Mold cycle liquid level control method, storage medium and computer device

Abstract

The application discloses a crystallizer cycle liquid level control method and a storage medium and a computer device, and the method comprises the following steps: converting a frequency characteristic vector of liquid level fluctuation data into a time domain signal, inputting a PID controller after pre-compensating the time domain signal through a feedforward controller; when the liquid level of molten steel in a target crystallizer is controlled based on the PID controller and the pre-compensated time domain signal, the parameters of the PID controller are adjusted in real time based on a fuzzy parameter regulator until the liquid level adjustment execution structure is regulated based on the PID controller to control the liquid level of the molten steel in the crystallizer to meet the control requirement. By adding the fuzzy parameter regulator in the PID control architecture and formulating fuzzy rules based on field big data analysis, the control parameters are changed in real time, different control effects can be achieved in different control stages, and the fuzzy feedforward parameter regulator is added, the output of the feedforward controller is adjusted according to the change speed of the signal, and the control accuracy of the liquid level fluctuation is improved.

Description

Technical Field

[0001] This application relates to the field of metallurgical continuous casting technology, and in particular to a method for controlling the liquid level in a crystallizer cycle, a storage medium, and a computer device. Background Technology

[0002] In modern steel production processes, continuous casting plays a crucial and irreplaceable role. As one of the core processes in steel manufacturing, continuous casting technology has greatly improved production efficiency and product quality. However, the complexity of the continuous casting process also brings many challenges, among which the surface defects of continuously cast billets are particularly prominent.

[0003] Numerous studies and practical experience have shown that fluctuations in the molten liquid level in the crystallizer have a crucial impact on the surface quality of continuously cast billets. Abnormal fluctuations in the molten liquid level often lead to a series of defects such as inclusions and surface cracks, severely damaging product quality. Furthermore, periodic abnormal fluctuations in the molten liquid level are among the most common anomalies in continuous casting, attracting widespread attention and in-depth discussion from researchers in the metallurgical field.

[0004] Traditional control systems fall short in the face of this challenge. Due to technological limitations, traditional systems often cannot respond quickly and accurately to periodic abnormal fluctuations in the liquid level of the crystallizer, thus failing to effectively solve the problem and significantly impacting production quality and efficiency. Summary of the Invention

[0005] In view of this, this application provides a crystallizer periodic liquid level control method, storage medium, and computer equipment. The method converts the frequency feature vector of liquid level fluctuation data into a time-domain signal. This time-domain signal is then fed forward compensated by a feedforward controller and input to a PID controller. When liquid level control is implemented on the molten steel in the target crystallizer based on the PID controller and the feedforward compensated time-domain signal, the PID controller parameters are adjusted in real time based on a fuzzy parameter regulator until the liquid level adjustment execution structure based on the PID controller controls the molten steel level in the crystallizer to meet the control requirements. By adding a fuzzy parameter regulator to the PID control architecture and formulating fuzzy rules based on on-site big data analysis, the control parameters can be changed in real time, achieving different control effects at different control stages. Furthermore, the addition of a fuzzy feedforward parameter regulator adjusts the output of the feedforward controller according to the rate of signal change, improving the accuracy of liquid level fluctuation control.

[0006] According to one aspect of this application, a method for controlling the liquid level in a crystallizer during a cycle is provided, the method comprising:

[0007] Collect data on the fluctuation of the liquid level of molten steel in the target crystallizer, wherein the fluctuation of the liquid level of molten steel is periodic, and the liquid level fluctuation data includes the liquid level of molten steel at each time point within a preset control period based on a preset acquisition frequency;

[0008] The liquid surface fluctuation data is subjected to a fast Fourier transform to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes. The initial frequency feature vector is then filtered to obtain a frequency feature vector.

[0009] The frequency feature vector is converted into a time-domain signal, and after the time-domain signal is fed forward by the feedforward controller, it is input into the PID controller. The PID control system includes a feedforward controller, a PID controller, and a fuzzy parameter regulator.

[0010] When the liquid level of the molten steel in the target crystallizer is controlled based on the time-domain signal after feedforward compensation by the PID controller, the parameters of the PID controller are adjusted in real time based on the fuzzy parameter regulator until the liquid level adjustment execution structure of the PID controller is adjusted to control the liquid level of the molten steel in the crystallizer to meet the control requirements.

[0011] Optionally, the step of performing feedforward compensation on the time-domain signal via a feedforward controller includes:

[0012] The time-domain signal is processed by a first-order circuit using a feedforward controller, and the processed time-domain signal is differentiated to obtain the rate of change of the time-domain signal.

[0013] The output of the feedforward controller is adjusted according to the rate of change.

[0014] Optionally, the real-time adjustment of the PID controller parameters based on the fuzzy parameter regulator includes:

[0015] Configure the initial parameters of the PID controller and set the output domains of the proportional coefficient, integral coefficient, and derivative coefficient of the fuzzy parameter adjuster. The initial parameters include the proportional coefficient kp, integral coefficient ki, and derivative coefficient kd. The output domain of the proportional coefficient is [-0.7, 0.7], the output domain of the integral coefficient is [-1.1, 1.1], and the output domain of the derivative coefficient is [-0.21, 0.21].

