A signal processing system and method for measuring molten steel level in a crystallizer

By optimizing the phase matching between the reference signal and the induced signal, a reference square wave signal is generated using a frequency accumulator and an FPGA module. Combined with phase-sensitive detection and low-pass filtering, high precision and stability of molten steel level measurement in the crystallizer are achieved, solving the signal distortion problem caused by phase mismatch and simplifying the hardware design.

CN123111408APending Publication Date: 2026-11-17BAOSHAN IRON & STEEL CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202510624929.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-05-15
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In the measurement of molten steel level in the crystallizer, the existing technology suffers from signal distortion caused by phase mismatch, which affects the stability and accuracy of signal detection, and also has high hardware cost and design complexity.

Method used

By optimizing the phase matching between the reference signal and the induced signal, a stable reference square wave signal is generated using a frequency accumulator and an FPGA module. An adjustable phase output is achieved through a phase delay chain. Combined with phase-sensitive detection, low-pass filtering, and ADC sampling, the phase is dynamically adjusted to achieve precise signal matching.

Benefits of technology

It significantly reduces signal distortion caused by phase mismatch, improves the stability and accuracy of signal detection, simplifies hardware design, and reduces implementation costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123111408A_ABST
    Figure CN123111408A_ABST
Patent Text Reader

Abstract

This invention discloses a signal processing system for measuring molten steel level in a crystallizer, comprising: a sensor for acquiring the molten steel level signal in the crystallizer; a phase-sensitive detection module for acquiring the level signal and extracting the modulation waveform; a low-pass filter for extracting the fundamental frequency component related to the level information in the modulation waveform; an ADC sampling module for converting the signal after the low-pass filter into a digital signal; an evaluation module for analyzing the signal from the ADC sampling module to determine the actual molten steel level; and a signal conditioning module for acquiring the optimized reference signal from the evaluation module and returning the reference signal to the phase-sensitive detection module to re-match it with the input signal of the sensor, achieving phase matching between the reference signal and the induced signal. This invention also discloses a method for processing molten steel level measurement signals in a crystallizer. By optimizing the phase matching between the reference signal and the induced signal, this invention reduces signal distortion caused by phase mismatch and enhances the stability of signal detection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to continuous casting technology, and more specifically, to a system and method for processing molten steel level measurement signals in a crystallizer. Background Technology

[0002] In continuous casting, real-time monitoring of the molten steel level in the crystallizer is crucial for ensuring stable production and improving product quality. The molten steel level in the crystallizer is typically measured using an electromagnetic method. A sinusoidal signal of a certain frequency is used as the excitation signal, and the induced signal from the level receiving sensor also appears as a sinusoidal wave, with its amplitude directly proportional to the actual molten steel level in the crystallizer. Therefore, accurately extracting the molten steel level information from the induced signal and converting it into an accurate DC signal is key to the molten steel level measurement system in the crystallizer.

[0003] Lock-in amplifiers (LIAs) are ideal tools for liquid level signal detection due to their ability to extract signals of specific frequencies and effectively suppress background noise. In practical applications, ensuring precise phase alignment between the liquid level signal and the reference signal is crucial to improving signal detection accuracy.

[0004] In the article "Design of a Dual-Phase Lock-in Amplifier Based on AD630", two AD630 chips were used in the phase-sensitive detection stage, with their reference signals 90° out of phase. The final voltage was obtained by calculating the sum of the squares of the signals output by the two AD630 chips after low-pass filtering. Although this method effectively improved the detection signal-to-noise ratio, the use of two AD630 chips and a low-pass filter increased hardware costs and design complexity.

