A process parameter intelligent optimization system for automobile steel wire production

Abstract

The application belongs to the technical field of industrial automation control, and particularly relates to a process parameter intelligent optimization system for automobile steel wire production, which comprises a data acquisition module, a deviation calculation module, a prediction module and a control module; the data acquisition module is used for constructing a Lagrange coordinate system with material particles as cores, collecting whole-process time sequence process data and mapping the data to the material particles; the deviation calculation module is used for calculating historical deviation modulus of the material particles based on arc length integral of state deviation trajectory; the prediction module is used for constructing a dissipation damping model based on the historical deviation modulus, and calculating performance loss amount of the material particles; and the control module is used for generating process parameter compensation amount through reverse projection and implementing feedforward control on equipment based on preset historical regression gradient vector and the performance loss amount. The application realizes closed-loop compensation on historical cumulative damage, and improves consistency of automobile steel wire product performance.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology. More specifically, this invention relates to an intelligent optimization system for process parameters in the production of automotive steel wire. Background Technology

[0002] As a core raw material for tire skeletons and suspension springs, automotive steel wires have mechanical properties that directly affect the safety and comfort of driving. In modern metallurgical industry, the production of automotive steel wires usually involves continuous and complex processing steps such as heat treatment, surface treatment, and multiple drawing processes. These processes require extremely high precision in controlling process parameters.

[0003] Currently, production sites commonly use programmable logic controllers (PLCs) combined with supervisory control and data acquisition and control systems (SCADA) to achieve automated control. The underlying core control algorithms are mostly based on proportional-integral-derivative (PID) control strategies. This traditional control system mainly focuses on maintaining the constant physical quantities such as temperature, speed, or tension at specific detection points of the equipment, that is, ensuring that the equipment operates within a preset steady-state range.

[0004] However, on a high-speed, continuous drawing production line, the automotive steel wire being processed is in a state of rapid flow. The evolution of its internal structure and properties is a continuous process with memory characteristics. Existing feedback control modes based on fixed-point monitoring often treat the material as a homogeneous fluid, ignoring the historical cumulative effects carried by the material as it flows between different processes. When there are slight fluctuations in the upstream process, even if the downstream equipment parameters are maintained at standard values ​​through PID adjustment, the microscopic damage or structural distortion left by temperature or stress fluctuations in the preceding process will not be automatically eliminated. Instead, it will continue to accumulate in subsequent processing through genetic effects.

[0005] This excessive focus on single-point steady state leads to a disconnect between the control system and the actual quality evolution of the material, making it difficult for the production line to cope with nonlinear quality fluctuations caused by multi-parameter coupling. Furthermore, traditional offline sampling inspection methods for finished products have significant time lags and cannot promptly prevent performance dispersion risks during the production process. Summary of the Invention

[0006] To address the technical problem of existing single-point feedback control modes neglecting the historical cumulative effects during material processing, thus failing to effectively guarantee the performance consistency of automotive steel wire products, this invention provides an intelligent optimization system for process parameters in automotive steel wire production. The system includes the following modules: a data acquisition module, used to construct a Lagrange coordinate system and discretize the automotive steel wire into material particles, acquire time-series data and map it to the material particles to form a full-process process parameter vector sequence; a deviation calculation module, used to obtain the state deviation trajectory based on the full-process process parameter vector sequence, and obtain the deviation modulus of the material particles at the current moment by calculating the arc length of the deviation trajectory relative to the ideal trajectory; a prediction module, used to obtain the damping equation based on the deviation modulus, and use the damping equation to calculate the performance loss of the material particles; and a control module, used to generate the process parameter compensation amount at the current moment through inverse projection based on a preset historical regression gradient vector and performance loss amount, and implement feedforward control of the equipment based on the process parameter compensation amount.

[0007] This invention establishes a Lagrange coordinate system through a data acquisition module and discretizes automotive steel wire into material particles. A deviation calculation module acquires the state deviation trajectory and calculates the deviation arc length relative to the ideal trajectory. A prediction module then assesses the performance loss of the material particles based on the damping equation. Finally, a control module uses inverse projection to generate process parameter compensation amounts and implements feedforward control of the equipment. This approach transforms time-based discrete monitoring into continuous historical tracking centered on the material itself. After assessing the accumulated non-ideal fluctuations during automotive steel wire production, it proactively generates compensation strategies, thereby reducing the performance fluctuations of the final product caused by historical accumulated damage and improving the overall quality consistency of the finished automotive steel wire.

