Processing control method and system for carbon fiber composite transmission shaft

Abstract

The invention discloses a processing control method and system for a carbon fiber composite transmission shaft, and relates to the technical field of intelligent transmission shaft manufacturing, and the method comprises the steps: dynamically adjusting the control current of a magneto-rheological damper through a PID controller based on a process control parameter set, and recording tension fluctuation and temperature gradient data; inputting tension fluctuation and temperature gradient data into an ANSYS coupling simulation engine, establishing a three-dimensional temperature field model, calculating a resin flow compensation coefficient, generating a temperature-pressure cooperative control curve through an NSGA-III multi-objective optimization algorithm based on the resin flow compensation coefficient, and calculating the resin flow compensation coefficient and the temperature-pressure cooperative control curve according to the resin flow compensation coefficient and the temperature-pressure cooperative control curve. And integrating through a time sequence control engine to generate a solidification control instruction. According to the invention, a multi-source data fusion algorithm is combined with the industrial Internet of Things edge computing gateway, so that accurate monitoring and synchronous processing of the process control parameter set are realized, and the dynamic behavior of the carbon fiber composite material in the processing process is comprehensively reflected.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transmission shaft manufacturing, and in particular to a processing control method and system for a carbon fiber composite material transmission shaft. Background Art

[0002] In recent years, carbon fiber composites have been widely used in aerospace, automotive, and high-end equipment manufacturing. As a core component in mechanical power transmission, the lightweighting and performance optimization design of drive shafts has been a hot topic of research. With the advancement of composite material processing technology, traditional metal drive shafts are gradually being replaced by carbon fiber composites to meet higher performance requirements. Currently, related technologies have evolved from single-parameter monitoring to multi-parameter coupled analysis. For example, three-dimensional temperature field modeling technology based on simulation tools such as ANSYS has been widely used in hot press forming processes.

[0003] Despite this, existing technologies still face numerous challenges in practical application. For one thing, traditional process control methods typically rely on single-sensor data for local adjustments and lack the ability to integrate and process multi-source data. Furthermore, during the curing process, existing temperature-pressure coordinated control often employs static or piecewise linear models, lacking a precise description of nonlinear dynamic characteristics, leading to uneven resin flow and curing defects. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a processing control method for a carbon fiber composite material transmission shaft to solve the problems of insufficient fusion of multi-source data and insufficient control accuracy of the curing process.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for controlling the processing of a carbon fiber composite drive shaft, comprising dynamically adjusting the control current of a magnetorheological damper by a PID controller based on a process control parameter set, and recording tension fluctuation and temperature gradient data;

[0008] The tension fluctuation and temperature gradient data were input into the ANSYS coupled simulation engine to establish a three-dimensional temperature field model and calculate the resin flow compensation coefficient. Based on the resin flow compensation coefficient, the NSGA-III multi-objective optimization algorithm was used to generate a temperature-pressure coordinated control curve.

[0009] The resin flow compensation coefficient and temperature-pressure coordinated control curve are integrated through the timing control engine to generate curing control instructions;

[0010] Based on the curing control instructions, the power output of the hot press furnace and the pressure distribution of the hydraulic equipment are dynamically adjusted, and strain energy data and curing monitoring data are simultaneously collected to generate a curing degree report;

[0011] Based on the curing degree report, the particle swarm optimization algorithm is used to generate the final process control parameter set to optimize the transmission shaft machining control process.

[0012] As a preferred solution of the processing control method of the carbon fiber composite material transmission shaft of the present invention, the specific steps of obtaining the process control parameter set are as follows:

[0013] The fiber bundle diameter, initial tension value, fiber delivery speed, and fiber surface temperature data are weightedly fused using a multi-source data fusion algorithm to obtain a composite sensing data set.

[0014] The composite sensor data set is subjected to feature dimensionality reduction through principal component analysis to generate a process parameter feature matrix;

[0015] Based on the process parameter feature matrix, the process control parameter set is generated through the process parameter optimization engine using the improved NSGA-III multi-objective optimization algorithm.

[0016] As a preferred solution of the processing control method of the carbon fiber composite material drive shaft of the present invention, wherein: the control current of the magnetorheological damper is dynamically adjusted by the PID controller, and the tension fluctuation and temperature gradient data are recorded. The specific steps are as follows:

[0017] Based on the process control parameter set, the improved Ziegler-Nichols critical proportion method is used to tune the parameters of the PID controller;

[0018] After the parameters are tuned, the integral and differential terms are turned on, and the anti-saturation integral algorithm is used to eliminate the accumulated error and generate the PID control signal;

[0019] The high-frequency noise interference of the PID control signal is eliminated by a fourth-order Butterworth filter, and a pulse width modulation signal is generated by a digital signal processor. At the same time, the pulse width modulation signal drives the power amplifier circuit of the magnetorheological damper to output the control current;

[0020] The magnetoelectric tension sensor is used to record the tension fluctuation data in real time, and the multi-spectral infrared thermal imager is used to record the surface temperature gradient data.

[0021] As a preferred solution of the processing control method of the carbon fiber composite material transmission shaft of the present invention, wherein: the establishment of the three-dimensional temperature field model and the calculation of the resin flow compensation coefficient specifically include the following steps:

[0022] The tension fluctuation data and temperature gradient data are mapped to node coordinates and unified in unit system using the External Data tool. The Kriging spatial interpolation algorithm is then used to convert the tension fluctuation data and temperature gradient data from discrete fields to continuous fields to generate coupled field data.