[0016] Based on the amplitude and direction of the input time-domain signal, the fuzziness level of the input time-domain signal is determined, wherein the fuzziness level includes negative giant, negative medium, negative small, zero, positive small, positive medium, and positive giant.

[0017] Based on preset fuzzy rules and the fuzzy level, the fuzzy output of the PID controller parameters is calculated, wherein the preset fuzzy rules include the PID controller parameter adjustment methods corresponding to different fuzzy levels;

[0018] After defuzzifying the fuzzy output, the obtained proportional coefficient error value Δkp, integral coefficient error value Δki, and differential coefficient error value Δkd are added to the initial parameters of the PID controller, so that the PID controller can adjust the liquid level adjustment actuator based on the latest parameters.

[0019] Optionally, the step of controlling the liquid level of the molten steel in the target crystallizer based on the PID controller and the time-domain signal after feedforward compensation includes:

[0020] The output of the PID controller is sent as a control signal to the liquid level adjustment actuator, so that the liquid level adjustment actuator adjusts the inflow of molten steel in the crystallizer according to the control signal.

[0021] The molten steel level is detected in real time by a sensor, and the detected molten steel level value is returned to the PID control system as a feedback signal. The PID control system then continues to adjust the control signal based on the difference between the feedback signal and the molten steel level setpoint.

[0022] Optionally, performing a fast Fourier transform on the liquid surface fluctuation data to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes includes:

[0023] According to the Fast Fourier Transform (FFT) formula, the liquid surface fluctuation data is subjected to a Fast Fourier Transform to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes. The Fast Fourier Transform formula is as follows:

[0024]

[0025] Y(i) represents the initial frequency feature vector obtained after performing a Fast Fourier Transform, where i represents the frequency index, K represents the length of the liquid surface fluctuation data, and n represents the acquisition time index of the liquid surface fluctuation data. even (n) represents the even-numbered sequence in the liquid level fluctuation data, y odd (n) represents the odd sequence in the liquid surface fluctuation data, j represents the imaginary unit, and k represents the ordinal number of the acquisition time node.

[0026] Optionally, the step of filtering the initial frequency feature vector to obtain the frequency feature vector includes:

[0027] In the initial frequency feature vector, frequencies greater than a preset frequency threshold are removed to obtain the frequency feature vector.

[0028] Optionally, after filtering the initial frequency feature vector to obtain the frequency feature vector, the method further includes:

[0029] Based on the frequency feature vector, the dominant frequency affecting the fluctuation of the molten steel surface is determined, and the amplitude of the molten steel surface fluctuation is determined based on the determined dominant frequency.

[0030] The molten steel level fluctuation is determined based on the amplitude of the molten steel level fluctuation, and the molten steel level in the crystallizer is controlled based on the molten steel level fluctuation to meet the control requirements.

[0031] Optionally, the liquid level fluctuation state includes abnormal and normal states. Determining the liquid level fluctuation state based on the amplitude of the molten steel level fluctuation, and controlling the molten steel level in the crystallizer based on the liquid level fluctuation state to achieve the control requirements, includes:

[0032] When the fluctuation range of the molten steel surface is outside the preset normal fluctuation range, the molten steel surface fluctuation is considered abnormal; when the fluctuation range of the molten steel surface is within the preset normal fluctuation range, the molten steel surface fluctuation is considered normal.

[0033] If the fluctuation of the molten steel level is abnormal, the liquid level adjustment mechanism will be activated until the fluctuation of the molten steel level in the crystallizer is within the preset normal fluctuation range.

[0034] According to another aspect of this application, a storage medium is provided that stores a computer program thereon, which, when executed by a processor, implements the above-described crystallizer periodic liquid level control method.

[0035] According to another aspect of this application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor executes the program to implement the above-described crystallizer periodic liquid level control method.

[0036] By employing the above technical solution, this application provides a crystallizer periodic liquid level control method, storage medium, and computer equipment. The method converts the frequency characteristic vector of liquid level fluctuation data into a time-domain signal. This time-domain signal is then fed forward compensated by a feedforward controller before being input into a PID controller. When liquid level control is implemented on the molten steel in the target crystallizer based on the PID controller and the feedforward compensated time-domain signal, the PID controller parameters are adjusted in real time using a fuzzy parameter regulator until the liquid level adjustment execution structure, based on the PID controller, controls the molten steel level in the crystallizer to meet the control requirements. By adding a fuzzy parameter regulator to the PID control architecture and formulating fuzzy rules based on on-site big data analysis, the control parameters can be changed in real time, achieving different control effects at different control stages. Furthermore, the addition of a fuzzy feedforward parameter regulator adjusts the output of the feedforward controller according to the rate of signal change, improving the accuracy of liquid level fluctuation control.

[0037] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0038] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0039] Figure 1 A schematic flowchart of a crystallizer periodic liquid level control method provided in an embodiment of this application is shown;

[0040] Figure 2 A schematic diagram of a frequency characteristic provided in an embodiment of this application is shown:

[0041] Figure 3 A flowchart illustrating another crystallizer cycle liquid level control method provided in the embodiments of this application is shown;

[0042] Figure 4 The present application provides a schematic diagram illustrating the input and inference, and defuzzified output relationship of a feedforward controller according to an embodiment of the present application.