[0005] Chinese patent application number 202410707181.5 discloses "a weak magnetic field detection device and method, and a non-contact current measurement method." This patented technology uses an excitation signal as a reference signal for a lock-in amplifier during the phase-sensitive detection stage to ensure that the target signal and the reference signal maintain phase consistency during demodulation. However, in the application scenario of crystallizer level measurement, due to the complexity of environmental conditions, the phase of the level signal can fluctuate significantly due to environmental changes. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide a signal processing system and method for measuring molten steel level in a crystallizer. By optimizing the phase matching between the reference signal and the inductive signal, the signal distortion problem caused by phase mismatch is significantly reduced, thereby enhancing the stability and accuracy of signal detection.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] The first aspect of the present invention provides a steel liquid level measurement signal processing system in a crystallizer, comprising:

[0009] Sensors acquire the liquid level signal of molten steel in the crystallizer;

[0010] The phase-sensitive detection module acquires the liquid level signal and extracts the modulation waveform;

[0011] A low-pass filter is used to extract the fundamental frequency component related to the liquid level information from the modulated waveform;

[0012] The ADC sampling module converts the signal after passing through the low-pass filter into a digital signal;

[0013] The evaluation module analyzes the signal from the ADC sampling module to determine the actual situation of the molten steel level.

[0014] The signal conditioning module acquires the optimized reference signal from the evaluation module and returns the reference signal to the phase-sensitive detection module to re-match it with the input signal of the sensor, thereby achieving phase matching between the reference signal and the sensed signal.

[0015] Preferably, the signal conditioning module includes:

[0016] The reference signal is generated by using a frequency accumulator and a frequency control word to achieve precise control of the signal frequency and generate a stable reference square wave signal.

[0017] The reference signal phase is shifted, and an adjustable phase reference signal output is provided through the FPGA module.

[0018] Preferably, the adjustable phase of the FPGA module is achieved through a phase delay chain.

[0019] The second aspect of the present invention provides a method for processing molten steel level measurement signals in a crystallizer based on the crystallizer molten steel level measurement signal processing system provided in the first aspect of the present invention;

[0020] The sensor acquires the liquid level signal of the molten steel in the crystallizer, and multiplies it with the generated reference signal in the phase-sensitive detection module to extract the modulation waveform containing the liquid level information.

[0021] The modulated waveform signal enters the low-pass filter to filter out high-frequency noise and extract the fundamental frequency component related to the liquid level information. The filtered signal is converted into a digital signal by the ADC sampling module and transmitted to the evaluation module.

[0022] The evaluation module analyzes the sampled signal to determine the actual situation of the molten steel level, and generates an optimized reference signal by sending a control word to the signal conditioning module.

[0023] The optimized reference signal is returned to the phase-sensitive detection module and re-matched with the input signal of the sensor to achieve phase matching between the reference signal and the sensed signal.

[0024] Preferably, in the phase-sensitive detector module, the input signal is the liquid level signal of the crystallizer, and its expression is:

[0025] V in (t)=A s sin(ω s t+φ s (1)

[0026] In equation (1), V in (t) is the input signal, A s ω is the amplitude of the input signal. s φ is the angular frequency of the input signal. s The phase of the input signal;

[0027] The reference signal for phase-sensitive detection is:

[0028]

[0029] In equation (2), V ref (t) is the input signal, A r ω is the amplitude of the reference signal. r φ is the angular frequency of the reference signal. r The phase of the reference signal is given by k, and the harmonic order is given by k.

[0030] The input signal is mixed with the reference signal to obtain the product signal:

[0031] V mix (t)=V in (t)·V ref (t) (3)

[0032] In equation (3), V mix (t) represents the output signal of the phase-sensitive detector. Based on the product-sum formula for trigonometric functions, the mixed signal is expanded as follows:

[0033]

[0034] Preferably, the output signal after filtering by the low-pass filter is expressed as:

[0035]

[0036] In equation (5), V LP The output signal of the low-pass filter, φ s -φ r This represents the phase difference between the input signal and the reference signal.

[0037] Preferably, the signal conditioning module includes:

[0038] The reference signal is generated by using a frequency accumulator and a frequency control word to achieve precise control of the signal frequency and generate a stable reference square wave signal.

[0039] The reference signal phase is shifted, and an adjustable phase reference signal output is provided through the FPGA module.

[0040] Preferably, the frequency control word determines the increment of the accumulator in each clock cycle, and its value is jointly determined by the target output frequency and the system clock frequency, calculated as follows:

[0041]

[0042] In equation (6), N is the bit width of the accumulator, f out f is the target output signal frequency. clk This is the system clock frequency.

[0043] Preferably, the FPGA module performs phase adjustment through a phase delay chain.

