Automobile cable insulation layer eccentricity closed-loop feedback control method based on data analysis

Abstract

The invention relates to the field of cable insulation layer detection, in particular to an automobile cable insulation layer eccentricity closed-loop feedback control method based on data analysis, and the method comprises the steps: obtaining a multi-dimensional technological parameter and an original eccentricity vector in a production process; calculating a rheological fluctuation index of the insulating layer melt; performing adaptive filtering on the original eccentric vector to obtain a filtered eccentric vector; calculating the gravity thermal sag amount along the vertical direction; building a mold physical deviation vector based on the filtered eccentric vector and the gravity thermal sag amount; and calculating a mold displacement vector and controlling the mold to move, wherein the mold displacement vector comprises a feed-forward item in the same direction as the gravity thermal droop amount and a feedback item in the reverse direction of the mold physical deviation vector. The concentricity control precision and stability in the automobile cable production process are improved.

Description

Technical Field

[0001] This invention relates to the field of cable insulation layer inspection, and more particularly to a closed-loop feedback control method for the eccentricity of automotive cable insulation layers based on data analysis. Background Technology

[0002] As a crucial component of motor vehicles, the concentricity of the insulation layer of automotive cables is a core indicator determining their withstand voltage rating, mechanical life, and assembly safety. In high-speed extrusion production lines, insulation layer eccentricity is typically caused by die (head) position deviations. Existing control systems often employ a closed-loop system: X-ray diameter measurement—PID algorithm calculation—servo motor adjustment. However, traditional single-variable feedback control suffers from the following problems in practical engineering: First, there is interference from the rheological noise of the insulation layer melt. The rotation of the extruder screw, the breaking of the solid bed, and the changes in the flow resistance of the screen changer will generate disordered fluctuations in the insulation layer melt pressure at the die head. These fluctuations will cause high-frequency random fluctuations in the insulation layer thickness. Traditional PID controllers cannot distinguish between this fluid noise and die displacement, and often forcefully adjust the noise, causing frequent oscillations of the servo motor, accelerating equipment wear and destroying concentricity. Second, the nonlinear misleading effect of gravitational thermal sag. After the high-temperature molten insulation layer leaves the mold and before cooling and solidifying, it will collapse vertically downwards due to gravity. This process deformation has a complex nonlinear relationship with the traction speed, the temperature of the molten insulation layer, and its viscosity. Existing technologies typically ignore this factor, directly adjusting the mold based on the final measured values. For example, during production line acceleration, the increased linear speed naturally reduces the sag, which traditional controllers may misinterpret as upward mold displacement and adjust accordingly, leading to severe control overshoot and concentricity fluctuations. Therefore, there is an urgent need for an intelligent control method that can isolate fluid noise and process deformation from measurement data, restoring the true physical position of the mold. Summary of the Invention

[0003] To address the problem of low adjustment accuracy of the eccentricity control mechanism in existing automotive cable eccentricity control methods due to neglecting the rheological noise of the insulation melt and the sag effect caused by gravity, this invention provides a closed-loop feedback control method for automotive cable insulation eccentricity based on data analysis.

[0004] This invention provides a closed-loop feedback control method for the eccentricity of automotive cable insulation based on data analysis, employing the following technical solution: A data-driven closed-loop feedback control method for the eccentricity of automotive cable insulation includes the following steps: acquiring multi-dimensional process parameters and the original eccentricity vector during production; the multi-dimensional process parameters include insulation melt pressure, melt temperature, traction line speed, main motor current, and screw speed; calculating the rheological fluctuation index of the insulation melt, the magnitude of which is positively correlated with the standard deviation of the melt pressure within a preset sliding window and negatively correlated with the mean; adaptively filtering the original eccentricity vector to obtain the filtered eccentricity vector; calculating the gravitational thermal sag along the vertical direction, the magnitude of which is positively correlated with the melt temperature and negatively correlated with the traction line speed; constructing a mold physical deviation vector based on the filtered eccentricity vector and the gravitational thermal sag, the vertical component of which is the sum of the vertical component of the filtered eccentricity vector and the gravitational thermal sag; calculating the mold displacement vector and controlling the mold movement, the mold displacement vector including a feedforward term in the same direction as the gravitational thermal sag and a feedback term in the opposite direction to the mold physical deviation vector; the gain of the feedback term is negatively correlated with the rheological fluctuation index and positively correlated with the magnitude of the mold physical deviation vector.