[0043] Figure 5 A schematic flowchart of another crystallizer cycle liquid level control method provided in an embodiment of this application is shown;

[0044] Figure 6 A schematic flowchart of another crystallizer cycle liquid level control method provided in an embodiment of this application is shown. Detailed Implementation

[0045] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0046] This embodiment provides a method for controlling the liquid level in a crystallizer during the cycle, such as... Figure 1 As shown, this method is applied to a crystallizer level control system. The crystallizer level control system includes a crystallizer, a PID control system, and a level adjustment actuator. When the PID control system regulates the level adjustment actuator, the molten steel level in the crystallizer changes. The method includes:

[0047] Step 101: Collect the liquid level fluctuation data of molten steel in the target crystallizer, wherein the liquid level fluctuation of the molten steel changes periodically, and the liquid level fluctuation data includes the liquid level of molten steel at each time node collected based on a preset acquisition frequency within a preset control period.

[0048] In the above embodiments of this application, liquid level fluctuation data can be extracted from the continuous casting billet crystallizer of a steel plant. Specifically, liquid level fluctuation data of molten steel in the target crystallizer is collected. The liquid level fluctuation of molten steel changes periodically. Therefore, the liquid level fluctuation frequency characteristics can be extracted from the collected liquid level fluctuation data so that a liquid level adjustment strategy can be determined based on the liquid level fluctuation frequency characteristics.

[0049] The preset acquisition frequency is, for example, 10Hz, and the preset control period is, for example, 2s. Subsequently, the liquid level of the molten steel can be controlled based on the liquid surface fluctuation data acquired at a frequency of 10Hz within 2s.

[0050] In particular, it can also obtain relevant parameters of the continuous casting machine, stopper rod position data in the crystallizer, argon blowing volume at different positions, and steel composition of the continuously cast billet, which can be used to assist in controlling the molten steel level.

[0051] The results of steel composition collection are shown in Table 1, and the relevant parameters of the continuous casting machine are shown in Table 2.

[0052] Table 1

[0053]

[0054] Table 2

[0055] type parameter Continuous casting machine model Vertical bending type Crystallizer length (mm) 900 Crystallizer vibration frequency (times / min) 25-400 Crystallizer amplitude (mm) 2-10 Billet width (mm) 1065-1522 Billet thickness (mm) 230 Pulling speed (m / min) 0.9-1.3

[0056] Step 102: Perform a fast Fourier transform on the liquid surface fluctuation data to obtain an initial frequency feature vector that reflects the frequency of liquid surface fluctuation changes. Then, filter the initial frequency feature vector to obtain a frequency feature vector.

[0057] Next, a Fast Fourier Transform (FFT) is performed on the periodic liquid surface fluctuation data to analyze its frequency domain characteristics, obtaining an initial frequency feature vector. The acquired liquid surface fluctuation data is then filtered to eliminate noise interference. Specifically, the initial frequency feature vector is filtered to obtain a frequency feature vector, and the final transformation result is shown below. Figure 2 As shown, Figure 2 The left side of the image shows the initial frequency characteristics, and the right side shows the results after filtering. It can be seen that there are very obvious peaks in the periodic liquid surface fluctuations, indicating that the frequency corresponding to the peak is the main frequency of the liquid surface fluctuations.

[0058] Step 103: Convert the frequency feature vector into a time-domain signal, perform feedforward compensation on the time-domain signal through a feedforward controller, and then input it into the PID controller. The PID control system includes a feedforward controller, a PID controller, and a fuzzy parameter adjuster.

[0059] Step 104: When the liquid level control of the molten steel in the target crystallizer is implemented based on the time domain signal after feedforward compensation by the PID controller, the parameters of the PID controller are adjusted in real time based on the fuzzy parameter regulator until the liquid level adjustment execution structure of the PID controller is adjusted to control the liquid level of the molten steel in the crystallizer to meet the control requirements.

[0060] Next, after converting the frequency characteristic vector into a time-domain signal, a feedforward controller is added to the control logic of the PID control system. The output of the feedforward controller is adjusted according to the rate of signal change; that is, by using the feedforward controller to compensate for the time-domain signal before inputting it into the PID controller, the accuracy of the PID controller's input signal can be improved. Then, a fuzzy control method is introduced, using a fuzzy parameter regulator to achieve real-time adjustment of the PID controller parameters. By adding a feedforward controller based on fuzzy rules and adjusting its output according to the rate of signal change, overshoot can be optimized. Ultimately, the liquid level adjustment actuator is controlled based on the PID controller to achieve the required liquid level in the crystallizer. Specifically, the liquid level adjustment actuator can be a hydraulic actuator, an AC brushless servo motor actuator, or a high-precision pneumatic digital cylinder. Hydraulic actuators offer high speed and good linearity, but require oil pumps, oil tanks, and other devices that must be fixed to the ground, which can be inconvenient when connected with flexible hoses. The AC brushless servo motor actuator is a type that has been developed in recent years and has many advantages. However, due to the presence of a speed reducer, its size and weight are not easy to control. The high-precision pneumatic digital cylinder applies the pneumatic digital cylinder technology of machine tools to continuous casting production, which has high control precision, simple structure, and is easy to maintain.

