Non-contact steady-state and deviation correction collaborative control method for amorphous alloy strip

Abstract

This invention discloses a non-contact, steady-state, and correction-coordinated control method for amorphous alloy strips, relating to automated control technology for amorphous strip winding processes. By acquiring signals of air-bearing pressure, strip position, temperature, and tension, the method utilizes an adaptive PID algorithm based on extended Kalman filtering in air-bearing steady-state control, combined with temperature compensation and tension disturbance feedforward, to achieve high-precision closed-loop regulation of the air-bearing pressure. In airflow correction control, a cascaded architecture of fuzzy control and predictive control is employed. Fuzzy control provides rapid response, while predictive control overcomes actuator delay through rolling optimization, significantly improving dynamic correction accuracy and speed. Experiments show that this invention can reduce steady-state fluctuations in air-bearing pressure by more than 80%, reduce the maximum strip position deviation from ±2mm to within ±0.5mm, shorten the dynamic response time by 30%, and reduce gas consumption by 20%~30%, providing a reliable solution for high-speed, high-precision amorphous strip winding.

Description

Technical Field

[0001] This invention relates to automated control technology for amorphous ribbon winding processes, specifically to a method for coordinated control of air-float steady-state and airflow correction of amorphous ribbon. Background Technology

[0002] Amorphous alloy strips, due to their unique atomic arrangement structure, possess excellent electromagnetic properties such as high magnetic permeability, low iron loss, and high saturation magnetic induction, and are widely used in power transformer cores, high-frequency inductors, sensors, and other fields. The winding process is a crucial step in the formation of amorphous strips, directly affecting the performance indicators of the final product. During the winding process, the strip needs to be wound onto the take-up rollers with constant tension and precise position control. Existing winding equipment typically uses mechanical roller contact correction, adjusting the position by laterally squeezing the strip edges. However, amorphous strips, due to their extremely thin thickness (typically 0.02mm-0.05mm), brittleness, and stress sensitivity, are easily damaged during high-speed winding processes.

[0003] During transport, the contact between hard rollers and brittle amorphous ribbon can easily cause edge chipping or micro-cracks on the surface, damaging its magnetic properties. The contact friction between the amorphous ribbon and the rollers can lead to sudden changes in local tension, which, for ultra-thin ribbons, can easily cause tensile deformation or even breakage. The mechanical correction mechanism of the winding machine has a large inertia, resulting in lag in response at high speeds and a tendency to overcorrect and oscillate, leading to uneven winding ends.

[0004] To address the aforementioned problems associated with contact winding of amorphous ribbons, non-contact support structures for amorphous ribbons are currently a key research focus. Non-contact support for amorphous ribbons is mainly divided into magnetic non-contact and gas non-contact methods. The principle of gas non-contact is to form a stable gas film between the ribbon and the winding roller, achieving non-contact support and guidance. However, under high-speed winding conditions, the air-bearing control system has some problems.

[0005] First, most air flotation control systems use simple PID control or fixed parameter feedback adjustment, which cannot effectively cope with the complex nonlinear dynamic changes during high-speed winding. When the winding speed changes, the ambient temperature fluctuates, or the strip tension is disturbed, the air flotation pressure is prone to large fluctuations, resulting in unstable air film thickness, which in turn causes scratches or wave deformation on the strip surface.

[0006] Secondly, the airflow correction system mainly relies on single position sensor detection and threshold switch control, lacking comprehensive perception and prediction capabilities of the strip's motion state. Under high-speed winding conditions, this simple control strategy struggles to achieve rapid and accurate correction, resulting in significant lateral positional deviations in the strip and affecting winding quality.

[0007] Third, there are multiple sources of interference in the actual production environment, including mechanical vibration, air pressure fluctuations, and ambient temperature changes. Existing control systems are not robust enough to these multi-source interferences, especially when dealing with complex disturbances such as tension fluctuations, and often cannot achieve effective feedforward compensation. Summary of the Invention

[0008] This invention addresses the shortcomings of existing technologies by providing a non-contact, steady-state and correction-coordinated control method for amorphous alloy strips. This method can maintain high-precision air-float steady-state of amorphous alloy strips and achieve rapid and accurate lateral position correction even under high-speed, wide-width, and variable operating conditions, thus solving the problem of easy damage to amorphous alloy strips caused by previous contact-based correction methods.