[0008] Preferably, the step of collecting time-series data and mapping it to material particles includes: using an encoder, an infrared thermometer, and a tension sensor to collect speed, temperature, tension, and mold compression ratio on the production line in real time; and mapping the collected time-series data to material particles by integrating the speed.

[0009] This invention utilizes sensors such as encoders and infrared thermometers to collect data and maps the time-series data onto material particles by integrating the speed. This method establishes a one-to-one correspondence between process data and the physical location of automotive steel wires, enabling the system to trace the actual experience of each segment of automotive steel wire in different processes, thereby providing a precise spatial positioning basis for subsequent assessment of local cumulative damage.

[0010] Preferably, the deviation modulus satisfies the expression: In the formula, for The deviation modulus of a material particle at any given moment; This refers to the initial moment when material particles enter the production line. Let be the integral variable, representing the time from the initial moment. At the time Any historical point in time between; For material particles at historical moments The actual strain rate; The ideal strain rate is set in the process standard for the corresponding process. It is the minimum of the maximum allowable operating rate of the equipment in the process and the allowable critical deformation rate of the material; For material particles at historical moments The actual instantaneous temperature; The ideal target temperature set in the process standard for the relevant process; This is the highest safe temperature allowed in the process. It is a time differential unit.

[0011] This invention obtains the deviation modulus by introducing an integral expression that includes the difference in strain rate and temperature. This approach integrates the dual effects of deformation rate fluctuations and thermal history fluctuations on the material, unifying physical quantities of different dimensions into a dimensionless space for evaluation, thereby more comprehensively reflecting the actual damage degree of automotive steel wires under complex processing environments.

[0012] Preferably, the performance loss is obtained by multiplying the standard evolution rate vector by the time step to obtain the theoretical performance improvement under the ideal damage-free state; calculating the memory dissipation factor using the natural exponential function, and determining the damping loss rate based on the memory dissipation factor; and multiplying the theoretical performance improvement by the damping loss rate to obtain the performance loss.

[0013] Preferably, the performance loss satisfies the expression: In the formula, for The performance loss at any given moment; The standard evolution rate vector; for The deviation modulus at time; It is the time step; It is a natural exponential function; This represents the historical sensitivity coefficient.

[0014] Preferably, the standard evolution rate vector is obtained by: acquiring the standard process curve pre-stored in the industrial control computer, and calculating the tangent slope of the standard process curve as the standard evolution rate vector.

[0015] Preferably, the process parameter compensation amount satisfies the expression: In the formula, This is the compensation amount for process parameters; This is the compensation gain coefficient; This represents the historical regression gradient vector; The square of the magnitude of the historical regression gradient vector; for The performance loss at any given moment; Let be the vector magnitude.

[0016] This invention uses the historical regression gradient vector and its magnitude to calculate the performance loss and determines the compensation amount of the process parameters through inverse projection. This method dynamically adjusts the compensation range based on the current sensitivity of the equipment to control commands, reducing control deviations caused by equipment state drift and thus improving the accuracy of process parameter adjustment.

[0017] Preferably, the historical regression gradient vector is obtained by selecting the most recent... Historical production data from each batch is used to obtain the input matrix of process parameter fluctuations and the output vector of performance status fluctuations. The slope of the linear regression equation between the input matrix and the output vector is calculated using online least squares regression, and this slope is used as the historical regression gradient vector. This is the preset batch quantity.

[0018] This invention selects historical production data from several recent batches and calculates the slope between the input matrix and the output vector using online least squares regression. This slope is then used as the historical regression gradient vector. This approach enables the control model to update in accordance with the real-time status of the production line, promptly capturing subtle trends caused by equipment aging or environmental changes, thereby enhancing the system's adaptability to the long-cycle production process of automotive steel wire.

[0019] Preferably, the historical sensitivity coefficient is set to... .