[0023] Based on the coupled field data, a hexahedral structured grid is generated through a parametric modeling engine and a bidirectional fluid-structure coupling algorithm. The hexahedral structured grid is then solved simultaneously with the heat conduction equation and the Navier-Stokes equation to generate a three-dimensional temperature field model.

[0024] According to the three-dimensional temperature field model, the resin flow compensation coefficient is obtained using the flow front tracking algorithm.

[0025] As a preferred solution of the processing control method of the carbon fiber composite material transmission shaft of the present invention, wherein: the generating of the temperature-pressure coordinated control curve specifically includes the following steps:

[0026] Based on the resin flow compensation coefficient, the fiber bundle component geometry is obtained and reverse calibration is performed using the 3D laser reverse reconstruction method.

[0027] After the reverse calibration is completed, the temperature gradient data, pressure fluctuation rate and flow uniformity variance are simultaneously optimized using the NSGA-III multi-objective optimization algorithm, and the Das-Dennis method is combined to generate three-dimensional structured reference points;

[0028] Based on three-dimensional structured reference points, the temperature-pressure coordinated control curve is generated through adaptive normalization processing and vertical distance association strategy.

[0029] As a preferred solution of the processing control method of the carbon fiber composite material drive shaft of the present invention, wherein: the power output of the hot pressing furnace and the pressure distribution of the hydraulic equipment are dynamically adjusted, and the specific acquisition steps are as follows:

[0030] Based on the curing control instructions, the segmented PID control algorithm sets the control temperature range, switches the proportional coefficient and integral time, and adjusts the power output of the hot pressing furnace. At the same time, the sliding mode variable structure control algorithm is used to adjust the opening of the electro-hydraulic proportional valve and adjust the pressure distribution of the hydraulic equipment.

[0031] As a preferred solution of the processing control method of the carbon fiber composite material drive shaft of the present invention, wherein: the particle swarm optimization algorithm is used to generate the final process control parameter set, specifically including the following steps:

[0032] Based on the curing degree report, the particle swarm optimization algorithm is used to construct a PSO-BP neural network, and the dynamic inertia weight and non-dominated sorting are used to screen out the Pareto solution set, which is then optimized and iterated to generate the final process control parameter set.

[0033] In a second aspect, the present invention provides a processing control system for a carbon fiber composite material transmission shaft, comprising:

[0034] Tension and temperature control module, coupling modeling module, instruction generation module, curing monitoring module, and process parameter optimization module;

[0035] The tension and temperature control module is used to dynamically adjust the control current of the magnetorheological damper through a PID controller based on the process control parameter set, and record the tension fluctuation and temperature gradient data;

[0036] The coupled modeling module is used to input tension fluctuation and temperature gradient data into the ANSYS coupled simulation engine, establish a three-dimensional temperature field model, and calculate the resin flow compensation coefficient. Based on the resin flow compensation coefficient, the NSGA-III multi-objective optimization algorithm is used to generate a temperature-pressure coordinated control curve.

[0037] The instruction generation module is used to integrate the resin flow compensation coefficient and the temperature-pressure coordinated control curve through the timing control engine to generate curing control instructions;

[0038] The curing monitoring module is used to dynamically adjust the power output of the hot press and the pressure distribution of the hydraulic equipment based on the curing control instructions, simultaneously collect strain energy data and curing monitoring data, and generate a curing degree report;

[0039] The process parameter optimization module is used to generate the final process control parameter set based on the curing degree report using the particle swarm optimization algorithm to optimize the transmission shaft processing control process.

[0040] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the processing control method of the carbon fiber composite material drive shaft as described in the first aspect of the present invention is implemented.

[0041] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the processing control method of the carbon fiber composite material drive shaft as described in the first aspect of the present invention is implemented.

[0042] The beneficial effects of the present invention are as follows: By adopting a multi-source data fusion algorithm combined with an industrial Internet of Things edge computing gateway, it achieves precise monitoring and synchronous processing of multiple parameters such as carbon fiber bundle diameter, initial tension value, fiber delivery speed, and surface temperature. This not only improves the accuracy of generating the process parameter characteristic matrix, but also more comprehensively reflects the dynamic behavior of the material during processing. During the curing process, the NSGA-III multi-objective optimization algorithm is used to generate a temperature and pressure coordinated control curve. In combination with adaptive normalization processing and vertical distance correlation strategy, an accurate description of nonlinear dynamic characteristics is achieved. This not only solves the problems of resin flow unevenness and curing defects, but also enhances the controllability and efficiency of the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 The figure is a flow chart of the processing control method of the carbon fiber composite material transmission shaft.

[0045] Figure 2 Schematic diagram of the machining control system for carbon fiber composite drive shaft.

[0046] Figure 3 Flowchart for generating the final set of process control parameters using the particle swarm optimization algorithm.

[0047] Figure 4 Flowchart for generating temperature-pressure coordinated control curve based on tension fluctuation data and temperature gradient data. DETAILED DESCRIPTION

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0049] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0051] Reference Figures 1 to 4 This embodiment provides a method for controlling the processing of a carbon fiber composite material transmission shaft, comprising the following steps:

[0052] S1. Based on the process control parameter set, the control current of the magnetorheological damper is dynamically adjusted through the PID controller, and the tension fluctuation and temperature gradient data are recorded.

[0053] Based on optical imaging and image processing technology, a high-precision visual inspection unit is designed using a high-resolution camera. The visual inspection unit is installed directly above the fiber bundle conveying path to continuously capture cross-sectional images of the fiber bundle and use an edge detection algorithm to collect fiber bundle diameter data in real time. The visual inspection unit not only provides a non-contact and accurate measurement method, but also can continuously monitor changes in fiber diameter without affecting the production process, greatly improving the accuracy and efficiency of process control.