[0005] Compared to traditional single-variable feedback control relying solely on PID algorithms, this invention proposes a data analysis control strategy that integrates multi-dimensional process parameters. By calculating the melt rheological fluctuation index of the insulation layer, it can effectively identify and distinguish between fluid noise and actual mold displacement, avoiding servo motor oscillation caused by forcibly adjusting high-frequency fluid noise. Simultaneously, by constructing a gravity-induced thermal sag model and combining it with the mold's physical deviation vector, the nonlinear process deformation caused by gravity is separated from the total deviation and compensated for through feedforward control. This solves the problem of misjudgment and reverse adjustment caused by ignoring gravity sag changes during production line acceleration and deceleration in existing technologies, improving the concentricity control accuracy and stability in automotive cable production.

[0006] Preferably, the expression for the rheological fluctuation index is:

[0007] in, express The rheological fluctuation index of the insulating layer melt at any given time; For length is The standard deviation of the insulation melt pressure within the sliding window; For length is The average value of the insulating layer melt pressure within the sliding window; for The molten temperature of the insulation layer is measured continuously. This is the standard temperature for cable processing.

[0008] By introducing a temperature correction factor into the pressure fluctuation calculation, not only the relationship between the pressure standard deviation and the mean is considered, but also the effect of temperature on viscosity. This allows the fluid to be identified as being in an unsteady state boundary in advance, even if the pressure variation coefficient has not deteriorated significantly when the measured temperature deviates from the standard value. This provides a robust reference index for the gain adjustment of the control system.

[0009] The preferred expression for adaptive filtering is:

[0010] in, for The eccentric vector after filtering at time step, for The eccentric vector after filtering at time step, for Dynamic filtering coefficients at any time for The original eccentric vector at time.

[0011] By introducing a dynamic filtering coefficient based on the rheological fluctuation index, the filtering intensity can be reduced when the extruder is operating stably to ensure rapid tracking of the actual minute displacement of the die; while the filtering intensity is automatically increased when severe pressure fluctuations occur, and the signal is smoothed using historical values, effectively filtering out high-frequency random noise and providing a clean and low-hysteresis data foundation for subsequent physical compensation.

[0012] Preferably, the expression for the dynamic filter coefficients is:

[0013] in, for Dynamic filtering coefficients at any time The preset base update rate, This is the filtering suppression coefficient. express The rheological fluctuation index of the insulating layer melt at a given time.

[0014] By establishing an inverse relationship between the dynamic filter coefficient and the rheological fluctuation index, the filter's inertia can be automatically adjusted according to the current fluid noise level. When the fluid fluctuation is large, the coefficient decreases, and the suppression capability is enhanced; when the fluid is stable, the coefficient increases, and the response speed is accelerated, thereby maximizing the system's dynamic response capability while ensuring signal quality.

[0015] Preferably, the base update rate ranges from [0.5, 0.7]; the filter suppression coefficient... The value range is [8.0, 12.0].

[0016] By limiting the specific numerical ranges of the base update rate and the filter suppression coefficient, the applicability of the algorithm in the extrusion scenario of automotive cable insulation was optimized. The setting of these empirical parameters ensures that the filter, under typical extruder operating conditions, neither masks the true deviation due to excessive smoothing nor introduces noise due to oversensitivity, thus guaranteeing the engineering practicality and stability of the control system.

[0017] Preferably, the expression for gravitational thermal sag is:

[0018] in, for The amount of thermal sag of the molten insulation layer at any given moment; The rheological constant of the insulating layer melt; The traction speed of the cable; This is the low-speed protection constant; The melting point of the insulating layer melt; for The molten temperature of the insulation layer is measured continuously. This is the relative viscosity index.

[0019] A rheological sag physical model incorporating low-speed boundary protection was established. Compared to existing techniques that ignore the effects of gravity, this formula accurately quantifies the nonlinear influence of traction speed, melt temperature, and relative viscosity on the amount of insulation sag. This allows the control system to calculate the "pseudo-bias" caused by gravity and independently compensate for it as a feedforward term in subsequent control, ensuring that the controller does not mistakenly interpret mold displacement when traction speed changes (such as acceleration leading to reduced sag).

[0020] Preferably, the relative viscosity index is the ratio of the main motor current to the screw speed.

[0021] Using the ratio of main motor current to screw speed as a relative viscosity index can intuitively reflect the change in melt flow resistance, enabling the gravity thermal sag model to adapt to viscosity changes caused by different batches of materials or process fluctuations, thus improving the adaptive capability of sag calculation.

[0022] Preferably, the expression for the mold displacement vector is:

[0023] in, for The displacement vector of the mold at any given moment; It is the gravity feedforward vector; for The physical deviation vector of the mold at any given time; The default base gain; express The rheological fluctuation index of the insulating layer melt at any given time; This is the error acceleration factor; This is the maximum permissible eccentricity threshold; Represents the magnitude of a vector.