[0020] Preferably, the control module includes: calculating the pseudo-inverse of the historical regression gradient vector, using the pseudo-inverse to reverse map the performance loss to the equipment process parameter space, and generating the process parameter adjustment direction and magnitude that can produce an equivalent performance improvement.

[0021] The beneficial effects of this invention are as follows: This invention constructs a Lagrange coordinate system with material particles as the core and calculates the deviation arc length of the state deviation trajectory. This allows discrete state data of automotive steel wire in multiple processes to be linked into a continuous cumulative effect file. After discovering minor deviations in early processes, targeted adjustments can be made in subsequent processes in a timely manner, thereby reducing the situation where product performance fails to meet standards due to the neglect of cumulative effects.

[0022] This invention evaluates performance loss based on deviation modulus and damping equations and implements inverse projection control. It can predict the performance deficit of automotive steel wire before all processing is completed and generate compensation amounts for process parameters including heating power or drawing speed adjustment. This feedforward processing logic changes the traditional post-inspection mode, thereby reducing the scrap rate in automotive steel wire production.

[0023] This invention establishes a standard process curve that includes ideal strain rate and ideal target temperature, and uses this as a benchmark to calculate the deviation of the entire process parameter vector sequence. This provides a standardized evaluation benchmark for complex industrial production environments, making the quality of automotive steel wire produced by different batches or different equipment comparable, thereby promoting the standardized management and optimization of automotive steel wire production processes. Attached Figure Description

[0024] Figure 1 This is a schematic diagram illustrating a system block diagram of an intelligent optimization system for process parameters in the production of automotive steel wire according to the present invention. Figure 2 It is a schematic diagram illustrating the cumulative state deviation of material particles during the processing. Figure 3 This is a schematic diagram illustrating the evolution of performance loss as a function of the deviation modulus. Figure 4 This is a schematic diagram illustrating the distribution of process parameter compensation amounts. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] This invention provides an intelligent optimization system for process parameters in the production of automotive steel wire. For example... Figure 1 As shown, an intelligent optimization system for process parameters in automotive steel wire production includes a data acquisition module 100, a deviation calculation module 200, a prediction module 300, and a control module 400, which are described in detail below.

[0028] The data acquisition module 100 is used to construct a Lagrange coordinate system and discretize the automotive steel wire into material particles, collect time-series data and map it to the material particles to form a vector sequence of process parameters for the entire process.

[0029] It should be noted that, in order to track the historical experience of automotive steel wire throughout the entire production process, this invention establishes a Lagrange perspective that follows the movement of automotive steel wire, recording complete time-series data of every tiny element, i.e., material particle, on the automotive steel wire from wire rod unwinding to finished product winding, laying the foundation for subsequent calculations of historical cumulative effects.

[0030] Specifically, this invention establishes a Lagrange coordinate system that follows the movement of the automotive steel wire, discretizing the entire automotive steel wire to be processed into several independent material particles along its length. Real-time drawing speed is collected synchronously using sensors such as rotary encoders and infrared thermometers installed on the production line. Actual instantaneous temperature of automotive steel wire and the current mold compression ratio To obtain the derivative physical quantities required by the model, the system calculates the actual strain rate at the current moment based on the principle of metal plastic forming and using the engineering-standard expression for the average strain rate. The system retrieves benchmark data matching the current process from the pre-stored process database on the industrial control computer, including the ideal strain rate. Ideal target temperature The maximum allowable ultimate strain rate of the equipment Maximum safe temperature Standard evolution rate vector and historical sensitivity coefficient By adjusting the drawing speed The cumulative displacement of the automotive steel wire is calculated by integration, and all real-time calculated actual values ​​are packaged and mapped with the retrieved reference values ​​to the specific material particles in the currently processed area. This forms a vector sequence of process parameters for the entire process.

[0031] The deviation calculation module 200 is used to obtain the state deviation trajectory based on the whole process parameter vector sequence, and obtain the deviation modulus of the material particle at the current moment by calculating the deviation arc length of the state deviation trajectory relative to the ideal trajectory.