[0061] By applying the technical solution of this embodiment, it can be integrated into a liquid level control device to achieve rapid and precise control of liquid level fluctuations during the control process. By adding a fuzzy parameter regulator to the PID control architecture and formulating fuzzy rules based on on-site big data analysis, control parameters can be changed in real time, achieving different control effects at different control stages. Furthermore, the addition of a fuzzy feedforward parameter regulator adjusts the output of the feedforward controller according to the rate of signal change, further improving the control accuracy of liquid level fluctuations.

[0062] Furthermore, as a refinement and extension of the specific implementation of the above embodiments, to fully illustrate the specific implementation process of this embodiment, another method for periodic liquid level control in a crystallizer is provided, applied to a crystallizer liquid level control system. The crystallizer liquid level control system includes a crystallizer, a PID control system, and a liquid level adjustment execution structure. When the PID control system regulates the liquid level adjustment execution structure, the molten steel level in the crystallizer changes; for example... Figure 3 As shown, the method includes:

[0063] Step 201: Collect the surface fluctuation data of molten steel in the target crystallizer. Perform a Fast Fourier Transform (FFT) on the surface fluctuation data according to the FFT formula to obtain an initial frequency feature vector reflecting the frequency of surface fluctuation changes. The surface fluctuation of the molten steel exhibits periodic changes. The surface fluctuation data includes the molten steel level at each time point within a preset control period, collected based on a preset acquisition frequency. The FFT formula is:

[0064]

[0065] Y(i) represents the initial frequency feature vector obtained after performing a Fast Fourier Transform, where i represents the frequency index, K represents the length of the liquid surface fluctuation data, and n represents the acquisition time index of the liquid surface fluctuation data. even (n) represents the even-numbered sequence in the liquid level fluctuation data, y odd (n) represents the odd sequence in the liquid surface fluctuation data, j represents the imaginary unit, and k represents the ordinal number of the acquisition time node.

[0066] In the above embodiments of this application, the Fast Fourier Transform method is used to perform frequency domain feature analysis on periodic liquid surface fluctuation data, to deeply explore the fluctuation characteristics, and to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes.

[0067] Step 202: Remove frequencies greater than a preset frequency threshold from the initial frequency feature vector to obtain a frequency feature vector, and convert the frequency feature vector into a time-domain signal.

[0068] Next, frequencies exceeding a preset frequency threshold, such as 3 Hz, are removed from the initial frequency feature vector to obtain the final frequency feature vector. This frequency feature vector is then converted into a time-domain signal. Specifically, when filtering the acquired liquid surface fluctuation data to eliminate noise interference, the FIR filter design method can be used. The basic idea of ​​the FIR filter design method is to approximate the desired frequency response curve. Currently, the window function method and the optimal approximation method are commonly used. The window function method can be expressed by the following formula:

[0069]

[0070] Among them, Fd(e jω) It is the required frequency response, f d [n] is derived from the unit impulse response through the inverse Fourier transform, where ω is the frequency and j is a random integer.

[0071] The selection principles for the window function are as follows: the main lobe width should be as narrow as possible to improve frequency domain resolution and reduce data loss, thereby achieving greater stopband attenuation. It has lower sideband amplitudes, especially the first sideband amplitude. The amplitude decays with frequency as quickly as possible. Furthermore, considering the periodic characteristics of liquid surface ripples, a combination of the Blackman window function and FIR can be chosen, defined as shown in the following formula.

[0072] w(n)=0.42-0.5cos(2πn / N)+0.08cos(4πn / N),

[0073] Where N is the window length, and n = 0, 1, ..., N-1.

[0074] Step 203: The time-domain signal is processed by a first-order circuit through a feedforward controller, and the processed time-domain signal is differentiated to obtain the rate of change of the time-domain signal. The output of the feedforward controller is adjusted according to the rate of change and then input to the PID controller. The PID control system includes a feedforward controller, a PID controller and a fuzzy parameter adjuster.

[0075] Next, a feedforward controller is added to the control logic of the PID control system. When the time-domain signal enters the feedforward controller, it first differentiates the time-domain signal after passing through a first-order circuit to obtain the rate of change of the time-domain signal. The output of the feedforward controller is then adjusted based on this rate of change. The specific adjustment value can be calculated using the following formula:

[0076] K = -0.0034e 2 +0.0369e-1.3322,

[0077] Where K is the adjustment value and e is the input time-domain signal.

[0078] The relationship between the input and inference of the feedforward controller and the defuzzified output is as follows: Figure 4 As shown, Figure 4 As shown, when the input signal (the input time-domain signal) changes significantly, the feedforward controller will output a small adjustment amount to help the system quickly reach the desired position. When the input signal (the input time-domain signal) tends to stabilize, the feedforward controller will output a larger adjustment amount to offset the overshoot in the feedback controller, thereby reducing the settling time of the PID control system and enabling the PID control system to reach stability more quickly.

[0079] Step 204: The output of the PID controller is sent as a control signal to the liquid level adjustment actuator, so that the liquid level adjustment actuator adjusts the inflow of molten steel in the crystallizer according to the control signal.

[0080] Step 205: The molten steel level is detected in real time by a sensor, and the detected molten steel level value is returned to the PID control system as a feedback signal so that the PID control system can continue to adjust the control signal according to the difference between the feedback signal and the molten steel level set value.