[0024] By incorporating a gravity feedforward term, zero-hysteresis dynamic compensation for the influence of gravity is achieved. By incorporating an adaptive feedback term that includes an error acceleration coefficient and a rheological fluctuation index, intelligent adjustment is achieved, which is stable for small deviations, fast for large deviations, and slow for high noise levels. When a real, large-scale mechanical fault is detected, the gain can be automatically increased to quickly correct the deviation, while the gain can be automatically reduced to prevent oscillations when there is high fluid noise, effectively reducing the scrap rate.

[0025] Preferably, the feedforward term is ,in for The amount of gravitational thermal sag at any given moment. This represents the transpose of a vector.

[0026] By directly using the calculated gravitational thermal sag as the vertical compensation component, the mold can be pre-raised by a certain sag amount. This targeted physical compensation method fundamentally eliminates the interference of gravity on concentricity measurement, allowing the feedback loop to focus solely on handling the physical alignment deviation of the mold, thus reducing the burden on the feedback controller.

[0027] Preferably, the method for obtaining the original eccentricity vector is as follows: deploy an X-ray eccentricity meter between the extruder head and the cooling water tank, and collect the eccentricity data of the cable cross-section at a preset frequency to obtain the original eccentricity vector.

[0028] The present invention has the following technical effects: By constructing a rheological fluctuation index model and adaptively adjusting the filter strength and feedback gain, the interference of extruder fluid noise on servo control is effectively isolated. At the same time, a gravity thermal sag physical model is established, and nonlinear deformation caused by gravity is calculated and fed forward using multidimensional process parameters in real time. This method solves the problems of oscillation caused by fluid pulsation and misadjustment caused by changes in process parameters in traditional control, and improves the concentricity of the cable insulation layer. Attached Figure Description

[0029] Figure 1 This is a flowchart of the closed-loop feedback control method for the eccentricity of automotive cable insulation based on data analysis, as proposed in this invention.

[0030] Figure 2 This is a diagram showing the multidimensional rheological characteristic signal and adaptive filtering effect of the present invention. Detailed Implementation

[0031] 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.

[0032] This invention discloses a closed-loop feedback control method for the eccentricity of automotive cable insulation based on data analysis, referring to... Figure 1 The process includes the following steps, as detailed below: S1: Obtain multi-dimensional process parameters and eccentricity data.

[0033] A coordinate system is constructed with any point on the mold as the origin, the vertical direction as the y-axis, and the horizontal direction as the x-axis. The upward direction of the y-axis is the positive direction, and the cable travel direction is perpendicular to the plane formed by the x and y axes. An X-ray eccentricity measuring instrument is deployed between the machine head and the cooling water tank. The original eccentric vector of the frequency acquisition cable cross section This vector contains a horizontal eccentric component. and vertical eccentricity component The extruder PLC control system synchronously collects the melt pressure of the die head insulation layer. Insulating layer melt temperature Screw speed Main motor current and traction line speed Establish a length of The sliding window in this preferred embodiment This is used for real-time calculation of subsequent statistics, and the process parameters and eccentricity data are normalized using a linear normalization algorithm.

[0034] S2: Calculate the melt rheological fluctuation index of the insulating layer.

[0035] To accurately assess the stability of the current flow field, the fluctuation index needs to be calculated during equipment operation. The expression for the fluctuation index is as follows:

[0036] in, express The rheological fluctuation index of the insulating layer melt at any given time; For length is The standard deviation of the insulation melt pressure within the sliding window; For length is The average value of the insulating layer melt pressure within the sliding window; for The molten temperature of the insulation layer is measured continuously. This refers to the standard temperature for cable processing, such as 180°C. This is the temperature sensitivity coefficient, whose value is set according to the actual situation. For example, its value is 5.0.

[0037] In the formula, The term characterizes the coefficient of variation of pressure, directly reflecting the intensity of fluid pulsation. A large value indicates possible solid bed breakage or feeding fluctuations within the extruder, resulting in a low signal-to-noise ratio for the eccentric measurement. As a linear correction factor, it is used to appropriately amplify the fluctuation index when the temperature is abnormal. When the measured temperature deviates from the standard value, even if the coefficient of variation of the current pressure has not deteriorated significantly, the viscosity reference of the insulating layer melt has already drifted, indicating that the fluid is in an unsteady state boundary.

[0038] S3: Adaptive pre-filtering of the original eccentric vector based on the fluctuation index.

