount onto the spatial coordinate sequence according to the adjustment direction to form the updated feed path coordinates.

[0084] 1054. Convert the updated feed path coordinates into control signals for the rotary drive, and execute the control signals through the rotary drive to make the rotary wheel move along the updated feed path coordinates to suppress wall thickness fluctuations caused by material springback.

[0085] In the above scheme, dynamic optimization refers to the process of real-time adjustment and optimization. The rotary feed path refers to the feed route of the rotary wheel. Material springback refers to the recovery phenomenon of material deformation. Wall thickness fluctuation refers to the change in wall thickness. Intelligent control refers to an automated control method. Path correction amount refers to the adjustment amount of the path. Material springback compensation coefficient refers to the proportional coefficient for compensating for springback. Adjustment direction refers to the direction of path adjustment. Spatial coordinate sequence refers to spatial coordinates arranged in sequence. Updated feed path coordinates refer to the adjusted path coordinates. Rotary wheel driver refers to the device that drives the rotary wheel. Control signal refers to the signal that controls the device.

[0086] In this embodiment, the system first analyzes the spatial deviation correction component and offset direction in the trajectory compensation vector: the vector analysis module extracts the spatial deviation correction component (a scalar value describing the magnitude of the position deviation) and the offset direction (unit vector) recorded in the trajectory compensation vector generated in step 104. The compensation fusion unit combines this component value with a preset material rebound compensation coefficient (a correction factor calibrated based on the material elastic modulus experiment) through scalar multiplication to generate a path correction amount (vector value) that integrates the position deviation and material rebound characteristics. This process achieves coupled correction of spatial deviation and material rebound through a dynamic weighting algorithm, ensuring that the correction amount simultaneously reflects the influence of instantaneous position error and material elastic recovery, providing accurate input for path updating.

[0087] Subsequently, the system determines the adjustment direction of the rotary feed path based on the offset direction: the direction mapper directly uses the offset direction in the trajectory compensation vector and sets it as the adjustment direction of the rotary feed path. The direction locking unit ensures that this direction is completely consistent with the original vector pointing from the thickness anomaly point to the matching trajectory point through vector inheritance technology, avoiding the destruction of the spatial correlation between the adjustment direction and the defect position, thereby ensuring the physical rationality of the correction action.

[0088] Next, the system acquires the spatial coordinate sequence of the current feed path of the rotary wheel and superimposes the path correction: the path acquisition module reads the spatial coordinate sequence (such as an axial / radial coordinate array) of the current feed path of the rotary wheel from the CNC system. The coordinate correction engine traverses each coordinate point in the sequence, decomposes the path correction into three-dimensional displacement components (such as x, y, z axial displacements) according to the adjustment direction, and superimposes them onto the original coordinate values ​​through vector addition. The sequence updater outputs the updated feed path coordinates including the position correction, forming a new motion trajectory after compensating for the springback effect, causing the actual path of the rotary wheel to converge towards the thickness anomaly region.

[0089] Finally, the system converts the updated feed path coordinates into control signals to drive the rotary wheel: the signal converter discretizes the updated feed path coordinates into a pulse sequence of the stepper motor based on a spline interpolation algorithm, generating the control signal for the rotary wheel driver. The motion execution unit converts the signal into the rotation angle and speed of the motor shaft through the driver, driving the rotary wheel to feed precisely along the corrected path. The springback suppression module synchronously adjusts the rotary wheel pressure parameters to counteract the elastic restoring force after material unloading, significantly reducing the fluctuation range of the tank wall thickness and achieving intelligent closed-loop control of the thickness.