[0081] Step 206: Adjust the parameters of the PID controller in real time based on the fuzzy parameter regulator until the liquid level adjustment execution structure based on the PID controller controls the liquid level in the crystallizer to meet the control requirements.

[0082] Next, the output of the PID controller is sent as a control signal to the level adjustment actuator, which then adjusts the inflow of molten steel into the crystallizer according to the control signal. The molten steel level is detected in real time by sensors, and the detected level value is returned as a feedback signal to the PID control system. The PID control system then adjusts the control signal based on the difference between the feedback signal and the setpoint for the molten steel level. The PID controller parameters are adjusted in real time using a fuzzy parameter regulator until the PID controller controls the level adjustment actuator, achieving the required molten steel level in the crystallizer.

[0083] By applying the technical solution of this embodiment, through on-site sampling, the liquid level fluctuation data of molten steel in the continuous casting crystallizer of a steel plant is analyzed in the frequency domain using the Fast Fourier Transform method to deeply explore the fluctuation characteristics. The collected liquid level fluctuation data is also filtered to eliminate interference from noise. A feedforward controller is added to the control logic of the PID control system. After the input signal enters the feedforward controller, it first differentiates the input signal after passing through a first-order loop to obtain the rate of change of the input signal. The output of the feedforward controller is then adjusted according to the rate of change, thereby improving the accuracy of liquid level control.

[0084] Furthermore, as a refinement and extension of the specific implementation of the above embodiments, to fully illustrate the specific implementation process of this embodiment, another method for periodic liquid level control in a crystallizer is provided, applied to a crystallizer liquid level control system. The crystallizer liquid level control system includes a crystallizer, a PID control system, and a liquid level adjustment execution structure. When the PID control system regulates the liquid level adjustment execution structure, the molten steel level in the crystallizer changes; for example... Figure 5 As shown, the method includes:

[0085] Step 301: Collect the liquid level fluctuation data of molten steel in the target crystallizer, perform a fast Fourier transform on the liquid level fluctuation data to obtain an initial frequency feature vector reflecting the frequency of liquid level fluctuation changes, and filter the initial frequency feature vector to obtain a frequency feature vector. The liquid level fluctuation of the molten steel changes periodically, and the liquid level fluctuation data includes the liquid level of molten steel at each time node collected based on a preset acquisition frequency within a preset control period.

[0086] In the above embodiments of this application, the liquid surface fluctuation data of molten steel in the target crystallizer is collected, the liquid surface fluctuation data is subjected to fast Fourier transform to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes, and the initial frequency feature vector is filtered to obtain a frequency feature vector, which is used to prepare for subsequent control of the molten steel surface.

[0087] Step 302: Convert the frequency feature vector into a time-domain signal, perform feedforward compensation on the time-domain signal through a feedforward controller, and then input it into a PID controller. The PID control system includes a feedforward controller, a PID controller, and a fuzzy parameter regulator.

[0088] Next, the frequency feature vector is converted into a time-domain signal. After the time-domain signal is fed forward by the feedforward controller, it is input into the PID controller, which then "controls" the molten steel level in the crystallizer based on the time-domain signal.

[0089] Step 303: When performing liquid level control on the molten steel in the target crystallizer based on the PID controller and the time-domain signal after feedforward compensation, configure the initial parameters of the PID controller and set the proportional coefficient output domain, integral coefficient output domain, and differential coefficient output domain of the fuzzy parameter adjuster. The initial parameters include the proportional coefficient kp, integral coefficient ki, and differential coefficient kd. The proportional coefficient output domain is [-0.7, 0.7], the integral coefficient output domain is [-1.1, 1.1], and the differential coefficient output domain is [-0.7, 0.7].

[0090] [-0.21, 0.21].

[0091] Step 304: Determine the fuzziness level of the input time-domain signal based on its amplitude and direction. The fuzziness level includes negative giant, negative medium, negative small, zero, positive small, positive medium, and positive giant.

[0092] Step 305: Calculate the fuzzy output of the PID controller parameters based on the preset fuzzy rules and the fuzzy level, wherein the preset fuzzy rules include the PID controller parameter adjustment methods corresponding to different fuzzy levels.

[0093] Step 306: After defuzzifying the fuzzy output, the obtained proportional coefficient error value Δkp, integral coefficient error value Δki, and differential coefficient error value Δkd are added to the initial parameters of the PID controller, so that the PID controller adjusts the liquid level adjustment execution structure based on the latest parameters until the liquid level adjustment execution structure is adjusted based on the PID controller to control the liquid steel level in the crystallizer to meet the control requirements.

[0094] By adding fuzzy rules to the traditional PID control system to dynamically adjust the PID parameters, control accuracy can be improved. This is achieved by introducing fuzzy control methods and using a fuzzy parameter regulator to adjust the PID controller parameters in real time. The output domains of the fuzzy parameter regulator are Δkp = [-0.7, 0.7], Δki = [-1.1, 1.1], and Δkd = [-0.21, 0.21]. The fuzzy subsets of the input and output are defined as: negative giant (NH), negative medium (NM), negative small (NS), zero (O), positive small (PS), positive medium (PM), and positive giant (PH). Fuzzy inference and defuzzification are then performed to obtain the parameters. After obtaining the fuzzy error information, the fuzzy regulator mimics human parameter tuning experience, performs fuzzy inference on the controller parameters based on the fuzzy information, and defuzzifies the inference results to finally obtain the actual adjustment amount.