[0054] A magnetoelectric tension sensor is installed at the entrance of the fiber bundle conveying path. The initial tension value is monitored using electromagnetic induction signals, and the initial tension value data is acquired through a signal acquisition circuit. After acquisition, the initial tension value data is further calibrated using a non-contact laser displacement meter to compensate for measurement deviations caused by environmental changes or equipment wear. The combined application of the magnetoelectric tension sensor and the non-contact laser displacement meter enables stable and reliable tension readings under dynamic conditions, avoiding the additional errors and complexity that may be introduced by traditional mechanical methods.

[0055] A fiber-optic photoelectric encoder is installed on the transmission wheel of the fiber bundle conveying path. The fiber conveying speed is monitored using the photoelectric sensing principle. The fiber conveying speed data is obtained through the signal acquisition circuit. At the same time, the fiber conveying speed data is further calibrated using synchronous time stamp technology to ensure that the fiber conveying speed readings provided are stable and accurate, thereby enhancing the real-time and reliability of the entire process control.

[0056] An infrared thermal imager is installed above the fiber bundle conveying path. The fiber surface temperature is monitored using the infrared radiation detection principle, and the fiber surface temperature data is collected using a high-speed data acquisition card. The infrared thermal imager ensures stable and accurate temperature readings under dynamic conditions, thereby improving the overall controllability of drive shaft manufacturing.

[0057] After data acquisition is completed, it is processed using a multi-source data fusion algorithm. Specifically, an adaptive sliding window is used to align the fiber bundle diameter, initial fiber bundle tension, fiber delivery speed, and fiber surface temperature data. The fiber bundle diameter data is processed through wavelet transformation to eliminate optical measurement noise, and the initial fiber bundle tension data is processed through Kalman filtering to remove mechanical vibration interference.

[0058] Then, in the fusion stage, based on the multi-source data fusion algorithm, the pre-processed fiber bundle diameter, fiber bundle initial tension value, fiber delivery speed, and fiber surface temperature data are dynamically weighted to obtain a multi-source feature vector. Based on the multi-source feature vector, feature association and conflict resolution are performed using the sliding window integration method to generate a composite sensing data set.

[0059] Based on the composite sensor data set, a process parameter feature matrix is ​​generated. Specifically, the composite sensor data set is reduced in dimension through principal component analysis to obtain principal component projection data. Based on the principal component projection data, feature screening is performed using the variance contribution rate threshold method to obtain core features. Subsequently, the kernel density estimation (KDE) algorithm is applied to model the probability distribution of the core features, and the Gaussian kernel function is used to obtain the probability density value. Subsequently, the obtained core features are smoothed using the time series sliding average method, and the smoothed core features and the probability density values ​​are aligned by timestamp to construct the process parameter feature matrix.

[0060] Based on the process parameter feature matrix, the initial process control parameter set is generated through the process parameter optimization engine. In the specific operation, the process parameter feature matrix is ​​transmitted to the process parameter optimization engine through the API data interface, and the improved NSGA-III multi-objective optimization algorithm is used to initialize the process parameter feature matrix population to generate initial population individuals. At the same time, fitness evaluation and population screening are performed through non-dominated sorting and crowding calculation to obtain the next generation optimized population. The next generation optimized population is then iteratively optimized to generate the process control parameter set.

[0061] Based on the process control parameter set, the control current of the magnetorheological damper is dynamically adjusted by the PID controller. It is necessary to combine the improved Ziegler-Nichols critical proportion method with multi-dimensional sensing technology. In the specific operation, the parameters of the PID controller are first tuned, the integral and differential terms of the PID controller are turned off, and the proportional gain is gradually increased in the magnetorheological damper control loop. In the process of gradually increasing the proportional gain, the tension fluctuation data is monitored by the sampling rate of the magnetoelectric tension sensor. When the tension fluctuation amplitude reaches the critical threshold and shows periodic oscillation, it is determined that the control loop of the magnetorheological damper enters the critical oscillation state, and the critical proportional gain and critical period are recorded. After the parameter tuning is completed, the integral and differential terms are enabled, and the anti-saturation integral algorithm is used to set the integral limit threshold. Based on the integral limit threshold, the cumulative error is eliminated by limiting the integral term cumulative acceleration and reverse error feedback compensation to generate a smooth PID control signal.

[0062] It should be noted that the critical threshold is defined based on the tension fluctuation amplitude during periodic oscillation and is used to identify the control instability state; the integral limit threshold is defined based on the accumulated value of the tension fluctuation energy and is used to prevent continuous overload.

[0063] Based on the generated PID control signal, high-frequency noise interference is eliminated by fourth-order Butterworth filtering. In the specific operation, the frequency domain parameters of the fourth-order Butterworth filter are first configured in the digital signal processor, and the cutoff frequency is set to one-fifth of the sampling frequency according to the operating frequency range of the magnetorheological damper (for example, the cutoff frequency is set to 2kHz when the sampling frequency is 10kHz). At the same time, the PID control signal is discretized into the analog filter transfer function by using the bilinear transformation method, and the continuous time variables in the analog filter transfer function are replaced by the discrete time mapping relationship in the bilinear transformation formula through algebraic substitution to generate the differential equation coefficients of the fourth-order Butterworth filter. Based on the differential equation coefficients, two second-order sections are cascaded to achieve fourth-order filtering, where each second-order section is implemented using a direct type II structure.