[0044] Preferably, the phase delay chain is based on LCELL chains connected in series to form a signal transmission path, specifically including:

[0045] Driven by the rising edge of the clock signal, the input signal propagates through the delay chain step by step; each stage register introduces a fixed time delay, and different delay points are selected as outputs through the phase control word, thereby realizing dynamic phase adjustment of the signal. The total delay time can be expressed as:

[0046] T delay =T clk * reg phase (7)

[0047] In equation (7), T delay T represents the total delay time. clk For the period of the clock signal, reg phase This is the phase control word.

[0048] Preferably, the evaluation module includes a search process, as follows:

[0049] Initialization phase: Determine the required initial conditions and parameters, including the search range, the frequency and phase of the initial reference signal.

[0050] Coarse search: Make rapid adjustments over a large range with large step sizes, and determine the approximate optimal area by observing the trend of the results;

[0051] Fine-tuning search: Within the initially determined area, the search range is narrowed down and the adjustment step size is reduced to get closer to the optimal state;

[0052] Precise search: By adjusting smaller steps, the range is narrowed down, and the results are precisely optimized;

[0053] Finally, confirm and record the optimal state, complete the adjustment process, and end the automatic phase alignment.

[0054] The phase-aligned output signal can be expressed by the following formula:

[0055]

[0056] This invention provides a signal processing system and method for measuring molten steel level in a crystallizer. Based on a phase-sensitive detection-based method for enhancing the level signal, it significantly reduces signal distortion caused by phase mismatch by optimizing the phase matching between the reference signal and the induced signal, thereby enhancing the stability and accuracy of signal detection. This invention offers a new technical solution for level detection in continuous casting crystallizers, improving the measurement accuracy of the level signal, simplifying hardware design, and reducing implementation costs. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the structural framework of the molten steel level measurement signal processing system in the crystallizer of the present invention;

[0058] Figure 2 This is a flowchart illustrating the steel liquid level measurement signal processing method in the crystallizer of the present invention.

[0059] Figure 3 This is a schematic diagram of the phase delay chain used in the signal conditioning module of the steel liquid level measurement signal processing method in the crystallizer of the present invention;

[0060] Figure 4 This is a schematic diagram of the electromagnetic wave pattern of the phase-sensitive detector module in an embodiment of the method for processing the molten steel level measurement signal in the crystallizer of the present invention;

[0061] Figure 5 This is a schematic diagram of the automatic phase alignment of the evaluation module in an embodiment of the method for processing the molten steel level measurement signal in the crystallizer of the present invention. Detailed Implementation

[0062] To better understand the above-mentioned technical solutions of the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0063] Combination Figure 1 As shown, the present invention provides a steel liquid level measurement signal processing system in a crystallizer, comprising:

[0064] Sensors acquire the liquid level signal of molten steel in the crystallizer;

[0065] Phase-sensitive detection module 1 acquires the liquid level signal and multiplies it with the generated reference signal to realize the phase-sensitive detection function and extract the modulation waveform containing liquid level information;

[0066] Low-pass filter 2 filters out high-frequency noise and extracts the fundamental frequency component related to the liquid level information in the modulated waveform;

[0067] The ADC sampling module 3 converts the signal after passing through the low-pass filter 2 into a digital signal and transmits it to the evaluation module 4;

[0068] Evaluation module 4 analyzes the signal from ADC sampling module 3 to determine the actual steel level and generates an optimized reference signal by sending a control word to signal conditioning module 5.

[0069] The signal conditioning module 5 acquires the reference signal optimized by the evaluation module 4 and returns the reference signal to the phase-sensitive detection module 1 to re-match it with the input signal of the sensor, thereby achieving phase matching between the reference signal and the sensed signal.

[0070] Signal conditioning module 5 includes:

[0071] The reference signal is generated by using a frequency accumulator and a frequency control word to achieve precise control of the signal frequency and generate a stable reference square wave signal.

[0072] The reference signal phase is shifted, and an adjustable phase reference signal output is provided through the FPGA module.

[0073] The adjustable phase of the aforementioned FPGA module is achieved through a phase delay chain.