[0095] Specifically, configuring the initial parameters of the PID controller involves setting a reasonable set of initial parameters for the PID controller, namely the proportional coefficient kp, integral coefficient ki, and derivative coefficient kd. The selection of these initial parameters is usually based on experience, system characteristics, and control objectives.

[0096] Regarding the proportional gain kp: it determines the speed of the system response and the overshoot. The initial value can be selected based on the dynamic characteristics of the system, generally choosing a moderate value to ensure that the system is neither too sluggish nor too oscillating.

[0097] Regarding the integral coefficient ki: it is used to eliminate static errors. The selection of the initial value should consider the static characteristics of the system and the impact of the integral action on the system stability.

[0098] Regarding the differential coefficient kd: it improves the system's stability and response speed, and is used to reduce overshoot. The selection of the initial value requires a trade-off between the system's dynamic performance and stability.

[0099] Next, the output domains of the fuzzy parameter regulator are set, namely Δkp = [-0.7, 0.7], Δki = [-1.1, 1.1] and Δkd = [-0.21, 0.21]. These output domains limit the range of adjustment of the PID controller parameters by the fuzzy parameter regulator, ensuring that the adjusted parameters are still within a reasonable range.

[0100] Next, the ambiguity level of the target time-domain signal is determined. Based on the amplitude and direction of the target time-domain signal, it is classified into a preset ambiguity level, including negative giant, negative medium, negative small, zero, positive small, positive medium, and positive giant. This is usually achieved by comparing the target signal with a preset threshold.

[0101] Next, preset fuzzy rules are formulated, defining how the PID controller parameters are adjusted under different fuzzy levels. These rules can be formulated based on experience, system characteristics, and control objectives. For example, when the target signal is at a negatively large level, it may be necessary to significantly increase the proportional coefficient kp to quickly respond to signal changes, while appropriately decreasing the integral coefficient ki to prevent integral saturation; when the target signal is at a zero level, it may not be necessary to adjust the PID controller parameters; when the target signal is at a positively small level, it may be necessary to appropriately decrease the proportional coefficient kp to reduce overshoot and appropriately increase the derivative coefficient kd to improve system stability.

[0102] Next, the fuzzy output of the PID controller parameters is calculated. Based on the fuzziness level of the target time domain signal and the preset fuzziness rules, the fuzzy output of the PID controller parameters, namely Δkp, Δki and Δkd, is calculated using the fuzzy parameter adjuster.

[0103] Next, deblurring is performed on the fuzzy output to obtain the specific values ​​of Δkp, Δki, and Δkd. Deblurring methods can include centroid method, maximum membership method, etc.

[0104] Finally, the parameters of the PID controller are updated by adding the obtained Δkp, Δki, and Δkd values ​​to the initial parameters of the PID controller, resulting in new PID controller parameters. These new parameters are then used to control the liquid level of the molten steel in the target crystallizer.

[0105] Next, monitoring and adjustments can be made. During actual operation, the system's response and control effectiveness need to be continuously monitored. If the control effect is unsatisfactory, adjustments and optimizations can be made to the initial parameters, fuzzy level, fuzzy rules, etc., based on the actual situation.

[0106] By following the steps above, the parameters of the PID controller can be adjusted and optimized in real time, thereby improving the accuracy and stability of the molten steel level control in the target crystallizer.

[0107] By applying the technical solution of this embodiment, by adding a fuzzy parameter regulator to the PID control architecture and formulating fuzzy rules based on on-site big data analysis, the control parameters can be changed in real time, achieving different control effects at different control stages. At the same time, by adding a fuzzy feedforward parameter regulator, the output of the feedforward controller is adjusted according to the rate of change of the signal, thereby improving the control accuracy of liquid surface fluctuations.

[0108] Furthermore, as a refinement and extension of the specific implementation of the above embodiments, and to fully illustrate the specific implementation process of this embodiment, another method for controlling the liquid level in a crystallizer during the cycle is provided, such as... Figure 6 As shown, the method includes:

[0109] Step 401: Collect the liquid level fluctuation data of molten steel in the target crystallizer, wherein the liquid level fluctuation of the molten steel changes periodically, and the liquid level fluctuation data includes the liquid level of molten steel at each time node collected based on a preset acquisition frequency within a preset control period.

[0110] Step 402: Perform a fast Fourier transform on the liquid surface fluctuation data to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes. Filter the initial frequency feature vector to obtain a frequency feature vector.

[0111] In this embodiment, data on the surface fluctuation of molten steel in the target crystallizer is collected. The surface fluctuation of the molten steel exhibits periodic changes, and the data includes the molten steel level at each time point within a preset control period, collected at a preset acquisition frequency. A Fast Fourier Transform is performed on the surface fluctuation data to obtain an initial frequency feature vector reflecting the frequency of surface fluctuation changes. This initial frequency feature vector is then filtered to obtain a further frequency feature vector. This process prepares for subsequent determination of the surface fluctuation state based on the amplitude of the molten steel surface fluctuations, and for controlling the molten steel level in the crystallizer to meet control requirements based on the surface fluctuation state.