[0032] It should be noted that this invention treats the non-ideal fluctuations of automotive steel wire during processing as a path in a high-dimensional state space, and evaluates the historical cumulative effect by calculating the arc length of the deviation of this path from the ideal trajectory.

[0033] Specifically, for each material particle This invention calculates the deviation modulus of a material particle at the current moment, and the deviation modulus satisfies the expression:

[0034] In the formula, for The deviation modulus of a material particle at any given moment; This refers to the initial moment when material particles enter the production line. Let be the integral variable, representing the time from the initial moment. At the time Any historical point in time between; For material particles at historical moments The actual strain rate; The ideal strain rate is set in the process standard for the corresponding process. It is the minimum of the maximum allowable operating rate of the equipment in the process and the allowable critical deformation rate of the material; For material particles at historical moments The actual instantaneous temperature; The ideal target temperature set in the process standard for the relevant process; This is the highest safe temperature allowed in the process. It is a time differential unit.

[0035] In the formula, and As the denominator, it normalizes physical quantities with different dimensions to a unified dimensionless space; the integrand calculates the state at a certain historical moment. The Euclidean distance between the actual production state point and the ideal process point in the normalized state space is calculated by integrating this distance in the time domain to determine the total path length taken by the state point to deviate from the ideal trajectory. The longer this arc length, the deeper the material has deviated from the ideal state in history, thereby assessing the irreversible damage or distortion energy accumulated inside the material.

[0036] For example, Figure 2 This is a schematic diagram illustrating the accumulation of state deviations during the processing of material particles. The diagram shows that within the processing time, as the material particles experience different process environments such as drawing and heating, the accumulated deviation modulus exhibits a monotonically increasing trend. The steep upward region in the middle of the curve reflects that the material particles encountered significant velocity or temperature fluctuations within a certain period, leading to a rapid accumulation of the deviation modulus. This verifies the deviation calculation module's ability to capture the integral effect of historical trajectories.

[0037] The prediction module 300 is used to obtain the damping equation based on the deviation modulus and to calculate the performance loss of material particles using the damping equation.

[0038] It should be noted that the deviation modulus reflects the degree of non-ideal state accumulated in the material during past processing. This invention believes that this historical accumulation effect will reduce the material's responsiveness to the current process, causing the actual evolution of material properties to lag behind the theoretical expected value. The purpose of this invention is to use known data to accurately calculate the amount of this performance improvement shortage, i.e., the amount of performance loss, thereby establishing a clear evaluation target for subsequent control and compensation.

[0039] Specifically, this invention defines the performance loss of the material property state vector, and the performance loss satisfies the expression:

[0040] In the formula, for The amount of performance loss of a material due to historical cumulative damage at any given time; The standard evolution rate vector is taken from the tangent slope of the standard process curve pre-stored in the industrial control computer; for The deviation modulus at time; It is the time step; It is a natural exponential function; This is the historical sensitivity coefficient, expressed as the reciprocal of time (seconds).

[0041] In the formula, This represents the ideal, undamaged state after... The theoretical performance improvement that the material should have achieved with time step; memory dissipation factor The value of the deviation modulus The decrease due to the increase reflects the degradation of the material's responsiveness; This is used to assess what proportion of the standard evolution rate vector is offset by historical damage; it is calculated by multiplying the theoretical performance improvement by this damping loss rate. This refers to the unrealized performance evolution caused by historically accumulated distortion energy or micro-damage, and the performance evolution is the part that the control system needs to compensate for.

[0042] In this embodiment, the performance state vector refers to the state vector composed of key mechanical properties of automotive steel wire. Specifically, the performance state vector includes three core components: tensile strength (MPa), torsion cycles (times / 360°), and reduction of area (%). Correspondingly, the standard evolution rate vector... It is a three-dimensional vector, whose three components represent the theoretical rate of change of three physical properties with respect to time under standard process conditions. For example, if the standard process curve shows that the tensile strength increases by 5 MPa per second in a certain drawing pass, then... The corresponding tensile strength component is 5 MPa / s.