[0074] Combined Figure 1 and Figure 2 As shown, the present invention also provides a method for measuring the molten steel level in a crystallizer based on the crystallizer molten steel level measurement signal processing system of the present invention. By optimizing the phase matching of the reference signal and the induction signal, the signal distortion problem caused by phase mismatch is significantly reduced, thereby enhancing the stability and accuracy of signal detection.

[0075] The liquid level signal of molten steel in the crystallizer is acquired by a sensor and multiplied with the generated reference signal in the phase-sensitive detector module 1 to extract the modulation waveform containing liquid level information.

[0076] The modulated waveform signal enters the low-pass filter 2 to filter out high-frequency noise and extract the fundamental frequency component related to the liquid level information. The filtered signal is converted into a digital signal by the ADC sampling module 3 and transmitted to the evaluation module 4.

[0077] The evaluation module 4 analyzes the sampled signal to determine the actual situation of the molten steel level, and generates an optimized reference signal by sending a control word to the signal conditioning module 5.

[0078] The optimized reference signal is returned to the phase-sensitive detector module 1 and re-matched with the sensor's input signal to achieve phase matching between the reference signal and the sensed signal.

[0079] The phase-sensitive detector module 1 in the steel level measurement signal processing method within the crystallizer of this invention is used to lock and amplify signals of a specific frequency while suppressing noise and interference at other frequencies. In steel level detection within the crystallizer, the phase-sensitive detector module 1 multiplies the reference signal with the level signal, shifting the main energy spectrum of the target signal to the vicinity of the DC component, while interfering signals are shifted to other frequency bands. After subsequent processing by the low-pass filter 2, the interfering signals are suppressed, and only signals directly related to the steel level change are output, achieving high-precision detection. The specific process is as follows:

[0080] The input signal of phase-sensitive detector module 1 is the liquid level signal of the crystallizer, and its expression is:

[0081] V in (t)=A s sin(ω s t+φ s (1)

[0082] In equation (1), V in (t) is the input signal, A s ω is the amplitude of the input signal. s φ is the angular frequency of the input signal. s The phase of the input signal;

[0083] The reference signal for phase-sensitive detection is:

[0084]

[0085] In equation (2), V ref (t) is the input signal, A r ω is the amplitude of the reference signal. r φ is the angular frequency of the reference signal. r The phase of the reference signal is given by k, and the harmonic order is given by k.

[0086] The input signal is mixed with the reference signal to obtain the product signal:

[0087] V mix (t)=V in (t)·V ref (t) (3)

[0088] In equation (3), V mix(t) represents the output signal of the phase-sensitive detector. Based on the product-sum formula for trigonometric functions, the mixed signal is expanded as follows:

[0089]

[0090] The low-pass filter 2 in the steel level measurement signal processing method in the crystallizer of this invention mainly functions to suppress high-frequency noise after spectral shift, while extracting low-frequency signals containing effective liquid level information. Its core objective is to ensure the stability and purity of the detection signal and effectively reduce the impact of high-frequency interference.

[0091] After processing by low-pass filter 2, only the low-frequency component corresponding to k=1 in the mixer signal is retained, while the remaining high-frequency components are sufficiently suppressed. The filtered output signal is expressed as:

[0092]

[0093] In equation (5), V LP The output signal of the low-pass filter, φ s -φ r This represents the phase difference between the input signal and the reference signal. The amplitude of the filtered signal is directly related to the phase of the liquid level, laying the foundation for subsequent signal processing and molten steel level measurement.

[0094] In the ADC sampling module 3 of the steel liquid level measurement signal processing method in the crystallizer of the present invention, the output voltage of the low-pass filter 2 is converted from a continuous analog quantity to a digital quantity and input into the process evaluation module 4 for further calculation and analysis.

[0095] The signal conditioning module 5 in the steel liquid level measurement signal processing method in the crystallizer of the present invention includes:

[0096] The reference signal is generated by using a frequency accumulator and a frequency control word to achieve precise control of the signal frequency and generate a stable reference square wave signal.

[0097] The reference signal phase is shifted, and an adjustable phase reference signal output is provided through the FPGA module.