[0090] In practical applications, during the spin forming of fire extinguisher canister heads, specifically in the scenario of suppressing springback during the high-pressure spin forming process, the system, based on a trajectory compensation vector (containing a spatial deviation correction component and an offset direction), first analyzes the spatial deviation correction component and offset direction within the trajectory compensation vector (this component quantifies the geometric deviation between the actual trajectory of the spinning wheel and the ideal forming position, while the offset direction indicates the reverse compensation requirement of the spinning wheel path caused by material springback). The spatial deviation correction component is then combined with a preset material springback compensation coefficient (calibrated based on the high elasticity of the aluminum alloy of the canister head) to generate a path correction amount (this correction amount integrates the combined effects of geometric deviation and material elastic recovery). Subsequently, the adjustment direction of the spinning wheel feed path is determined based on the offset direction (this adjustment direction strictly maintains consistency with the offset direction to ensure that the compensation path and the material springback vector work in opposite directions). Finally, the spatial coordinates of the current spinning wheel feed path are obtained. The system reads data from the spindle encoder and displacement sensor in real time through the CNC system of the spinning machine, and superimposes the path correction amount onto the spatial coordinate sequence according to the adjustment direction (the correction amount is decomposed into axial and radial coordinate increments through a three-dimensional vector superposition algorithm) to form the updated feed path coordinates (this coordinate sequence dynamically corrects the spatial position of the spinning wheel trajectory and suppresses local thickening in the rounded corner area of ​​the end cap caused by material elastic recovery); finally, the updated feed path coordinates are converted into control signals for the spinning wheel drive (the CNC system parses the coordinate sequence into flow commands for the hydraulic servo valve), and the control signals are executed by the spinning wheel drive (driving the spinning wheel hydraulic cylinder to move along the updated path), so that the spinning wheel moves along the updated feed path coordinates to suppress wall thickness fluctuations caused by material rebound (such as uneven wall thickness after unloading in the transition area of ​​the end cap, which is forced to homogenize the material flow by the reverse feed of the compensation path, ensuring the consistency of wall thickness in the end cap area of ​​the tank).

[0091] The scheme described in step 105 above achieves intelligent dynamic optimization control of the spinning forming process. An innovative adaptive feed control system is constructed through the analysis of the compensation vector and path correction. This technology employs an intelligent combination of the material springback coefficient and the path correction amount to ensure the process rationality of the compensation scheme. An innovative coordinate sequence update mechanism achieves precise conversion of compensation commands into drive signals. This technical approach, which directly converts detection and analysis results into control commands, effectively suppresses thickness fluctuations caused by material springback and significantly improves the stability and consistency of the tank forming quality.

[0092] The following are specific examples of steps 101 to 105:

[0093] In the closed-loop control scenario of thickness measurement during the high-pressure spinning process of fire extinguisher cylinders, the system first continuously collects thickness measurements at various locations on the cylinder surface using thickness sensors (such as a laser thickness gauge array) arranged in a ring on the spinning device. Simultaneously, it records the spatial coordinates and sampling timestamps of each measurement point, forming a thickness fluctuation signal dataset. This dataset is then divided into multiple data segments according to a fixed time window. Each segment contains thickness measurements and corresponding spatial coordinates at consecutive time points. The system then calculates the thickness variation between adjacent measurement points within each data segment and generates a thickness variation trend value based on the distance between measurement points. All thickness variation trend values ​​are arranged sequentially according to the time window to construct a thickness gradient tensor. Next, the system extracts the average value of the thickness variation trend values ​​from the thickness gradient tensor and calculates the difference between the trend value at each spatial location and this average value as the thickness fluctuation intensity value. Adjacent spatial locations are located within the thickness gradient tensor, and the time difference between the time points when their thickness variation trend values ​​exceed a preset threshold is recorded as a time lag. This time lag is then combined with the sampling time interval of the thickness fluctuation signal to generate a phase offset value, and all phase offset values ​​are integrated according to their spatial relationships to form a phase offset feature.

[0094] Based on the aforementioned thickness fluctuation intensity value and phase shift characteristics, the system combines the two to generate a laser scanning lens curvature adjustment value. By adjusting the pressure value of the hydraulic regulator of the laser scanning lens, the lens curvature radius is changed to the target value. Subsequently, the adjusted laser scanning lens is activated to perform a focused scan on the tank surface, so that the laser beam is perpendicularly incident on the tank surface and the reflected beam is captured by the receiver. By calculating the time difference between the laser beam emission and reception, it is converted into the thickness value of each scanning point on the tank surface. Finally, the thickness values ​​and corresponding spatial coordinates of all scanning points are extracted, a three-dimensional lattice structure is constructed in grid order, and integrated into a three-dimensional wall thickness cloud map. In this cloud map, continuous abnormal points with thickness values ​​lower than the preset thickness threshold are identified and connected to form a thickness abnormality region.