[0009] To solve the above problems, the present invention adopts the following technical solution: a non-contact steady-state and correction-coordinated control method for amorphous alloy strip, comprising the following steps: Step A: Signal acquisition, real-time acquisition of multi-point pressure signals of the air flotation system, strip lateral position signals, ambient temperature signals, and strip tension signals; Step B: Using the desired air flotation pressure... To achieve this goal, establish a system that includes air flotation pressure. and pressure change rate The state-space model is used to estimate the pressure state online using an extended Kalman filter; Step C: Adaptively adjust the proportional, integral, and derivative coefficients of the PID controller based on the estimated pressure deviation and the rate of change of deviation; Step D: Correct the desired pressure based on the current temperature using a pre-calibrated temperature compensation relationship; Step E: Estimate the equivalent disturbance to the air flotation pressure from the tension signal F using a robust extended Kalman filter. The disturbance is then fed forward and superimposed onto the output of the PID controller. Because the tension signal sampling frequency is inconsistent with the air flotation pressure control frequency, the disturbance estimate differs between two adjacent tension samples. Keeping the most recent update value unchanged, the air flotation pressure control signal is updated again by the robust extended Kalman filter after the new tension data arrives. Step F: Determine the desired lateral position of the strip. Calculate the positional deviation for the target. The rate of change of the strip and its lateral kinematics are used as inputs to the fuzzy control algorithm to generate the basic airflow regulation amount; Step G: Establish an incremental state-space model of the strip's lateral kinematics, predict the strip position for multiple future control cycles based on this model, and minimize the objective function of the position tracking error and control increment weighting through rolling optimization to obtain the optimal control increment sequence. The first term of this sequence is superimposed with the basic airflow regulation amount to obtain the final airflow regulation command; the final airflow regulation command , This is the basic airflow regulation output of the fuzzy controller. This increment is defined relative to the final control value at the previous moment. The correction amount; Step H: Drive the air flotation air supply proportional valve according to the air flotation pressure control signal, and drive the lateral airflow adjustment nozzle according to the final airflow adjustment command to realize the coordinated control of air flotation steady state and strip position correction.

[0010] In the preferred embodiment of the above technical solution, the state-space model in step B specifically comprises: state equations. ; Observation equation, ;in, , This is the control signal for the proportional valve. It's process noise. Here, A is the measurement noise, B is the state transition matrix, and C is the input matrix.

[0011] In the preferred embodiment of the above technical solution, the expression for adaptively adjusting the PID controller parameters in step C is: , , ,in, It is the air flotation pressure deviation at time k. , , For the desired air flotation pressure, This is the current pressure measurement value. , , As the baseline PID parameters, , , and This is a preset positive coefficient.

[0012] In the preferred embodiment of the above technical solution, the temperature compensation relationship in step D is obtained by fitting the temperature-pressure characteristic curve under standard operating conditions: The revised expected pressure is: ,in, Refers to the current ambient temperature. For reference temperature, This is the current temperature. This is the original expected air flotation pressure without temperature compensation. The compensation amount for the pressure setpoint caused by temperature. The desired air flotation pressure after temperature correction. and The fitting coefficients are denoted as .

[0013] In the preferred embodiment of the above technical solution, the disturbance is fed forward to the PID output in step E: ,in The feedforward gain is derived from the balance between buoyancy and strip tension: , that is, the negative value of the partial derivative of pressure with respect to tension, characterizes the amount of pressure required to compensate for a unit change in tension. The raw control quantity calculated by the PID controller. This is the air float pressure control signal that ultimately acts on the proportional valve.