[0112] Step 403: Based on the frequency feature vector, determine the dominant frequency affecting the fluctuation of the molten steel surface, and determine the fluctuation amplitude of the molten steel surface based on the determined dominant frequency, wherein the fluctuation state includes stable fluctuation and unstable fluctuation.

[0113] Step 404: When the fluctuation amplitude of the molten steel surface is outside the preset normal fluctuation range, the molten steel surface fluctuation is in an abnormal state; when the fluctuation amplitude of the molten steel surface is within the preset normal fluctuation range, the molten steel surface fluctuation is in a normal state.

[0114] Step 405: If the fluctuation of the molten steel level is abnormal, adjust the liquid level adjustment mechanism until the fluctuation of the molten steel level in the crystallizer is within the preset normal fluctuation range.

[0115] Next, a normal fluctuation range is preset, for example, ±5mm. When the fluctuation amplitude of the molten steel surface is outside the preset normal fluctuation range, the molten steel surface fluctuation is in an abnormal state. When the fluctuation amplitude of the molten steel surface is within the preset normal fluctuation range, the molten steel surface fluctuation is in a normal state. When an abnormal state occurs, the liquid level needs to be adjusted to adjust the execution structure until the fluctuation amplitude of the molten steel surface in the crystallizer is within the preset normal fluctuation range.

[0116] By applying the technical solution of this embodiment, the dominant frequency affecting the fluctuation of the molten steel surface is determined based on the frequency feature vector, and the fluctuation characteristics of the molten steel are judged accordingly. This enables precise control of the crystallizer operating parameters. That is, when the molten steel fluctuates violently and the fluctuation of the molten steel surface is abnormal, the surface fluctuation can be effectively stabilized and the quality of the molten steel can be improved.

[0117] It should be noted that other corresponding descriptions of the functional units involved in the crystallizer cycle liquid level control device provided in this application embodiment can be found in the following references. Figure 1 , Figure 3 , Figure 5 and Figure 6 The corresponding descriptions in the method will not be repeated here.

[0118] Based on the above, Figure 1 , Figure 3 , Figure 5 and Figure 6 Accordingly, this application also provides a storage medium storing a computer program, which, when executed by a processor, implements the above-described method. Figure 1 , Figure 3 , Figure 5 and Figure 6 The crystallizer cycle liquid level control method is shown.

[0119] Based on this understanding, the technical solution of this application can be embodied in the form of a software product. This software product can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or portable hard drive), and includes several instructions to cause a computer device (such as a personal computer, server, or network device) to execute the methods described in the various implementation scenarios of this application.

[0120] Based on the above, Figure 1 , Figure 3 , Figure 5 and Figure 6 To achieve the above objectives, this application also provides a computer device, specifically a personal computer, server, network device, etc., as shown in the method. The computer device includes a storage medium and a processor; the storage medium stores a computer program; the processor executes the computer program to achieve the above-described objectives. Figure 1 , Figure 3 , Figure 5 and Figure 6 The crystallizer cycle liquid level control method is shown.

[0121] Optionally, the computer device may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB interfaces, card reader interfaces, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Bluetooth interfaces, Wi-Fi interfaces), etc.

[0122] Those skilled in the art will understand that the computer device structure provided in this embodiment does not constitute a limitation on the computer device, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0123] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages and stores the hardware and software resources of a computer device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various components within the storage medium, as well as communication with other hardware and software within the physical device.

[0124] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms, or by hardware implementation to convert the frequency feature vector of liquid level fluctuation data into a time-domain signal, and then inputting the time-domain signal into a PID controller after feedforward compensation by a feedforward controller. When liquid level control is implemented on the molten steel in the target crystallizer based on the PID controller and the feedforward compensated time-domain signal, the parameters of the PID controller are adjusted in real time based on the fuzzy parameter regulator until the liquid level adjustment execution structure is adjusted based on the PID controller to control the liquid level in the crystallizer to meet the control requirements. By adding a fuzzy parameter regulator to the PID control architecture and formulating fuzzy rules based on on-site big data analysis, the control parameters can be changed in real time, achieving different control effects at different control stages. At the same time, the addition of a fuzzy feedforward parameter regulator adjusts the output of the feedforward controller according to the rate of signal change, improving the control accuracy of liquid level fluctuations.

[0125] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application. Those skilled in the art will understand that the modules in the apparatus of the embodiment can be distributed within the apparatus of the embodiment as described, or can be modified to be located in one or more apparatuses different from this embodiment. The modules of the above-described embodiment can be combined into one module, or further divided into multiple sub-modules.

[0126] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.