[0064] The PID control signal processed by the fourth-order Butterworth filter is used to generate a pulse width modulation signal through a digital signal processor. In the specific operation, the period register and comparison register are first configured in the digital signal processor, and the base frequency of the pulse width modulation signal is set through the period register. The duty cycle of the filtered PID signal is mapped in real time using the comparison register. At the same time, to prevent the upper and lower bridge arms of the power amplifier circuit from being directly connected, an adaptive dead zone compensation strategy is adopted. The dead zone time range (such as 50-500ns) is dynamically adjusted through a closed-loop feedback mechanism. The power amplifier circuit adopts an H-bridge topology structure and uses N-channel MOSFET as the switching device. At the same time, a parallel RC snubber network is used to suppress the switching spike voltage. Finally, a high-precision pulse width modulation signal is generated by an integrated driver chip and transmitted to the power amplifier circuit through a high-speed optocoupler (such as 6N137) for isolation. The magnetorheological damper coil is driven to generate a controllable magnetic field to achieve dynamic adjustment of the control current.

[0065] While dynamically adjusting the control current, the magnetoelectric tension sensor obtains tension fluctuations through the strain gauge detection unit and converts the tension fluctuations into electrical signals using a full-bridge circuit. After being amplified by an instrumentation amplifier, the electrical signal is input into a 24-bit high-precision ADC for digitization. The Kalman filter algorithm is used to suppress noise interference and finally transmitted to the digital signal processor through the SPI interface. The fast Fourier transform (FFT) algorithm is used to perform spectral analysis of the tension fluctuations. At the same time, the thermal imager is calibrated using a blackbody furnace (800-2200K), and the true temperature field is calculated based on Planck's multispectral infrared thermal imager law and the multispectral radiation temperature measurement algorithm. Based on the calculated true temperature field, the fiber surface temperature gradient is synchronously collected.

[0066] S2. Input the tension fluctuation and temperature gradient data into the ANSYS coupled simulation engine, establish a three-dimensional temperature field model and calculate the resin flow compensation coefficient. At the same time, based on the resin flow compensation coefficient, the temperature-pressure coordinated control curve is generated through the NSGA-III multi-objective optimization algorithm.

[0067] The specific operations include the following:

[0068] The collected tension fluctuation data and temperature gradient data are mapped to node coordinates using the coordinate conversion algorithm through the ExternalData tool. Based on the unit conversion table, the unit system is unified through automatic unit matching and standardization operations. At the same time, the Kriging spatial interpolation algorithm is used to perform weighted averaging on the neighboring points of the tension fluctuation data and temperature gradient data to convert the tension fluctuation data and temperature gradient data from discrete fields to continuous fields.

[0069] After the conversion is completed, the coupled field data is generated through the multi-source data fusion technology. In the specific operation, first, the tension fluctuation data and the temperature gradient data are temporally and spatially aligned, and the aligned tension fluctuation data and temperature gradient data are dynamically weighted fused using the multi-source data fusion technology to obtain the multi-physical field eigenvector; based on the multi-physical field eigenvector, nonlinear feature mapping is performed through the radial basis function to generate the coupled feature matrix; based on the coupled feature matrix, the finite element-control volume coupling algorithm (FE-CV) is used to perform multi-field collaborative solution to generate the coupled field data;

[0070] Based on the generated coupling field data, a hexahedral structured grid is generated using a parametric modeling engine and a bidirectional fluid-solid coupling algorithm. In specific operations, the coupling field data is input into the parametric modeling engine through the API data interface, and the bidirectional fluid-solid coupling algorithm is used to set boundary conditions for the coupling field data. For example, in the interface area where the fiber bundle contacts the resin, the temperature boundary is set to a constant heat flux input to simulate the curing heating process, and stress constraints are applied to reflect the influence of fiber tension on resin flow. After the boundary conditions are set, an initial geometric framework is generated through geometric mapping and node interpolation. The initial geometric framework is subjected to node interpolation and unit segmentation through a meshing algorithm to generate a hexahedral structured grid. The hexahedral structured grid can not only improve the resolution and accuracy of the three-dimensional temperature field model, but also allow efficient simulation of complex geometric shapes.

[0071] The heat conduction equation and the Navier-Stokes equation are solved simultaneously to generate a three-dimensional temperature field model. In the specific operation, based on the heat conduction equation, the finite element analysis method is used to solve the steady-state and transient heat distribution of the hexahedral structured grid to generate a temperature distribution field. At the same time, based on the Navier-Stokes equation, the finite volume method is used to solve the fluid dynamic characteristics of the hexahedral structured grid to generate a pressure distribution field. The generated temperature distribution field and pressure distribution field are input into the ANSYS coupled simulation environment, and the bidirectional fluid-solid coupling iterative calculation is started to generate transient thermal-fluid coupled field data.

[0072] Based on transient thermal-fluid coupling field data, the viscosity parameters are dynamically adjusted through an adaptive PID control algorithm, and the flow front tracking algorithm is used to dynamically adjust the adaptive encryption area of ​​the hexahedral structured grid to generate an optimized multi-physics field coupling grid model. The field variable reconstruction algorithm in the ANSYS CFX-Post post-processor is used to perform multi-physics field data fusion and spatiotemporal interpolation calculations on the optimized multi-physics field coupling grid model. At the same time, the isosurface extraction algorithm and volume rendering technology are combined to generate a three-dimensional temperature field model including temperature gradients, heat flow vectors and solidification fronts.