[0014] In the preferred embodiment of the above technical solution, the fuzzy control algorithm in step F employs Mamdani-type inference, with input variable position deviation... and rate of change of deviation and output variable basic airflow regulation amount Each set is divided into 7 fuzzy subsets, and the fuzzy rule is in the form of: If for ,and for ,but for ,in This refers to the lateral position deviation of the strip. , For the desired lateral position of the strip, The sensor measures the lateral position of the strip. The rate of change of deviation , The basic airflow regulation output by the fuzzy controller, where , and For predefined fuzzy language values, the centroid method is used for defuzzification.

[0015] In the preferred embodiment of the above technical solution, the incremental state-space model of the strip's lateral kinematics in step G is as follows: , ,in Include , This includes the actual lateral position of the strip measured by the sensor. Lateral drift velocity of the strip A is the airflow regulation increment; A is the state transition matrix, B is the input matrix, and C is the observation matrix; the objective function is: ,in, For the prediction time domain, i.e., the number of steps to predict forward, To control the time domain, i.e., the number of steps in which the control variable is allowed to change, This refers to the increment of airflow regulation, i.e., the change in control quantity between adjacent sampling periods. Let be the predicted value of the lateral position of the strip at time k for the (k+i)th future step. For the desired lateral position, Let be the weight matrix of the output error in step i. The weight matrix represents the control increment at step j; the objective function has input increment constraints. and output constraints ;in, For the maximum allowable control increment, Maximum permissible lateral position offset; prediction time domain and control time domain Based on the current winding speed Adaptive selection: , ,in This refers to the physical distance between the strip position detection point and the correction nozzle. The current winding speed, This is the sampling period of the control system.

[0016] In the above technical solution, preferably, in step A, the sampling frequency of the multi-point pressure signal is 1kHz, the sampling frequency of the strip position signal is 500Hz, and the sampling frequency of the temperature and tension signals is 10Hz.

[0017] This application deploys multiple pressure sensors, strip lateral position sensors, ambient temperature sensors, and tension sensors on the winding equipment to acquire multi-point air flotation pressure signals, strip lateral position signals, ambient temperature signals, and strip tension signals in real time. Then, with a preset desired air flotation pressure as the target, a system state-space model is established with air flotation pressure and pressure change rate as state variables. An extended Kalman filter is used to estimate the actual state of the current air flotation system online. Based on the estimated pressure deviation and its change, the proportional, integral, and derivative gains of the PID controller are adjusted in real time according to a preset adaptive law. Simultaneously, the desired air flotation pressure is corrected based on the measured ambient temperature using a pre-calibrated temperature compensation relationship. This application also uses a robust extended Kalman filter to estimate the equivalent disturbance of tension fluctuations on the air flotation pressure online from the tension signal, and directly superimposes this disturbance estimate onto the output of the PID controller in a feedforward manner to obtain the final air flotation supply proportional valve control signal.

[0018] This application takes the desired lateral position of the strip as the central objective. First, it calculates the current position deviation and the rate of change of deviation, which are used as inputs to a fuzzy controller. Fuzzy inference generates a basic airflow regulation quantity with fast response capability. At the same time, an incremental state-space model describing the lateral motion characteristics of the strip is established. Based on this model, the strip position in the future multiple control cycles is predicted. In each sampling cycle, a quadratic programming problem is solved online with the objective function of minimizing the weighted average of position tracking error and control increment as the objective. A set of optimal control increment sequences is obtained. Only the first term of the sequence is taken as the optimized increment, which is added to the aforementioned basic airflow regulation quantity to form the final airflow regulation command.

[0019] This application outputs an air flotation pressure control signal to the air flotation proportional valve to adjust the air supply pressure within the air flotation pad. Then, it outputs a final airflow adjustment command to the lateral airflow nozzles on both sides of the strip, changing the airflow velocity and direction acting on the strip's sides, thereby generating the required lateral correction force. These two control paths are executed in parallel within the same sampling period, achieving coordinated control of the air flotation steady-state and strip position correction.