Claims

1. A method for periodic liquid level control in a crystallizer, characterized in that, An application is made to a crystallizer level control system, the crystallizer level control system comprising a crystallizer, a PID control system, and a level adjustment actuator, wherein when the PID control system regulates the level adjustment actuator, the molten steel level in the crystallizer changes; the method includes: Collect data on the fluctuation of the liquid level of molten steel in the target crystallizer, wherein the fluctuation of the liquid level of molten steel is periodic, and the liquid level fluctuation data includes the liquid level of molten steel at each time point within a preset control period based on a preset acquisition frequency; The liquid surface fluctuation data is subjected to a fast Fourier transform to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes. The initial frequency feature vector is then filtered to obtain a frequency feature vector. The frequency feature vector is converted into a time-domain signal, and after the time-domain signal is fed forward by a feedforward controller, it is input into a PID controller. The PID control system includes a feedforward controller, a PID controller, and a fuzzy parameter regulator. When the liquid level of molten steel in the target crystallizer is controlled based on the PID controller and the time-domain signal after feedforward compensation, the parameters of the PID controller are adjusted in real time based on the fuzzy parameter regulator until the liquid level adjustment execution structure based on the PID controller controls the liquid level of molten steel in the crystallizer to meet the control requirements.

2. The method according to claim 1, characterized in that, The feedforward compensation of the time-domain signal via a feedforward controller includes: The time-domain signal is processed by a first-order circuit using a feedforward controller, and the processed time-domain signal is differentiated to obtain the rate of change of the time-domain signal. The output of the feedforward controller is adjusted according to the rate of change.

3. The method according to claim 1, characterized in that, The real-time adjustment of PID controller parameters based on a fuzzy parameter regulator includes: Configure the initial parameters of the PID controller and set the proportional coefficient output domain, integral coefficient output domain, and derivative coefficient output domain of the fuzzy parameter regulator. The initial parameters include the proportional coefficient kp, integral coefficient ki, and derivative coefficient kd. The proportional coefficient output domain is [-0.7, 0.7], the integral coefficient output domain is [-1.1, 1.1], and the derivative coefficient output domain is [-0.21, 0.21]. Based on the amplitude and direction of the input time-domain signal, the fuzziness level of the input time-domain signal is determined, wherein the fuzziness level includes negative giant, negative medium, negative small, zero, positive small, positive medium, and positive giant. Based on preset fuzzy rules and the fuzzy level, the fuzzy output of the PID controller parameters is calculated, wherein the preset fuzzy rules include the PID controller parameter adjustment methods corresponding to different fuzzy levels; After defuzzifying the fuzzy output, the obtained proportional coefficient error value Δkp, integral coefficient error value Δki, and derivative coefficient error value Δkd are added to the initial parameters of the PID controller so that the PID controller can adjust the liquid level adjustment execution structure based on the latest parameters.

4. The method according to claim 1, characterized in that, The method of controlling the liquid level of molten steel in the target crystallizer based on a PID controller and a time-domain signal after feedforward compensation includes: The output of the PID controller is sent as a control signal to the liquid level adjustment actuator, so that the liquid level adjustment actuator adjusts the inflow of molten steel in the crystallizer according to the control signal. The molten steel level is detected in real time by a sensor, and the detected molten steel level value is returned to the PID control system as a feedback signal. The PID control system then continues to adjust the control signal based on the difference between the feedback signal and the molten steel level setpoint.

5. The method according to claim 1, characterized in that, The step of performing a Fast Fourier Transform on the liquid surface fluctuation data to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes includes: According to the Fast Fourier Transform (FFT) formula, the liquid surface fluctuation data is subjected to a Fast Fourier Transform to obtain an initial frequency feature vector reflecting the frequency of liquid surface fluctuation changes. The Fast Fourier Transform formula is as follows: Y(i) represents the initial frequency feature vector obtained after performing a Fast Fourier Transform, where i represents the frequency index, K represents the length of the liquid surface fluctuation data, and n represents the acquisition time index of the liquid surface fluctuation data. even (n) represents the even-numbered sequence in the liquid level fluctuation data, y odd (n) represents the odd sequence in the liquid surface fluctuation data, j represents the imaginary unit, and k represents the ordinal number of the acquisition time node.

6. The method according to claim 1, characterized in that, The step of filtering the initial frequency feature vector to obtain the frequency feature vector includes: In the initial frequency feature vector, frequencies greater than a preset frequency threshold are removed to obtain the frequency feature vector.

7. The method according to claim 1, characterized in that, After filtering the initial frequency feature vector to obtain the frequency feature vector, the method further includes: Based on the frequency feature vector, the dominant frequency affecting the fluctuation of the molten steel surface is determined, and the amplitude of the molten steel surface fluctuation is determined based on the determined dominant frequency. The molten steel level fluctuation is determined based on the amplitude of the molten steel level fluctuation, and the molten steel level in the crystallizer is controlled based on the molten steel level fluctuation to meet the control requirements.

8. The method according to claim 7, characterized in that, The liquid level fluctuation state includes abnormal and normal states. Determining the liquid level fluctuation state based on the amplitude of the molten steel level fluctuation, and controlling the molten steel level in the crystallizer based on the liquid level fluctuation state to achieve the control requirements, includes: When the fluctuation range of the molten steel surface is outside the preset normal fluctuation range, the molten steel surface fluctuation is considered abnormal; when the fluctuation range of the molten steel surface is within the preset normal fluctuation range, the molten steel surface fluctuation is considered normal. If the fluctuation of the molten steel level is abnormal, the liquid level adjustment mechanism will be activated until the fluctuation of the molten steel level in the crystallizer is within the preset normal fluctuation range.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for periodic liquid level control of the crystallizer as described in any one of claims 1 to 8.

10. A computer device, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for periodic liquid level control of the crystallizer as described in any one of claims 1 to 8.

Images