[0095] Subsequently, the system acquires the spatial coordinate sequence of the spinning wheel's motion trajectory through a spinning wheel trajectory recording device (such as a spinning machine spindle encoder and displacement sensor). Within the thickness anomaly region, it locates the spatial coordinates of each anomaly point and matches them with the spinning wheel trajectory coordinate sequence to obtain matching trajectory points. It calculates the spatial straight-line distance between the thickness anomaly point and the matching trajectory point, generates a spatial deviation correction component based on the phase offset value, and determines the offset direction by the line connecting the thickness anomaly point to the matching trajectory point. It integrates the correction components and offset directions of all anomaly points to form a trajectory compensation vector. Finally, the system analyzes the spatial deviation correction component and offset direction in the trajectory compensation vector, combines the correction component with a preset material springback compensation coefficient to generate a path correction amount, and superimposes it onto the spatial coordinate sequence of the spinning wheel's current feed path in an adjustment direction consistent with the offset direction, forming updated feed path coordinates. This coordinate is converted into a control signal for the spinning wheel driver (such as a hydraulic servo valve command), driving the spinning wheel to move along the new path, thereby suppressing wall thickness fluctuations caused by material springback and achieving intelligent control of the tank thickness.

[0096] Figure 2 This application provides a schematic diagram of the structure of an intelligent control system for the thickness of a tank body during spinning, as shown in the embodiment of the present application. Figure 2 As shown, the system includes: Separation module 21 is used to collect the thickness fluctuation signal on the surface of the tank during the spinning process, and to separate the thickness fluctuation signal through time-series decomposition technology to generate a thickness gradient tensor. The calculation module 22 is used to obtain the thickness fluctuation intensity value in the thickness gradient tensor and obtain the temporal lag of the thickness change between adjacent scan points to calculate the phase offset characteristics. The adjustment module 23 is used to dynamically adjust the curvature of the laser scanning lens that is arranged in a ring on the rotating wheel of the tank according to the thickness fluctuation intensity value and phase offset characteristics, and to perform focused scanning on the surface of the tank according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map, wherein the three-dimensional wall thickness cloud map identifies areas with abnormal thickness. The generation module 24 is used to obtain the rotation trajectory of the tank and, based on the thickness anomaly region and the phase offset feature, associate the spatial deviation between the thickness change and the rotation trajectory to generate a trajectory compensation vector for correcting the rotation trajectory. The optimization module 25 is used to dynamically optimize the rotary feed path according to the trajectory compensation vector, so as to suppress the wall thickness fluctuation caused by material springback and realize intelligent control of the tank thickness.

[0097] Figure 2 The aforementioned intelligent control system for tank body spinning thickness can perform... Figure 1The implementation principle and technical effects of the intelligent control method for the thickness of tank spinning forming described in the above embodiment will not be repeated here. The specific operation methods of each module and unit in the intelligent control system for tank spinning forming thickness in the above embodiment have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0098] In one possible design, Figure 2 The intelligent control system for the thickness of a tank body in the illustrated embodiment can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0099] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0100] The processing component 32 is used for the above Figure 1 The embodiment describes an intelligent control method for the thickness of a tank body during spinning.

[0101] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0102] Storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0103] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0104] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0105] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0106] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0107] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The embodiment shown illustrates a method for intelligent control of the thickness of a tank body during spinning.

[0108] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0109] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0110] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for intelligent control of thickness during tank spinning, characterized in that, include: The thickness fluctuation signal on the surface of the tank during the spinning process is collected, and the thickness fluctuation signal is separated by time-series decomposition technology to generate a thickness gradient tensor; Obtain the thickness fluctuation intensity value in the thickness gradient tensor, and obtain the temporal lag of thickness change between adjacent scan points to calculate phase offset characteristics; Based on the thickness fluctuation intensity value and phase shift characteristics, the curvature of the laser scanning lens arranged in a ring on the tank is dynamically adjusted, and the surface of the tank is focused and scanned according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map, which identifies areas with abnormal thickness. The rotation trajectory of the tank is obtained, and based on the thickness anomaly region and the phase offset feature, the spatial deviation between the thickness change and the rotation trajectory is correlated to generate a trajectory compensation vector for correcting the rotation trajectory. Based on the trajectory compensation vector, the rotary feed path is dynamically optimized to suppress wall thickness fluctuations caused by material springback and achieve intelligent control of tank thickness.