[0073] According to the three-dimensional temperature field model, the level set method is first used to capture the position of the resin flow front, and local grid encryption is implemented at the resin flow front. At the same time, the velocity vector field is solved by the finite difference method to obtain high-precision velocity field data. The simulated flow path is reconstructed based on the high-precision velocity field data using the inverse particle tracking algorithm. The real-time flow path is obtained through high-speed industrial cameras and image processing technology. The least squares method is used to fit the deviation curve between the simulated flow path and the real-time flow path, and the Savitzky-Golay filtering algorithm is used to extract the slope of the deviation curve as the resin flow compensation coefficient. The specific mathematical formula is as follows:

[0074]

[0075] Where C represents the resin flow compensation coefficient, N represents the number of all sampling points used to compare the simulated flow path with the actual flow path, and i represents the index of the sampling point. Represents the slope value of the i-th sampling point after Savitzky-Golay filtering;

[0076] After obtaining the resin flow compensation coefficient, the component geometry is obtained and reverse calibrated through the three-dimensional laser reverse reconstruction method. In the specific operation, a line laser scanner is first used to scan the component surface at multiple angles with a resolution of 0.05mm, and a three-dimensional point cloud is generated through the triangulation principle. At the same time, an industrial camera is used to collect the component texture image, and the texture image is mapped to the three-dimensional point cloud through the SIFT feature matching algorithm. Then, the three-dimensional point cloud is scanned using the ICP (iterative closest point) algorithm, and the CAD design model of the drive shaft is collected through the enterprise PLM. The three-dimensional point cloud and the CAD design model are matched for feature points using the KD-Tree nearest neighbor search algorithm.

[0077] Based on the matched feature points, a bidirectional distance constraint strategy is used to complete high-precision registration to obtain the component geometry. After registration, NURBS surface reconstruction technology is used to reconstruct the surface through the control point mesh optimization algorithm, and curvature continuity analysis is used to correct local geometric deformation to complete the reverse calibration of the component geometry.

[0078] After the reverse calibration is completed, the energy distribution of the pressure fluctuation is analyzed by the fast Fourier transform (FFT) spectrum analysis method to generate the pressure fluctuation rate; the permeability variance is calculated by the Darcy's law flow simulation algorithm to generate the flow uniformity variance; the temperature gradient, pressure fluctuation rate and flow uniformity variance are simultaneously optimized using the NSGA-III multi-objective optimization algorithm, and the Das-Dennis method is combined to generate three-dimensional structured reference points. In the specific operation, the NSGA-III algorithm parameters are first initialized in the MATLAB environment, the population size is set to 200 and the maximum evolution generation is set to 100, and 120 uniformly distributed reference points are generated in the three-dimensional target space (temperature gradient, pressure fluctuation rate, flow uniformity variance) by the Das-Dennis method. Then, the ACT extension function of ANSYS Workbench is called to read the component geometry and resin flow compensation coefficient after reverse calibration in real time, and use them as optimization boundary conditions.

[0079] Based on the optimization boundary conditions, in each iteration of the NSGA-III multi-objective optimization algorithm, Latin hypercube was used to sample the temperature field control parameters, pressure curve adjustment coefficients, and injection speed, generating 200 sets of experimental schemes.

[0080] It should be noted that the pressure curve adjustment coefficient is defined based on the dynamic pressure compensation equation, and the injection rate is generated by real-time permeability data fed back from the flow front;

[0081] Based on the generated test plan, the temperature gradient is optimized through transient heat conduction and finite element analysis, the pressure fluctuation rate is optimized through fluid-solid coupling simulation calculation, and the flow uniformity variance is optimized through the flow front monitoring algorithm. After the optimization is completed, the non-dominated sorting and crowding distance calculation built into the NSGA-III multi-objective optimization algorithm are used to output the Pareto optimal solution set. Based on the Pareto optimal solution set, the Das-Dennis method is used to generate uniformly distributed three-dimensional structured reference points.

[0082] Based on three-dimensional structured reference points, a temperature-pressure coordinated control curve is generated through adaptive normalization processing and vertical distance association strategy. Specifically, the maximum-minimum normalization method is first used to adaptively normalize the Pareto optimal solution set to eliminate dimensional differences. Then, the vertical distance association strategy is applied to calculate the minimum vertical distance between the three-dimensional structured reference point and the Pareto frontier solution using the L2 norm distance metric. Based on the minimum vertical distance, the K-means clustering algorithm uses the silhouette coefficient method to automatically determine the optimal number of clusters and screen out the optimal solution subset that balances each objective.

[0083] The B-spline curve fitting technology is used to optimize the control points and perform piecewise continuity verification on the optimal solution subset to generate a temperature-pressure coordinated control curve. In the specific operation, based on the B-spline curve fitting technology, the least squares method is used to fit the optimal solution subset to achieve control point optimization; the curvature continuity of the optimal solution subset is optimized through the node vector adjustment algorithm to generate a smooth parameterized curve; based on the smooth parameterized curve, the piecewise continuity verification is performed through the second-order derivative boundary condition;

[0084] Next, the curvature of the smooth parameterized curve is homogenized by the node insertion algorithm, and the slope constraint of the smooth parameterized curve is optimized by the piecewise cubic Hermite interpolation method to generate the temperature-pressure coordinated control curve.

[0085] S3. Integrate the resin flow compensation coefficient and the temperature-pressure coordinated control curve through a timing control engine to generate a curing control instruction.