[0043] It should be further explained that the historical sensitivity coefficient The range of values ​​is usually 100. The history sensitivity coefficient is determined by the microstructure stability of the automotive steel wire raw materials: for materials such as high-carbon steel or alloy steel that are highly sensitive to thermal history, Take as A large value means that even a small historical deviation can lead to a significant decline in response efficiency; while for materials such as low-carbon steel, which are not sensitive to process fluctuations, Take as The minimum value reflects its strong process robustness; in this embodiment, the historical sensitivity coefficient is set to [value missing]. .

[0044] For example, Figure 3 This is a schematic diagram of the evolution model of performance loss as a function of deviation modulus, reflecting the characteristics of the damping equation in the prediction module. Specifically, as the deviation modulus increases, the performance loss of the material exhibits a non-linear growth trend. The curve shows that when the deviation is small, the performance loss increases rapidly; however, when the deviation modulus reaches a certain level, the growth of performance loss tends to level off, reflecting the damping saturation effect of internal material damage on performance evolution.

[0045] The control module 400 is used to generate the process parameter compensation amount at the current moment by inverse projection based on the preset historical regression gradient vector and performance loss amount, and to implement feedforward control of the equipment based on the process parameter compensation amount.

[0046] It should be noted that, as a control execution link, this invention uses reverse calculation based on the amount of performance loss to determine how much adjustment needs to be made to the equipment process parameters to offset the deficit, thereby achieving precise control over the performance of the final product.

[0047] Specifically, this invention obtains the historical regression gradient vector. The method for obtaining it is as follows: select the most recent one from the historical production database. Production data and experience values ​​for each batch Extracting fluctuation data of process parameters to obtain the input matrix And extract performance fluctuation data to obtain the output vector. Establish a linear regression equation , It is the bias term. The linear regression equation is solved using the online least squares method, and the slope of the solution is determined as the historical regression gradient vector.

[0048] Furthermore, the present invention obtains The process parameter compensation amount at time t, the process parameter compensation amount satisfies the expression:

[0049] In the formula, The compensation amount for the process parameters at time t; This is the compensation gain coefficient; The historical regression gradient vector reflects in real time the actual impact of a unit control adjustment on the performance status under the current equipment condition; for The performance loss at any given moment; The length is the module length.

[0050] In the formula, This term mathematically determines the direction and unit influence factor that can most efficiently affect the performance state in the multidimensional process parameter space; multiplying this by the performance loss... This allows us to work backwards to calculate how to fill the deficit. Process parameters at time The specific adjustment range and direction are determined; the calculation process constitutes a closed-loop compensation for the historical cumulative effect, ensuring that the control system can actively offset the performance lag caused by historical damage.

[0051] It should be noted that the compensation gain coefficient It is a key damping parameter that determines the strength of feedforward control. Its value range is usually set to [0.4, 0.85]. The specific value of the compensation gain coefficient needs to be tuned on-site according to the response lag of the production equipment: for new heating or drawing equipment with rapid response, A larger value, such as 0.7, can be used to achieve rapid compensation; for older equipment with significant inertial lag, The minimum value should be selected, such as 0.4, to prevent system oscillations caused by overcompensation. During the initial debugging, it is recommended to use 0.5 as the baseline value and fine-tune it through step response testing until the system does not exhibit overshoot during the compensation process.

[0052] Furthermore, the control module will calculate the compensation amount of the process parameters. The corrective control command is generated by overlaying the current baseline settings of the equipment in real time. This command is then sent to the underlying programmable logic controller (PLC) via the industrial fieldbus, directly driving the corresponding actuators. The system translates the virtual algorithm calculation results into physical equipment actions, correcting performance deviations caused by historical cumulative effects.

[0053] For example, Figure 4 This is a schematic diagram of the process parameter compensation distribution. The horizontal axis represents the sequence of material particles distributed along the length of the automotive steel wire, and the vertical axis represents the process parameter compensation amount generated by the control module through reverse projection. The obvious peak areas in the diagram correspond to the particle segments that produced large deviations in historical processes. The control system generates high-amplitude compensation commands for these specific segments, thereby achieving feedforward control to ensure the consistency of the performance of the entire automotive steel wire.