2. The method according to claim 1, characterized in that, The thickness fluctuation signal on the surface of the tank during the spinning process is collected, and the thickness fluctuation signal is separated by time-series decomposition technology to generate a thickness gradient tensor, including: Thickness measurements are continuously collected at various locations on the surface of the tank by thickness sensors arranged in a ring on the spinning forming device. At the same time, the spatial coordinates and sampling timestamps of each measurement point are recorded. The thickness measurements, spatial coordinates and sampling timestamps are combined to form a thickness fluctuation signal dataset. The thickness fluctuation signal dataset is divided into multiple data segments according to a fixed time window. Each data segment contains thickness measurement values ​​at consecutive time points and their corresponding spatial coordinates. The thickness variation between adjacent measurement points is calculated based on the thickness measurement value of the data segment, and the spacing value of the measurement points is obtained. Based on the thickness change and the distance between measurement points, a thickness change trend value is generated, and the thickness change trend value is arranged in the order of the time window of the data segment so that each time window corresponds to a set of thickness change trend values. Integrate the set of thickness change trend values ​​from all time windows to construct the thickness gradient tensor.

3. The method according to claim 1, characterized in that, Obtain the thickness fluctuation intensity value in the thickness gradient tensor, and obtain the temporal hysteresis of thickness change between adjacent scan points to calculate phase shift characteristics, including: Extract the average value of the thickness change trend value in the thickness gradient tensor, and calculate the difference between the thickness change trend value at each spatial location point and the average value as the thickness fluctuation intensity value; In the thickness gradient tensor, adjacent spatial locations are located, and the time difference between the thickness change trend values ​​of adjacent spatial locations exceeding a preset threshold is recorded as the time lag. The sampling time interval of the thickness fluctuation signal is obtained, and a phase offset value is generated based on the time lag and the sampling time interval. The phase offset values ​​of all adjacent spatial locations are integrated, and the phase offset values ​​are arranged according to their spatial positional relationship to form a phase offset feature.

4. The method according to claim 1, characterized in that, Based on the thickness fluctuation intensity value and phase shift characteristics, the curvature of the laser scanning lens, which is arranged in a ring on the rotating wheel of the tank, is dynamically adjusted. The adjusted laser scanning lens then performs a focused scan of the tank surface to output a three-dimensional wall thickness cloud map. This three-dimensional wall thickness cloud map identifies areas of abnormal thickness, including: The thickness fluctuation intensity value is combined with the phase shift value in the phase shift feature to generate the curvature adjustment value of the laser scanning lens; Adjust the hydraulic regulator pressure of the laser scanning lens according to the curvature adjustment value so that the radius of curvature of the laser scanning lens is changed to the target value; The adjusted laser scanning lens performs a focused scan on the surface of the tank. During the focused scan, the laser beam emitted by the laser scanning lens is incident perpendicularly on the surface of the tank and then reflected and captured by the receiver. Calculate the time difference between the laser beam's emission and its capture by the receiver, and convert the time difference into the thickness value of each scanning point on the surface of the tank; The thickness values ​​of all scan points are integrated and arranged according to spatial coordinates to form a three-dimensional wall thickness cloud map. The areas of continuous thickness anomalies in the three-dimensional wall thickness cloud map with thickness values ​​lower than a preset thickness threshold are marked as thickness anomaly areas.