[0098] The frequency control word determines the increment of the accumulator per clock cycle. Its value is determined by both the target output frequency and the system clock frequency, and the calculation formula is as follows:

[0099]

[0100] In equation (6), N is the bit width of the accumulator, f out f is the target output signal frequency. clk This refers to the system clock frequency. In the system design, dynamic control of the output frequency is achieved by configuring FCW.

[0101] The task of the frequency accumulator is to add the current accumulated value to the frequency control word in each clock cycle. When the accumulated value exceeds the maximum bit width of the accumulator, an overflow occurs and it restarts from zero, forming a periodic output signal. The most significant bit of the accumulator is directly used to generate the state of the square wave signal, ensuring that the output digital signal has extremely high frequency stability and consistency.

[0102] In liquid level control applications, the initial phase of the liquid level signal may dynamically shift due to sensor replacement or changes in operating conditions. To ensure the consistency of the reference signal with the liquid level signal in frequency and phase, the FPGA module needs to provide an adjustable-phase reference signal output.

[0103] Combination Figure 3 As shown, the FPGA module adjusts the phase through a phase delay chain.

[0104] The phase delay chain is based on LCELL chains connected in series to form a signal transmission path, specifically including:

[0105] Driven by the rising edge of the clock signal, the input signal propagates through the delay chain step by step; each stage register introduces a fixed time delay, and different delay points are selected as outputs through the phase control word, thereby realizing dynamic phase adjustment of the signal. The total delay time can be expressed as:

[0106] T delay =T clk * reg phase (7)

[0107] In equation (7), T delay T represents the total delay time. clk For the period of the clock signal, reg phase This is the phase control word. The delay time depends on the period of the clock signal and the phase control word. Different delay times are selected through the phase control word, and a phase-shifted reference signal is output.

[0108] The evaluation module 4 in the molten steel level measurement signal processing method in the crystallizer of this invention is responsible for determining whether the phases are aligned. Its detection result directly depends on the initial phase difference between the measured signal and the reference signal. Therefore, after the molten steel level measurement signal processing system in the crystallizer of this invention is powered on, the initial phase needs to be set, and real-time acquisition and evaluation are performed through the ADC sampling module 3. When the initial phase difference between the reference signal and the measured signal is adjusted to zero, the cosine value of the two reaches its maximum, that is, cos(φ) in equation (5). s -φ r =1. At this point, the signal output amplitude is at its maximum, and the system detection accuracy also reaches its optimal level.

[0109] To achieve this functionality, evaluation module 4 includes a search process, which is as follows:

[0110] Initialization phase: Determine the required initial conditions and parameters, including the search range, the frequency and phase of the initial reference signal.

[0111] Coarse Search Phase: Within one complete cycle of the reference signal (0°–360°), the signal conditioning module gradually adjusts the phase of the reference signal, starting from 0°, in 22.5° steps. After each step, the reference signal is multiplied by the liquid level signal, low-pass filtered, and then sent to the ADC module for sampling. The evaluation module determines the alignment degree of the current phase based on the magnitude of the sampled amplitude, and accordingly determines a rough optimal phase range.

[0112] Fine-tuning search phase: Using the phase obtained in the coarse search phase as the center, expand the search range to the left and right by 45° to form a search range. Within this range, make more detailed adjustments in steps of 2.8125°. After each step, repeat the same process: multiplier mixing, low-pass filtering, ADC sampling, and the evaluation module evaluates the phase alignment based on the amplitude, thereby further narrowing down the possible optimal phase range.

[0113] Precise search phase: Within the phase center value obtained in the fine search phase (±5.625°), a higher resolution search is performed with a step size of 0.3515625°. Similarly, after each sampling step, the evaluation module determines the signal amplitude and finally locks the optimal phase point, achieving precise alignment.

[0114] Once the precise search is completed and the optimal phase is determined, the system records the phase parameter, thus completing the entire automatic alignment process.

[0115] When the voltage value acquired by ADC sampling module 3 reaches its maximum value, it indicates that the phases of the reference signal and the signal under test are fully aligned, and the system enters the optimal detection state, thus verifying the correctness and accuracy of the detection. The output signal after phase alignment can be expressed by the following formula:

[0116]

[0117] The signal processing method for measuring molten steel level in the crystallizer of this invention can reduce signal distortion caused by phase mismatch by optimizing the phase matching of the reference signal and the induction signal, thereby enhancing the accuracy and stability of signal detection.