[0020] This application's air-float steady-state control section uses an extended Kalman filter (EDF) as the core state observation tool, solving the problem that traditional PID controllers cannot perceive the dynamic changes within the system. The EDF, through iterative calculation of state prediction and measurement updates, can reconstruct the true pressure and its changing trend from noisy pressure measurements, providing an accurate basis for adaptive parameter adjustment. When the pressure deviation or its change is large, the system automatically increases the proportional and derivative gains to enhance damping and quickly suppress fluctuations; when the deviation decreases to near steady state, the integral action is appropriately enhanced to eliminate residual steady-state error. This mechanism fundamentally solves the problem of decreased control accuracy caused by dynamic characteristic drift due to changes in operating conditions in air-float systems.

[0021] This application is designed to address the physical properties of gas viscosity changing with temperature. By changing the ambient temperature in a constant temperature chamber in advance and recording the optimal air flotation pressure setting value, a temperature-pressure compensation curve is fitted. During online operation, the desired pressure is corrected in real time based on this curve, avoiding steady-state deviation caused by the decrease in air film stiffness due to temperature increase.

[0022] This application utilizes a robust extended Kalman filter, which exhibits better numerical stability in handling outliers and model uncertainties compared to the standard extended Kalman filter. The filter takes the tension sensor signal as an external input, estimating its impact on the air flotation pressure as an equivalent time-varying disturbance state. The control signal is then injected in a negative feedforward manner, enabling the air flotation system to quickly maintain pressure stability even during large tension fluctuations, thus reducing the undesirable coupling between air flotation and tension.

[0023] Furthermore, this application employs a cascaded structure of fuzzy control and predictive control. The fuzzy controller directly uses the position deviation and its rate of change as input, and outputs a basic correction quantity based on a pre-summarized operational experience rule base, ensuring that the system reacts almost without delay when any deviation occurs. Simultaneously, this application introduces a model predictive controller, which utilizes an online identified strip lateral kinematics model to explicitly incorporate this hysteresis distance and excitation response characteristics into the optimization framework. At each sampling moment, it calculates the optimal airflow adjustment sequence for a future prediction time domain. Through reasonable constraints and smooth weighting of the control increment, it generates an optimized compensation increment that simultaneously satisfies the requirements of rapid response and control stability. This superimposes the rapid response capability of the fuzzy output with the look-ahead optimization capability of predictive control. The prediction and control time domains are adaptively calculated in real time based on the current winding speed; the faster the speed, the more prediction steps are required, ensuring that the optimization always covers the complete time period from detection to execution.

[0024] Compared with existing technologies, this application, based on the synergistic effect of adaptive PID with extended Kalman filtering, temperature compensation, and tension feedforward, reduces the steady-state fluctuation of air flotation pressure from ±0.15MPa to within ±0.03MPa, improving the steady-state accuracy by about 80%, thus significantly improving the steady-state accuracy of air flotation.

[0025] The fuzzy control in this application provides a zero-delay basic response and predictive control compensates for execution lag, which shortens the convergence time of strip position deviation by about 30%, reduces the maximum lateral position deviation from ±2mm to within ±0.5mm, and improves the correction accuracy by about 75%, resulting in a significantly faster and more accurate dynamic response for correction.

[0026] The temperature compensation and perturbation feedforward based on robust extended Kalman filtering in this application effectively suppress the interference of ambient temperature changes and strip tension fluctuations on the steady state of air flotation, and have strong anti-interference ability.

[0027] This application mainly relies on sensors and industrial controllers, with low equipment modification costs. Furthermore, due to more precise and efficient air flotation pressure and airflow regulation, gas consumption is reduced by 20%–30%, and operating costs are reduced by 10%–15%, significantly improving the overall economic benefits of the production line. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be discussed below. Obviously, the technical solutions described in conjunction with the accompanying drawings are only some embodiments of the present invention. For those skilled in the art, other embodiments and their accompanying drawings can be obtained based on the embodiments shown in these drawings without creative effort.

[0029] Figure 1This is a flowchart of the control method steps of the present invention. Detailed Implementation

[0030] The technical solutions of various embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments described in 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.