5. The method according to claim 4, characterized in that, The thickness values ​​of all scanned points are integrated and arranged according to spatial coordinates to form a three-dimensional wall thickness cloud map. Regions in the three-dimensional wall thickness cloud map with continuous thickness anomalies whose thickness values ​​are below a preset thickness threshold are marked as thickness anomaly regions, including: Extract the thickness values ​​of all scanning points and their corresponding spatial coordinates, and mark the thickness value of each scanning point on the corresponding spatial coordinates, so that the spatial coordinates and the thickness values ​​form a corresponding mapping relationship. Arrange all spatial coordinate points with labeled thickness values ​​according to the grid order of the spatial coordinates to construct a three-dimensional lattice structure, and integrate all three-dimensional lattice structures to form a three-dimensional wall thickness cloud map. In the three-dimensional wall thickness cloud map, thickness anomalies with thickness values ​​lower than a preset thickness threshold are identified, and consecutive adjacent thickness anomalies are connected to form a closed region as a thickness anomaly region.

6. The method according to claim 1, characterized in that, The trajectory of the rotating wheel in the tank is obtained, and based on the thickness anomaly region and the phase shift characteristics, the spatial deviation between the thickness change and the rotating wheel trajectory is correlated to generate a trajectory compensation vector for correcting the rotating wheel trajectory, including: A spatial coordinate sequence of the wheel's motion trajectory is obtained by a wheel trajectory recording device. The spatial coordinate sequence includes the three-dimensional position coordinates of the wheel's center point at each time point. The spatial coordinates of each abnormal point within the thickness abnormality area are located, and the spatial coordinates of the thickness abnormal points are matched with the spatial coordinate sequence of the wheel motion trajectory to obtain the matching trajectory points. Calculate the spatial straight-line distance between the thickness anomaly point and the matching trajectory point, and combine the spatial straight-line distance with the phase offset value to generate a spatial deviation correction component; The offset direction of the wheel motion trajectory is determined based on the spatial deviation correction component, and the offset direction is determined by the direction of the line connecting the thickness anomaly point to the matching trajectory point; The spatial deviation correction components and offset directions corresponding to all thickness anomalies are integrated to form a trajectory compensation vector describing the trajectory of the spinning wheel.

7. The method according to claim 1, characterized in that, Based on the trajectory compensation vector, the rotary feed path is dynamically optimized to suppress wall thickness fluctuations caused by material springback and achieve intelligent control of the tank thickness, including: The spatial deviation correction component and offset direction in the trajectory compensation vector are analyzed, and the spatial deviation correction component is combined with the preset material springback compensation coefficient to generate the path correction amount; The adjustment direction of the rotary feed path is determined based on the offset direction, and the adjustment direction is consistent with the offset direction; Obtain the spatial coordinate sequence of the current feed path of the rotary wheel, and superimpose the path correction amount onto the spatial coordinate sequence according to the adjustment direction to form the updated feed path coordinates; The updated feed path coordinates are converted into control signals for the rotary drive, and the control signals are executed by the rotary drive to make the rotary wheel move along the updated feed path coordinates to suppress wall thickness fluctuations caused by material springback.

8. A smart control system for the thickness of tank body spin forming, characterized in that, include: The separation module is used to collect the thickness fluctuation signal on the surface of the tank during the spinning process, and to separate the thickness fluctuation signal through time-series decomposition technology to generate a thickness gradient tensor. The calculation module is used to obtain the thickness fluctuation intensity value in the thickness gradient tensor and obtain the temporal lag of the thickness change between adjacent scan points to calculate the phase shift characteristics. The adjustment module is used to dynamically adjust the curvature of the laser scanning lens that is arranged in a ring on the rotating wheel of the tank according to the thickness fluctuation intensity value and phase offset characteristics, and to perform focused scanning on the surface of the tank according to the adjusted laser scanning lens to output a three-dimensional wall thickness cloud map, which identifies areas with abnormal thickness. The generation module is used to obtain the rotation trajectory of the tank's wheel, and based on the thickness anomaly region and the phase offset feature, associate the spatial deviation between the thickness change and the rotation trajectory to generate a trajectory compensation vector for correcting the rotation trajectory. The optimization module is used to dynamically optimize the rotary feed path based on the trajectory compensation vector, so as to suppress the wall thickness fluctuation caused by material springback and realize intelligent control of the tank thickness.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement the intelligent control method for the thickness of tank spinning as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The device contains a computer program that, when executed by a computer, implements a method for intelligent control of the thickness of a tank body during spinning as described in any one of claims 1 to 7.

Images