[0118] Through the design and optimization of each signal processing module in the molten steel level measurement signal processing in the crystallizer of this invention, the system can meet the requirement of automatic phase alignment after changing different sensors and crystallizers.

[0119] Example

[0120] In this embodiment, in the crystallizer measuring instrument, when the instrument is initially powered on, the signal conditioning module has a default value and generates a default reference signal, which is multiplied with the liquid level signal in the phase-sensitive detector, and after low-pass filtering and ADC sampling, it reaches the evaluation module. The evaluation module gradually executes the automatic phase alignment search process to ensure that the reference signal and the signal to be measured achieve the best match, thereby realizing high-precision measurement.

[0121] The signal processing method for measuring the molten steel level in the crystallizer in this embodiment is as follows: Figure 2 As shown.

[0122] In this embodiment, the evaluation module 4 in the molten steel level measurement signal processing system in the crystallizer controls the signal adjustment module through frequency control word and phase control word to generate a reference signal and adjust its phase and frequency.

[0123] In this embodiment, the signal conditioning module 5 in the molten steel level measurement signal processing system in the crystallizer is implemented in an FPGA. It uses a 32-bit accumulator with a main frequency of 50MHz. The maximum counting frequency it can achieve is 25MHz, and the minimum frequency is 0.012Hz, which meets the system's requirements for the stability and accuracy of the reference signal frequency.

[0124] Regarding phase control, the molten steel level measurement signal processing system in this embodiment uses a delay chain to achieve phase adjustment. The clock period of the delay chain is 1.25μs, and it contains 1024 register stages. The total delay time of the output signal of a certain register stage in the delay chain can be calculated using the following formula:

[0125] T delay =1.25*reg phase (us) (9)

[0126] In equation (9), reg phase This is the delay chain control word, used to select a specific register in the delay chain as the source of the output signal. The typical period of the liquid level signal is 1250 μs, while the total delay time of the delay chain registers in the molten steel level measurement signal processing system within the crystallizer of this embodiment can reach a maximum of 1280 μs, exceeding the period of the liquid level signal. Therefore, by adjusting reg... phase In this embodiment, the molten steel level measurement signal processing system in the crystallizer can achieve full-range phase adjustment from 0 to nearly 360 degrees, thereby ensuring that the reference signal and the signal to be measured are accurately aligned in phase.

[0127] In this embodiment, the phase-sensitive detector module 1 in the molten steel level measurement signal processing system within the crystallizer uses an AD630 as the core multiplier for phase-sensitive detection, with its gain configured to 1. During the initial power-on phase, the frequency control word of the reference signal is consistent with the liquid level signal, and the phase control word is set to 0, meaning the reference signal and liquid level signal have the same frequency and are directly multiplied. At this time, since crystallizer vibration interference has not yet been introduced, the input signal mainly consists of the liquid level signal. After phase-sensitive detection, the output signal exhibits a modulated waveform containing liquid level information, reflecting the initial state of the liquid level.

[0128] like Figure 4 The image shows the signal waveform when the instrument is powered on. However, because the reference signal and the liquid level signal are not perfectly aligned in the initial stage, the output amplitude of the detector signal after low-pass filtering has not yet reached its maximum value, resulting in an inability to fully reflect the true characteristics of the liquid level signal. Therefore, to improve measurement accuracy, the phase of the reference signal needs to be optimized and adjusted.

[0129] In this embodiment, the low-pass filter 2 in the molten steel level measurement signal processing system within the crystallizer is a second-order Sallen-Key filter with a cutoff frequency set to 10Hz to effectively filter the level signal. A FET-type operational amplifier is selected in the filtering circuit; its high input impedance not only prevents signal attenuation during transmission but also significantly reduces the impact of capacitor leakage current, thereby improving the filter's stability and performance.