Claims

1. A smart optimization system for process parameters in automotive steel wire production, characterized in that, include: The data acquisition module is used to construct a Lagrange coordinate system and discretize automotive steel wire into material particles, collect time-series data and map it to the material particles to form a vector sequence of process parameters for the entire process. The deviation calculation module is used to obtain the state deviation trajectory based on the entire process parameter vector sequence. By calculating the deviation arc length of the state deviation trajectory relative to the ideal trajectory, the deviation modulus of the material particles at the current moment is obtained. The prediction module is used to obtain the damping equation based on the deviation modulus, and then use the damping equation to calculate the performance loss of the material particles; the performance loss is obtained in the following way: Multiply the standard evolution rate vector by the time step to obtain the theoretical performance improvement under the ideal damage-free state; calculate the memory dissipation factor using the natural exponential function, and determine the damping loss rate based on the memory dissipation factor; multiply the theoretical performance improvement by the damping loss rate to obtain the performance loss. The control module is used to generate the current process parameter compensation amount by inverse projection based on the preset historical regression gradient vector and performance loss amount, and to implement feedforward control of the equipment based on the process parameter compensation amount. The process parameter compensation amount satisfies: ； This is the compensation amount for process parameters; This is the compensation gain coefficient; This represents the historical regression gradient vector; The square of the magnitude of the historical regression gradient vector; for The performance loss at any given moment; The vector magnitude is used; the value of the compensation gain coefficient is adjusted on-site based on the response lag of the production equipment: for new heating or drawing equipment with rapid response, take... To achieve rapid compensation; for older equipment with significant inertial lag, take To prevent system oscillations caused by overcompensation.

2. The intelligent optimization system for process parameters in automotive steel wire production according to claim 1, characterized in that, The acquisition of time-series data and mapping it to material particles includes: The speed, temperature, tension, and mold compression ratio on the production line are collected in real time using encoders, infrared thermometers, and tension sensors; the collected time-series data are mapped onto material particles by integrating the speed.

3. The intelligent optimization system for process parameters in automotive steel wire production according to claim 1, characterized in that, The deviation modulus satisfies the expression: ； In the formula, for The deviation modulus of a material particle at any given moment; This refers to the initial moment when material particles enter the production line. Let be the integral variable, representing the time from the initial moment. At the time Any historical point in time between; For material particles at historical moments The actual strain rate; The ideal strain rate is set in the process standard for the corresponding process. It is the minimum of the maximum allowable operating rate of the equipment in the process and the allowable critical deformation rate of the material; For material particles at historical moments The actual instantaneous temperature; The ideal target temperature set in the process standard for the relevant process; This is the highest safe temperature allowed in the process. It is a time differential unit.

4. The intelligent optimization system for process parameters in automotive steel wire production according to claim 1, characterized in that, The performance loss satisfies the expression: ； In the formula, for The performance loss at any given moment; The standard evolution rate vector; for The deviation modulus at time; It is the time step; It is a natural exponential function; This represents the historical sensitivity coefficient.

5. The intelligent optimization system for process parameters in automotive steel wire production according to claim 4, characterized in that, The standard evolution rate vector is obtained as follows: Obtain the pre-stored standard process curve in the industrial control computer, and calculate the tangent slope of the standard process curve as the standard evolution rate vector.

6. The intelligent optimization system for process parameters in automotive steel wire production according to claim 1, characterized in that, The historical regression gradient vector is obtained as follows: Select the most recent Historical production data from each batch is used to obtain the input matrix of process parameter fluctuations and the output vector of performance status fluctuations. The slope of the linear regression equation between the input matrix and the output vector is calculated using online least squares regression, and this slope is used as the historical regression gradient vector. This is the preset batch quantity.

7. The intelligent optimization system for process parameters in automotive steel wire production according to claim 4, characterized in that, Set the historical sensitivity coefficient to .

8. The intelligent optimization system for process parameters in automotive steel wire production according to claim 1, characterized in that, The process parameter compensation amount for the current moment is generated by reverse projection, including: Calculate the pseudo-inverse of the historical regression gradient vector, and use the pseudo-inverse to map the performance loss back to the equipment process parameter space, generating the process parameter adjustment direction and magnitude that can produce an equivalent performance improvement.

Images