[0031] Example 1: As Figure 1 As shown, a non-contact steady-state and correction-coordinated control method for amorphous alloy strip is applied to a winding machine for amorphous alloy strip. The winding machine includes a take-up roller and unwinding mechanism driven by a servo motor, a correction device, a tension roller, and a deviation detection device. It also includes an array of micro-holes on each roller, which create air cushions on the rollers; an air-blowing correction system on both sides of the correction device; and a laser displacement sensor on the deviation detection device. Multiple piezoresistive pressure sensors, tension sensors, and ambient temperature sensors are also present along the amorphous alloy strip's travel path. The controller uses an embedded industrial computer.

[0032] The process includes the following steps: Step A: Signal acquisition. Multiple air flotation pressure signals are simultaneously acquired at a frequency of 1kHz, and the average value is taken as the current air flotation pressure value. Laser displacement sensor signals are acquired at a frequency of 500Hz to obtain the lateral position of the strip. Ambient temperature and strip tension are acquired at a frequency of 10Hz. All signals are sent to the controller after passing through a low-pass anti-aliasing filter.

[0033] Step B: First, based on the calibrated second-order state-space model of the air flotation system, perform one prediction and update using extended Kalman filtering within each 1ms control cycle to obtain the optimal state estimate. The desired air flotation pressure is then determined. To achieve this goal, establish a system that includes air flotation pressure. and pressure change rate The state-space model is used to estimate the pressure state online using an extended Kalman filter; in step B, the state-space model specifically consists of the state equations. ; Observation equation, ;in, , This is the control signal for the proportional valve. It's process noise. It measures noise.

[0034] Step C: Then, based on the estimated current pressure deviation and the change in deviation, update the three gains of the PID controller and calculate the PID controller output. The expression for adaptively adjusting the PID controller parameters in step C is: , , ,in, , , For the desired air flotation pressure, This is the current pressure measurement value. , , As the baseline PID parameters, , , and This is a preset positive coefficient.

[0035] Step D: Simultaneously, using the temperature compensation curve pre-stored in the controller, calculate the pressure correction corresponding to the current temperature and add it to the setpoint. This is obtained by fitting the temperature-pressure characteristic curve under standard operating conditions. The revised expected pressure is: ,in, For reference temperature, This is the current temperature. The compensation amount for the pressure setpoint caused by temperature. The desired air flotation pressure after temperature correction. and The fitting coefficients are denoted as .

[0036] Step E: Estimate the equivalent perturbation to the air flotation pressure from the tension signal F using a robust extended Kalman filter. , ,in The feedforward gain is derived from the balance between buoyancy and strip tension: , that is, the negative value of the partial derivative of pressure with respect to tension, characterizes the amount of pressure required to compensate for a unit change in tension. The raw control quantity calculated by the PID controller. This is the final air float pressure control signal applied to the proportional valve. This disturbance is then superimposed onto the output of the PID controller using a feedforward method to obtain the air float pressure control signal. The disturbance estimate is multiplied by the feedforward gain determined by the buoyancy force-tension balance relationship, then added to the PID output, and finally output to the air flotation proportional valve.

[0037] Step F: Determine the desired lateral position of the strip Calculate the positional deviation for the target. The positional deviation and its rate of change are used as inputs to the fuzzy control algorithm to generate the basic airflow regulation. The fuzzy control algorithm employs Mamdani-type inference, with the input variable being the positional deviation. and rate of change of deviation and output variable basic airflow regulation amount Each set is divided into 7 fuzzy subsets, and the fuzzy rule is in the form of: If for ,and for ,but for ,in This refers to the lateral position deviation of the strip. , For the desired lateral position of the strip, The sensor measures the lateral position of the strip. The rate of change of deviation , The basic airflow regulation output by the fuzzy controller, where , and For predefined fuzzy language values, the centroid method is used for defuzzification.

[0038] The Mamdani-type fuzzy controller is a controller based on fuzzy logic. In this application, the Mamdani-type fuzzy controller is used in the front-end of airflow correction control. Its task is to quickly provide a reasonable airflow adjustment direction and approximate amplitude based on the lateral position deviation of the strip and its rate of change, thus providing a starting point for subsequent predictive control optimization.