[0086] The specific operations include the following:

[0087] The process of integrating the resin flow compensation coefficient and the temperature-pressure collaborative control curve to generate curing control instructions requires the collaborative operation of multiple technologies. In the specific operation, a real-time data channel is first established through the industrial real-time Ethernet protocol, and the resin flow compensation coefficient is read and stored in the memory buffer at a frequency of 10 times per second through the real-time data channel. At the same time, the temperature-pressure collaborative control curve is loaded from the XML configuration file using the DOM parser and saved in the memory buffer. Data synchronization is achieved by storing the resin flow compensation coefficient and the temperature-pressure collaborative control curve together in the memory buffer.

[0088] After data synchronization is complete, the resin flow compensation coefficient and temperature-pressure coordinated control curve are input into the timing control engine through memory mapping (Memory-Mapped File). The timing control engine uses a double buffering mechanism to process the resin flow compensation coefficient and the temperature-pressure coordinated control curve. The resin flow compensation coefficient is updated in real time by the priority thread to update the injection molding machine screw speed parameter. The temperature-pressure coordinated control curve is parsed by the motion control card into servo motor position instructions.

[0089] It should be noted that the injection molding machine screw speed parameter is defined based on the resin flow characteristics and the real-time position of the injection molding machine screw;

[0090] Through the motion controller's trajectory planning algorithm, the real-time updated injection molding machine screw speed parameters and servo motor position instructions are subjected to spatiotemporal interpolation calculation, kinematic smoothing, and spatiotemporal synchronization integration to generate curing control instructions containing multi-dimensional parameters such as timestamps, temperature gradients, and pressure curves.

[0091] S4. Based on the curing control instructions, dynamically adjust the power output of the hot press furnace and the pressure distribution of the hydraulic equipment, synchronously collect strain energy data and curing monitoring data, and generate a curing degree report.

[0092] The specific operations include the following:

[0093] Dynamic adjustment of the hot press furnace power output and hydraulic equipment pressure distribution is achieved based on the curing control instructions. In specific operations, the curing control instructions are transmitted to the hot press furnace controller via the PROFINET industrial bus. The segmented PID control algorithm uses the Arrhenius reaction kinetic equation to perform temperature-time integration based on the temperature gradient parameter in the curing control instructions to obtain the curing reaction rate. Based on the curing reaction rate, the control range threshold is defined using a three-point slope detection algorithm.

[0094] Based on the control interval threshold, the temperature control range is divided into three control intervals: low temperature interval, medium temperature interval, and high temperature interval. In the low temperature interval, a large proportional coefficient is used to quickly increase the temperature, and full-power pulse width modulation is used to adjust the power output of the hot press furnace. In the medium temperature interval, a proportional integral (PI) anti-saturation algorithm is used to dynamically adjust the on-off ratio of the solid-state relay to adjust the power output of the hot press furnace. In the high temperature interval, a composite control strategy combining proportional integral differential (PID) and feedforward compensation is used to adjust the power output of the hot press furnace.

[0095] A sliding-mode variable structure control algorithm is used to adjust the pressure distribution of hydraulic equipment. Specifically, the pressure curve parameters in the fixed control instructions are first converted into a desired pressure trajectory using a cubic spline interpolation algorithm. Based on the desired pressure trajectory, a PWM modulation signal with an adjustable duty cycle is generated by a digital signal processor. The PWM modulation signal drives the pilot-stage solenoid of the electro-hydraulic proportional valve to generate a pilot pressure. This pilot pressure, through the hydraulic amplification unit, pushes the main valve core to overcome the spring force and move, generating valve core displacement. The valve core displacement is detected in real time by an LVDT displacement sensor and converted into a voltage signal. The voltage signal is compared with a reference voltage through a differential amplifier to generate a displacement error signal.

[0096] Based on the displacement error signal, the control variable is calculated through a sliding mode variable structure controller, which drives the main valve core to adjust the opening of the electro-hydraulic proportional valve. Based on the adjusted opening of the electro-hydraulic proportional valve, the flow-pressure differential characteristic curve is solved in real time and dynamic feedforward compensation is performed through the throttle flow equation to generate a hydraulic flow control instruction. Based on the hydraulic flow control instruction, the servo amplifier drives the electromagnet to adjust the valve core displacement and adjust the pressure distribution of the hydraulic equipment.

[0097] In the process of dynamically adjusting the power output of the hot press furnace and the pressure distribution of the hydraulic equipment, strain energy data and curing monitoring data are synchronously collected to generate a curing degree report. In the specific operation, strain data is first collected in real time at a sampling rate of 1kHz through a fiber Bragg grating sensor array, and amplified and analog-to-digital converted by a high-speed signal conditioning circuit to obtain the original strain signal. The original strain signal is input into the strain energy calculation unit, and wavelet threshold denoising and baseline correction are performed to generate filtered strain data. The filtered strain data is then used to collect the resin dielectric loss tangent value at a frequency of 500Hz through a dielectric sensor array. Based on the resin dielectric loss tangent value, the reaction kinetic parameters are inversely calculated using the Arrhenius equation. At the same time, curing monitoring data is obtained by combining differential scanning calorimetry. The strain data and curing monitoring data are aligned by time stamp and input into the data fusion processor. Multi-source data fusion and outlier elimination are performed using the Kalman filter algorithm to obtain the final process data set.

[0098] Based on the final process data set, the principal component analysis method is used to extract the coupling feature matrix, and the feature dimension reduction of the coupling feature matrix is ​​performed through the threshold screening method to obtain the key feature matrix; the key feature matrix is ​​input into the curing degree prediction algorithm, and the predicted curing degree curve is output. Finally, combined with the hot pressing furnace temperature gradient data and hydraulic pressure distribution data, a curing degree report containing indicators such as residual stress distribution and curing uniformity is generated through a three-dimensional interpolation algorithm.