[0130] In this embodiment, the ADC sampling module 3 in the molten steel level measurement signal processing system within the crystallizer uses the AD7710 chip, supporting high-precision and high-dynamic-range signal conversion. Since the Sallen-Key filter cutoff frequency is set to 10Hz, and the ADC sampling frequency is set to 100Hz, this design effectively suppresses aliasing noise while ensuring the real-time performance and accuracy of the level signal detection, thus meeting the high-precision requirements for crystallizer level measurement.

[0131] In this embodiment, the evaluation module 4 in the molten steel level measurement signal processing system within the crystallizer uses a step-by-step search mechanism to find the optimal phase value that maximizes the output amplitude of the ADC module. The core idea is to dynamically adjust the phase control word of the phase adjustment module, thereby achieving precise alignment between the reference signal and the signal under test. The search process is as follows:

[0132] 1) Initialize variables

[0133] CurrentPhase: The phase value currently being tested.

[0134] AdcValue: The AD7710 read value corresponding to the current phase.

[0135] MaxAdcValue: Records the maximum value read during the current search process.

[0136] PhaseStart and PhaseEnd: The start and end ranges of the phase search.

[0137] OptimalPhase: The optimal phase value found at the end.

[0138] 2) Rough search

[0139] Within the search range of 0 to 1023, a relatively large step size of 64 is used for fast searching: the evaluation module gradually adjusts the value of CurrentPhase, setting the corresponding parameters to the signal conditioning module. The ADC output value AdcValue is read and compared with the currently recorded MaxAdcValue. If AdcValue exceeds the historical maximum value, MaxAdcValue and the corresponding OptimalPhase are updated.

[0140] 3) Fine-grained search

[0141] Near the maximum amplitude determined in the first step, reduce the step size to 8, and adjust the search range to CurrentPhase ±128. Repeat the phase setting, ADC value reading, and update the maximum value and OpticalPhase.

[0142] 4) Precise Search

[0143] Based on the refined search results, the step size is further reduced to 1, narrowing the search range to OptimalPhase±16. The step size is further reduced to 1, and CurrentPhase is adjusted step by step. The ADC output value is read to ensure that the exact maximum value is found and OptimalPhase is further optimized.

[0144] 5) Set the optimal phase and return the result.

[0145] Finally, the optimal phase value (Opt imalPhase) is written to the signal conditioning module, and this value is returned for subsequent liquid level signal measurement. After the automatic alignment program completes, the waveform is as follows: Figure 5 As shown.

[0146] In summary, the present invention provides a signal processing system and method for measuring molten steel level in a crystallizer. It employs a phase-sensitive detection scheme combining dynamic phase search and automatic alignment, along with a high-precision phase control mechanism and a staged search algorithm, to achieve accurate phase matching between the reference signal and the level signal. Through a three-stage strategy combining coarse, fine, and precise searches, the accuracy and anti-interference capability of the level measurement are effectively improved. The coordinated design of phase-sensitive detection, low-pass filtering, and a high-precision ADC not only optimizes signal processing performance but also ensures the stability and reliability of the system under complex environments.

[0147] Those skilled in the art should recognize that the above embodiments are merely illustrative of the present invention and are not intended to limit the present invention. Any variations or modifications to the above embodiments that are within the spirit and essence of the present invention will fall within the scope of the claims of the present invention.

Claims

1. A signal processing system for measuring molten steel level in a crystallizer, characterized in that, include: Sensors acquire the liquid level signal of molten steel in the crystallizer; The phase-sensitive detection module acquires the liquid level signal and extracts the modulation waveform; A low-pass filter is used to extract the fundamental frequency component related to the liquid level information from the modulated waveform; The ADC sampling module converts the signal after passing through the low-pass filter into a digital signal; The evaluation module analyzes the signal from the ADC sampling module to determine the actual situation of the molten steel level. The signal conditioning module acquires the optimized reference signal from the evaluation module and returns the reference signal to the phase-sensitive detection module to re-match it with the input signal of the sensor, thereby achieving phase matching between the reference signal and the sensed signal.

2. The steel liquid level measurement signal processing system in the crystallizer according to claim 1, characterized in that, The signal conditioning module includes: The reference signal is generated by using a frequency accumulator and a frequency control word to achieve precise control of the signal frequency and generate a stable reference square wave signal. The reference signal phase is shifted, and an adjustable phase reference signal output is provided through the FPGA module.