[0039] The two inputs to the fuzzy controller are: and The fuzzy universes of both input and output variables are normalized to [-1, 1].

[0040] Among them, the 7 fuzzy subsets are: Based on 7 fuzzy subsets having 49 rules, among which , , These are fuzzy subsets on the corresponding universes of discourse. The centroid method is... ,in, To output the number of discrete points in the universe of discourse, For the first The universe value of a point This represents the membership degree of the point.

[0041] Step G: Establish an incremental state-space model of the strip's lateral kinematics. Based on this model, predict the strip's position for multiple future control cycles. Minimize the objective function weighted by the position tracking error and control increments through rolling optimization to obtain the optimal control increment sequence. Take the first term of this sequence and superimpose it with the basic airflow adjustment amount to obtain the final airflow adjustment command. In step G, the incremental state-space model of the strip's lateral kinematics is: , ,in Include , This includes the actual lateral position of the strip measured by the sensor. Lateral drift velocity of the strip The airflow adjustment increment; the objective function is: ,in, For the prediction time domain, i.e., the number of steps to predict forward, To control the time domain, i.e., the number of steps in which the control variable is allowed to change, This refers to the increment of airflow regulation, i.e., the change in control quantity between adjacent sampling periods. Let be the predicted value of the lateral position of the strip at time k for the (k+i)th future step. For the desired lateral position, Let be the weight matrix of the output error in step i. The weight matrix represents the control increment at step j; the objective function has input increment constraints. and output constraints ;in, For the maximum allowable control increment, Maximum permissible lateral position offset; prediction time domain and control time domain Based on the current winding speed Adaptive selection: , ,in This refers to the physical distance between the strip position detection point and the correction nozzle. The current winding speed, This is the sampling period for the control system. The final airflow adjustment command is: Control increments in predictive models It is the final airflow adjustment amount actually performed relative to the previous sampling time. In other words, the output of the fuzzy controller To provide a reference bias at the current moment for predictive control, the optimal increment is obtained through optimization. When superimposed on this benchmark, the final control quantity that satisfies both incremental and output constraints is formed.

[0042] Step H: Drive the air flotation proportional valve according to the air flotation pressure control signal, and drive the lateral airflow adjustment nozzle according to the final airflow adjustment command to achieve coordinated control of air flotation steady state and strip position correction. The air flotation pressure control loop operates at a frequency of 1kHz, completing the calculations of steps B to E and updating the proportional valve drive signal in each control cycle; the airflow correction control loop operates at a frequency of 500Hz, triggered by the position signal update time, completing the calculations of steps F to G and updating the nozzle drive signal; the two controls are executed asynchronously at the time of each free sensor signal drive, and the output signal is synchronously sent to the execution unit through the same controller to achieve coordinated control of air flotation steady state and strip position correction.

Claims

1. A non-contact, steady-state and correction-coordinated control method for amorphous alloy strips, characterized by: Includes the following steps: Step A: Signal acquisition, real-time acquisition of multi-point pressure signals, strip lateral position signals, ambient temperature signals, and strip tension signals of the air flotation system; Step B: At the desired air flotation pressure To achieve this goal, establish a system that includes air flotation pressure. and pressure change rate The state-space model is used to estimate the pressure state online using an extended Kalman filter; Step C: Adaptively adjust the proportional, integral, and derivative coefficients of the PID controller based on the estimated pressure deviation and the rate of change of deviation; Step D: Using a pre-calibrated temperature compensation relationship, adjust the desired pressure according to the current temperature; Step E: Estimate the equivalent perturbation to the air flotation pressure from the tension signal F using a robust extended Kalman filter. The disturbance is then fed forward and superimposed onto the output of the PID controller. Because the tension signal sampling frequency is inconsistent with the air flotation pressure control frequency, the disturbance estimate differs between two adjacent tension samples. Keeping the most recent update value unchanged, the air flotation pressure control signal is updated again by the robust extended Kalman filter after the new tension data arrives. ; Step F: Determine the desired lateral position of the strip Calculate the positional deviation for the target. Its rate of change, as input to the fuzzy control algorithm, generates the basic airflow regulation amount; Step G: Establish an incremental state-space model of the strip's lateral kinematics. Based on this model, predict the strip's position for multiple future control cycles. Minimize the objective function weighted by the position tracking error and control increments through rolling optimization to obtain the optimal control increment sequence. Take the first term of this sequence and superimpose it with the basic airflow adjustment amount to obtain the final airflow adjustment command. , This is the basic airflow regulation output of the fuzzy controller. This increment is defined relative to the final control value at the previous moment. The amount of correction; Step H: Drive the air flotation proportional valve according to the air flotation pressure control signal, and drive the lateral airflow adjustment nozzle according to the final airflow adjustment command to achieve coordinated control of air flotation steady state and strip position correction.

2. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, in The state-space model in step B is as follows: Equations of state ; Observation equations ; in, , Let A be the proportional valve control signal, B be the state transition matrix, B be the input matrix, and C be the observation matrix. It's process noise. It measures noise.

3. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, in The expression for adaptively adjusting the PID controller parameters in step C is: ， ， ， in, It is the air flotation pressure deviation at time k. , , For the desired air flotation pressure, This is the current pressure measurement value. , , As the baseline PID parameters, , , and This is a preset positive coefficient.

4. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, In step D, the temperature compensation relationship is obtained by fitting the temperature-pressure characteristic curve under standard operating conditions: ， The revised expected pressure is: ， in, Refers to the current ambient temperature. For reference temperature, This is the current temperature. This is the original expected air flotation pressure without temperature compensation. The compensation amount for the pressure setpoint caused by temperature. The desired air flotation pressure after temperature correction. and The fitting coefficients are denoted as .

5. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, in In step E, the disturbance is fed forward to the PID output: ， in The feedforward gain is derived from the balance between buoyancy and strip tension: , that is, the negative value of the partial derivative of pressure with respect to tension, characterizes the amount of pressure required to compensate for a unit change in tension. The raw control quantity calculated by the PID controller. This is the air float pressure control signal that ultimately acts on the proportional valve.

6. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, in In step F, the fuzzy control algorithm employs Mamdani-type inference, with input variable position deviation. and rate of change of deviation and output variable basic airflow regulation amount Each set is divided into 7 fuzzy subsets, and the fuzzy rule is in the form of: like for ,and for ,but for , in This refers to the lateral position deviation of the strip. , For the desired lateral position of the strip, The sensor measures the lateral position of the strip. The rate of change of deviation , This is the basic airflow regulation output of the fuzzy controller. in , and For predefined fuzzy language values, the centroid method is used for defuzzification.

7. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, in The incremental state-space model of the strip's lateral kinematics in step G is as follows: ， ， in Include , This includes the actual lateral position of the strip measured by the sensor. Lateral drift velocity of the strip Let A be the airflow regulation increment, B be the state transition matrix, C be the input matrix, and C be the observation matrix. The objective function is: ， in, For the prediction time domain, i.e., the number of steps to predict forward, To control the time domain, i.e., the number of steps in which the control variable is allowed to change, This refers to the increment of airflow regulation, i.e., the change in control quantity between adjacent sampling periods. Let be the predicted value of the lateral position of the strip at time k for the (k+i)th future step. For the desired lateral position, Let be the weight matrix of the output error in step i. The weight matrix for controlling the increment in step j; The objective function has an input increment constraint. and output constraints ; in, For the maximum allowable control increment, The maximum permissible lateral position offset; Prediction Time Domain and control time domain Based on the current winding speed Adaptive selection: ， ， in This refers to the physical distance between the strip position detection point and the correction nozzle. The current winding speed, This is the sampling period of the control system.

8. The non-contact steady-state and correction-coordinated control method for amorphous alloy strips according to claim 1, characterized in that, in In step A, the sampling frequency of the multi-point pressure signal is 1kHz, the sampling frequency of the strip position signal is 500Hz, and the sampling frequency of the temperature and tension signals is 10Hz.

Images