[0099] S5. Based on the curing degree report, the particle swarm optimization algorithm is used to generate the final process control parameter set to optimize the transmission shaft processing control process.

[0100] The specific steps include:

[0101] Based on the cure degree report, the particle swarm optimization algorithm was used to generate the final set of process control parameters. Specifically, the residual stress distribution data in the cure degree report was first converted into a nodal stress matrix through three-dimensional mesh discretization. The nodal stress matrix was then input into the particle swarm optimization algorithm, and an initial population of 50 sets of process parameters was generated using the Latin hypercube sampling method. Each set of parameters in the initial population included three dimensions: temperature gradient, pressure curve, and cure time.

[0102] The initial population is used to construct a three-layer feedforward structure through a back-propagation neural network, with 3 nodes in the input layer, 20 nodes in the hidden layer, and 3 nodes in the output layer. At the same time, a particle swarm optimization weight update mechanism is embedded to obtain the PSO-BP neural network architecture. The PSO-BP neural network architecture uses the Sigmoid activation function and the mean square error loss function. The initial training is performed through the mini-batch gradient descent method to generate a pre-trained neural network model. The pre-trained neural network model is then subjected to online incremental learning and dynamic weight adjustment to finally output the PSO-BP neural network.

[0103] After the PSO-BP neural network is constructed, a linear reduction strategy is adjusted using dynamic inertia weights to generate an optimized search strategy. Based on this optimized search strategy, the NSGA-II algorithm framework is used to perform rapid grading and congestion calculation using non-dominated sorting, generating a Pareto solution set with good distribution. This Pareto solution set is then sub-optimized iteratively using an elite retention strategy to generate an improved parameter population.

[0104] The improved parameter population is physically verified in a hydraulic servo test unit, and material flow images are collected by a high-speed industrial camera. At the same time, the image processing algorithm is used to generate process performance indicators, which are fed back to the PSO-BP neural network for adaptive correction. The calibrated PSO-BP neural network is output. The calibrated PSO-BP neural network reinitializes the population boundary through the elite retention strategy to generate an enhanced parameter combination. Based on the enhanced parameter combination, multi-objective optimization is performed using a non-dominated sorting genetic algorithm to obtain a non-dominated solution set. Based on the non-dominated solution set, the final process control parameter set is screened out using the TOPSIS decision method.

[0105] This embodiment also provides a carbon fiber composite material transmission shaft processing control system, including: a tension and temperature control module, a coupling modeling module, an instruction generation module, a curing monitoring module, and a process parameter optimization module;

[0106] The tension and temperature control module is used to dynamically adjust the control current of the magnetorheological damper through a PID controller based on the process control parameter set, and record the tension fluctuation and temperature gradient data;

[0107] The coupled modeling module is used to input tension fluctuation and temperature gradient data into the ANSYS coupled simulation engine, establish a three-dimensional temperature field model, and calculate the resin flow compensation coefficient. Based on the resin flow compensation coefficient, the NSGA-III multi-objective optimization algorithm is used to generate a temperature-pressure coordinated control curve.

[0108] The instruction generation module is used to integrate the resin flow compensation coefficient and the temperature-pressure coordinated control curve through the timing control engine to generate curing control instructions;

[0109] The curing monitoring module is used to dynamically adjust the power output of the hot press and the pressure distribution of the hydraulic equipment based on the curing control instructions, simultaneously collect strain energy data and curing monitoring data, and generate a curing degree report;

[0110] The process parameter optimization module is used to generate the final process control parameter set based on the curing degree report using the particle swarm optimization algorithm to optimize the transmission shaft processing control process.

[0111] This embodiment also provides a computer device suitable for the case of a method for controlling the processing of a carbon fiber composite material drive shaft, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for controlling the processing of a carbon fiber composite material drive shaft as proposed in the above embodiment.

[0112] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0113] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the processing control method for a carbon fiber composite material drive shaft as proposed in the above embodiment; the storage medium 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 memory, flash memory, magnetic disk or optical disk.

[0114] In summary, the present invention achieves precise monitoring and synchronous processing of multiple parameters such as carbon fiber bundle diameter, initial tension value, fiber delivery speed, and surface temperature by adopting a multi-source data fusion algorithm combined with an industrial Internet of Things edge computing gateway. This not only improves the accuracy of generating the process parameter characteristic matrix, but also more comprehensively reflects the dynamic behavior of the material during processing. During the curing process, the NSGA-III multi-objective optimization algorithm is used to generate a temperature and pressure coordinated control curve, and adaptive normalization processing and vertical distance correlation strategy are combined to achieve an accurate description of nonlinear dynamic characteristics. This not only solves the problems of resin flow unevenness and curing defects, but also enhances the controllability and efficiency of the production process.

[0115] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for controlling the processing of a carbon fiber composite material transmission shaft, characterized by: include, Based on the process control parameter set, the control current of the magnetorheological damper is dynamically adjusted through the PID controller, and the tension fluctuation and temperature gradient data are recorded; The tension fluctuation and temperature gradient data were input into the ANSYS coupled simulation engine to establish a three-dimensional temperature field model and calculate the resin flow compensation coefficient. Based on the resin flow compensation coefficient, the NSGA-III multi-objective optimization algorithm was used to generate a temperature-pressure coordinated control curve. The resin flow compensation coefficient and temperature-pressure coordinated control curve are integrated through the timing control engine to generate curing control instructions; Based on the curing control instructions, the power output of the hot press furnace and the pressure distribution of the hydraulic equipment are dynamically adjusted, and strain energy data and curing monitoring data are simultaneously collected to generate a curing degree report; Based on the curing degree report, the particle swarm optimization algorithm is used to generate the final process control parameter set to optimize the transmission shaft machining control process.