[0039] For the original eccentric vector The random jitter caused by pressure pulsation and the non-gravitational jitter in the vertical direction are included. A dynamic low-pass filter is established, with the following expression:

[0040] Among them, dynamic filter coefficients The expression is:

[0041] in, for The eccentricity vector after time-lapse filtering, including Horizontal eccentricity vector after time filtering and vertical eccentric vector , for The eccentric vector after filtering at time step, for Dynamic filtering coefficients at any time for The original eccentricity vector at time; The preset base update rate is 0.6. The filtering suppression coefficient is set to 10.0 in this embodiment. express The rheological fluctuation index of the insulating layer melt at a given time.

[0042] When the extruder is operating stably Approaching 0, Approximately equal to The filter has a fast response speed, enabling it to track the actual minute displacements of the mold in a timely manner; when severe pressure fluctuations occur, Increase, leading to The filter primarily relies on historical values ​​to reduce this. This allows for the filtering out of high-frequency random noise during the signal preprocessing stage, providing a clean data foundation for subsequent physical compensation.

[0043] S4: Calculate the gravity thermal sagging correction vector.

[0044] A rheological sag model incorporating low-speed boundary protection is constructed to estimate non-mold-related eccentricity caused by gravity during equipment operation. The calculation formula is as follows:

[0045] in, for The amount of thermal sag of the insulation layer at any given time, expressed in mm. is the rheological constant of the insulating layer melt, the value of which is obtained from field experiments. For example, for PVC material, the value is 0.05. The traction speed of the cable; The low-speed protection constant is set to 0.5 m / min to prevent the denominator from being zero. The melting point of the insulating layer melt; for The molten temperature of the insulation layer is measured continuously. The relative viscosity index is the value of the main motor current. With screw speed According to the principle of motor drive, the current of the main motor of the extruder is positively correlated with the load torque, and the load torque is determined by both the melt viscosity and the screw speed. When constant, if the main motor current An increase indicates an increase in melt flow resistance, i.e., an increase in viscosity; when the production line speed increases or decreases, the screw speed increases. When it changes, the main motor current This will also cause corresponding changes, affecting the main motor current. With screw speed The ratio remained within a certain range without significant change; therefore, the main motor current... With screw speed The ratio is an independent state index that is only related to the rheological properties of the material.

[0046] When traction speed When it rises, As the traction speed increases, the gravitational thermal sag decreases, which aligns with the a priori physical law that the faster the traction speed, the smaller the sag of the molten insulation layer; when the measured temperature of the molten insulation layer... The higher the viscosity, the greater the gravitational thermal sag, which aligns with the a priori physical law that higher temperatures lead to better fluidity and increased gravitational thermal sag; when viscosity... As the size increases, the material's resistance to deformation increases, and the amount of gravitational thermal sag decreases.

[0047] S5: Construct the physical deviation vector of the mold.

[0048] The mold deviation vector is constructed as follows: ;in, for The physical deviation vector of the mold at time t, Represents the transpose of a vector; for The horizontal eccentricity vector after filtering at time step, for The vertical eccentricity vector after filtering at time step. for The amount of thermal sag of the melted insulation layer at any given time. For example, the measured value. The insulation layer melt deflects downwards, and the calculated sag is... ,but This means that although the cable is off-center, the mold is centered and there is no offset. The downward deviation of the insulation melt is caused by gravity. The physical deviation vector of the mold eliminates the natural sag term caused by gravity.

[0049] S6: Calculate the displacement vector of the mold to control the eccentricity of the automotive cable insulation layer.

[0050] To suppress oscillations while ensuring a rapid response to large deviations, the displacement vector of the mold is calculated, as shown in the following expression:

[0051] in, for The displacement vector of the mold at any given moment; It is the gravity feedforward vector, vertically upward, with an amplitude of ,Right now , Represents the transpose of a vector; for The physical deviation vector of the mold at any given time; The preset base gain is 0.8 in this embodiment, which reflects the reference sensitivity of the control system. express The rheological fluctuation index of the insulating layer melt at any given time; This is the error acceleration factor, with a value of 2.0; The maximum permissible eccentricity threshold is determined by the cable manufacturing process; for example, it is 2 mm. Represents the magnitude of a vector.

[0052] As a feedforward term, the mold is directly raised by a sag to account for the inefficient effect caused by gravity; its value is based on the traction linear velocity. The calculation is done in real time. When the production line speeds up, causing the sag to decrease, this item automatically decreases, achieving zero-lag dynamic compensation for the impact of gravity.