2. The carbon fiber composite material transmission shaft processing control method according to claim 1, characterized in that: The specific steps for obtaining the process control parameter set are as follows: The fiber bundle diameter, initial tension value, fiber delivery speed, and fiber surface temperature data are weightedly fused using a multi-source data fusion algorithm to obtain a composite sensing data set. The composite sensor data set is subjected to feature dimensionality reduction through principal component analysis to generate a process parameter feature matrix; Based on the process parameter feature matrix, the process control parameter set is generated through the process parameter optimization engine using the improved NSGA-III multi-objective optimization algorithm.

3. The carbon fiber composite material transmission shaft processing control method according to claim 1, characterized in that: The PID controller is used to dynamically adjust the control current of the magnetorheological damper and record the tension fluctuation and temperature gradient data. The specific steps are as follows: Based on the process control parameter set, the improved Ziegler-Nichols critical proportion method is used to tune the parameters of the PID controller; After the parameters are tuned, the integral and differential terms are turned on, and the anti-saturation integral algorithm is used to eliminate the accumulated error and generate the PID control signal; The high-frequency noise interference of the PID control signal is eliminated by a fourth-order Butterworth filter, and a pulse width modulation signal is generated by a digital signal processor. At the same time, the pulse width modulation signal drives the power amplifier circuit of the magnetorheological damper to output the control current; The magnetoelectric tension sensor is used to record the tension fluctuation data in real time, and the multi-spectral infrared thermal imager is used to record the surface temperature gradient data.

4. The carbon fiber composite material transmission shaft processing control method according to claim 1, characterized in that: The establishment of the three-dimensional temperature field model and calculation of the resin flow compensation coefficient specifically include the following steps: The tension fluctuation data and temperature gradient data are mapped to node coordinates and unified in unit system using the External Data tool. The Kriging spatial interpolation algorithm is then used to convert the tension fluctuation data and temperature gradient data from discrete fields to continuous fields to generate coupled field data. Based on the coupled field data, a hexahedral structured grid is generated through a parametric modeling engine and a bidirectional fluid-structure coupling algorithm. The hexahedral structured grid is then solved simultaneously with the heat conduction equation and the Navier-Stokes equation to generate a three-dimensional temperature field model. According to the three-dimensional temperature field model, the resin flow compensation coefficient is obtained using the flow front tracking algorithm.

5. The carbon fiber composite material transmission shaft processing control method according to claim 1, characterized in that: The generation of the temperature-pressure coordinated control curve specifically includes the following steps: Based on the resin flow compensation coefficient, the fiber bundle component geometry is obtained and reverse calibration is performed using the 3D laser reverse reconstruction method. After the reverse calibration is completed, the temperature gradient data, pressure fluctuation rate and flow uniformity variance are simultaneously optimized using the NSGA-III multi-objective optimization algorithm, and the Das-Dennis method is combined to generate three-dimensional structured reference points; Based on three-dimensional structured reference points, the temperature-pressure coordinated control curve is generated through adaptive normalization processing and vertical distance association strategy.

6. The carbon fiber composite material transmission shaft processing control method according to claim 1, characterized in that: The specific steps for dynamically adjusting the power output of the hot pressing furnace and the pressure distribution of the hydraulic equipment are as follows: Based on the curing control instructions, the segmented PID control algorithm sets the control temperature range, switches the proportional coefficient and integral time, and adjusts the power output of the hot pressing furnace. At the same time, the sliding mode variable structure control algorithm is used to adjust the opening of the electro-hydraulic proportional valve and adjust the pressure distribution of the hydraulic equipment.

7. The carbon fiber composite material transmission shaft processing control method according to claim 6, characterized in that: The method of using the particle swarm optimization algorithm to generate the final process control parameter set specifically includes the following steps: Based on the curing degree report, the particle swarm optimization algorithm is used to construct a PSO-BP neural network, and the dynamic inertia weight and non-dominated sorting are used to screen out the Pareto solution set, which is then optimized and iterated to generate the final process control parameter set.

8. A carbon fiber composite material transmission shaft processing control system, based on the carbon fiber composite material transmission shaft processing control method according to any one of claims 1 to 7, characterized in that: Including tension and temperature control module, coupling modeling module, instruction generation module, curing monitoring module, and process parameter optimization module; The tension and temperature control module is used to dynamically adjust the control current of the magnetorheological damper through a PID controller based on the process control parameter set, and record the tension fluctuation and temperature gradient data; The coupled modeling module is used to input tension fluctuation and temperature gradient data into the ANSYS coupled simulation engine, establish a three-dimensional temperature field model, and calculate the resin flow compensation coefficient. Based on the resin flow compensation coefficient, the NSGA-III multi-objective optimization algorithm is used to generate a temperature-pressure coordinated control curve. The instruction generation module is used to integrate the resin flow compensation coefficient and the temperature-pressure coordinated control curve through the timing control engine to generate curing control instructions; The curing monitoring module is used to dynamically adjust the power output of the hot press and the pressure distribution of the hydraulic equipment based on the curing control instructions, simultaneously collect strain energy data and curing monitoring data, and generate a curing degree report; The process parameter optimization module is used to generate the final process control parameter set based on the curing degree report using the particle swarm optimization algorithm to optimize the transmission shaft processing control process.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the carbon fiber composite material transmission shaft processing control method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for controlling the processing of a carbon fiber composite material transmission shaft according to any one of claims 1 to 7 are implemented.

Images