[0053] This is an adaptive feedback term; it only addresses physical deviations after removing gravity. Adjustment is required; base gain The standard response strength of the control system to deviations under ideal laminar flow conditions is represented by the melt rheological fluctuation index of the insulating layer. This represents the current fluid noise level. When the melt is stable, the denominator approaches 1, maintaining full gain response. When the melt pulsates violently, the denominator increases, which will automatically reduce the gain. This represents the current actual mechanical deviation. When the physical deviation is small, normal production is possible. When the value is close to 1, stable production is maintained. When the deviation is large, the value increases, indicating a real mechanical failure, such as a loose mold. Then, the gain is automatically increased to correct the mold position as quickly as possible to avoid the generation of defective products.

[0054] Finally, the displacement vector is sent to the underlying motion controller to adjust the position of the mold, thereby controlling the eccentricity of the automotive cable insulation layer.

[0055] like Figure 2 As shown, a comparison is made between the original eccentric signal containing high-frequency noise and the signal processed by the adaptive filter of this invention. It can be clearly seen that in the region with a high fluctuation index, the filtered curve is smoother, indicating that the algorithm effectively suppresses fluid noise; in the region with a stable fluctuation index, the filtered curve closely follows the original signal without significant lag.

[0056] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A closed-loop feedback control method for the eccentricity of automotive cable insulation based on data analysis, characterized in that, The steps include: acquiring multidimensional process parameters and the original eccentricity vector during the production process; the multidimensional process parameters include the insulation layer melt pressure, melt temperature, traction line speed, main motor current, and screw speed; calculating the rheological fluctuation index of the insulation layer melt, the magnitude of which is positively correlated with the standard deviation of the melt pressure within the preset sliding window and negatively correlated with the mean; and performing adaptive filtering on the original eccentricity vector to obtain the filtered eccentricity vector. Calculate the gravitational thermal sag along the vertical direction, the magnitude of which is positively correlated with the melt temperature and negatively correlated with the traction linear velocity; construct the mold physical deviation vector based on the filtered eccentric vector and the gravitational thermal sag, the vertical component of which is the sum of the vertical component of the filtered eccentric vector and the gravitational thermal sag. The mold displacement vector is calculated and the mold movement is controlled. The mold displacement vector includes a feedforward term in the same direction as the gravitational thermal sag and a feedback term in the opposite direction to the mold physical deviation vector. The gain of the feedback term is negatively correlated with the rheological fluctuation index and positively correlated with the magnitude of the mold physical deviation vector.

2. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 1, characterized in that, The expression for the rheological fluctuation index is: in, express The rheological fluctuation index of the insulating layer melt at any given time; For length is The standard deviation of the insulation melt pressure within the sliding window; For length is The average value of the insulating layer melt pressure within the sliding window; for The molten temperature of the insulation layer is measured continuously. This is the standard temperature for cable processing.

3. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 1, characterized in that, The expression for adaptive filtering is: in, for The eccentric vector after filtering at time step, for The eccentric vector after filtering at time step, for Dynamic filtering coefficients at any time for The original eccentric vector at time.

4. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 3, characterized in that, The expression for the dynamic filter coefficients is: in, for Dynamic filtering coefficients at any time The preset base update rate, This is the filtering suppression coefficient. express The rheological fluctuation index of the insulating layer melt at any given time.

5. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 4, characterized in that, The base update rate ranges from [0.5, 0.7]; the filter suppression coefficient... The value range is [8.0, 12.0].

6. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 1, characterized in that, The expression for gravitational thermal sag is: in, for The amount of thermal sag of the molten insulation layer at any given moment; The rheological constant of the insulating layer melt; The pulling speed of the cable; This is the low-speed protection constant; The melting point of the insulating layer melt; for The molten temperature of the insulation layer is measured continuously. This is the relative viscosity index.

7. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 6, characterized in that, The relative viscosity index is the ratio of the main motor current to the screw speed.

8. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 1, characterized in that, The expression for the mold displacement vector is: in, for The displacement vector of the mold at any given moment; It is the gravity feedforward vector; for The physical deviation vector of the mold at any given time; The default base gain; express The rheological fluctuation index of the insulating layer melt at any given time; This is the error acceleration factor; This is the maximum permissible eccentricity threshold; Represents the magnitude of a vector.

9. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 8, characterized in that, Feedforward term is ,in for The amount of gravitational thermal sag at any given moment. This represents the transpose of a vector.

10. The closed-loop feedback control method for eccentricity of automotive cable insulation layer based on data analysis according to claim 1, characterized in that, The method for obtaining the original eccentricity vector is as follows: deploy an X-ray eccentricity meter between the extruder head and the cooling water tank, and collect the eccentricity data of the cable cross-section at a preset frequency to obtain the original eccentricity vector.

Images