3. The steel liquid level measurement signal processing system in the crystallizer according to claim 2, characterized in that: The adjustable phase of the FPGA module is achieved through a phase delay chain.

4. A method for processing molten steel level measurement signals in a crystallizer based on the crystallizer molten steel level measurement signal processing system according to claim 1, characterized in that: The sensor acquires the liquid level signal of the molten steel in the crystallizer, and multiplies it with the generated reference signal in the phase-sensitive detection module to extract the modulation waveform containing the liquid level information. The modulated waveform signal enters the low-pass filter to filter out high-frequency noise and extract the fundamental frequency component related to the liquid level information. The filtered signal is converted into a digital signal by the ADC sampling module and transmitted to the evaluation module. The evaluation module analyzes the sampled signal to determine the actual situation of the molten steel level, and generates an optimized reference signal by sending a control word to the signal conditioning module. The optimized reference signal is returned to the phase-sensitive detection module and re-matched with the input signal of the sensor to achieve phase matching between the reference signal and the sensed signal.

5. The method for processing molten steel level measurement signals in a crystallizer according to claim 4, characterized in that, In the phase-sensitive detection module, the input signal is the liquid level signal of the crystallizer, and its expression is: V in (t)=A s sin(ω s t+φ s ) (1) In equation (1), V in (t) is the input signal, A s ω is the amplitude of the input signal. s φ is the angular frequency of the input signal. s The phase of the input signal; The reference signal for phase-sensitive detection is: In equation (2), V ref (t) is the input signal, A r ω is the amplitude of the reference signal. r φ is the angular frequency of the reference signal. r The phase of the reference signal is given by k, and the harmonic order is given by k. The input signal is mixed with the reference signal to obtain the product signal: V mix (t)=V in (t)·V ref (t) (3) In equation (3), V mix (t) represents the output signal of the phase-sensitive detector. Based on the product-sum formula for trigonometric functions, the mixed signal is expanded as follows:

6. The method for processing the molten steel level measurement signal in the crystallizer according to claim 5, characterized in that, The output signal after being filtered by the low-pass filter is expressed as follows: In equation (5), V LP The output signal of the low-pass filter, φ s -φ r This represents the phase difference between the input signal and the reference signal.

7. The method for processing molten steel level measurement signals in a crystallizer according to claim 6, characterized in that, In the precise search phase, the same process is repeated after each step: The phase-sensitive detector module mixes the signal, the low-pass filter filters the signal, the ADC sampling module samples the signal, and the evaluation module evaluates the phase alignment based on the amplitude.

8. The method for processing the molten steel level measurement signal in the crystallizer according to claim 4, characterized in that, The signal conditioning module includes: The reference signal is generated by using a frequency accumulator and a frequency control word to achieve precise control of the signal frequency and generate a stable reference square wave signal. The reference signal phase is shifted, and an adjustable phase reference signal output is provided through the FPGA module.

9. The method for processing the molten steel level measurement signal in a crystallizer according to claim 8, characterized in that, The frequency control word determines the increment of the accumulator for each clock cycle, and its value is determined by both the target output frequency and the system clock frequency, calculated as follows: In equation (6), N is the bit width of the accumulator, f out f is the target output signal frequency. clk This is the system clock frequency.

10. The method for processing the molten steel level measurement signal in a crystallizer according to claim 8, characterized in that, The FPGA module adjusts the phase through a phase delay chain.

11. The method for processing the molten steel level measurement signal in a crystallizer according to claim 10, characterized in that, The phase delay chain is based on LCELL chains connected in series to form a signal transmission path, specifically including: Driven by the rising edge of the clock signal, the input signal propagates through the delay chain step by step; each stage register introduces a fixed time delay, and different delay points are selected as outputs through the phase control word, thereby realizing dynamic phase adjustment of the signal. The total delay time can be expressed as: T delay =T clk * reg phase (7) In equation (7), T delay T represents the total delay time. clk For the period of the clock signal, reg phase This is the phase control word.

12. The method for processing the molten steel level measurement signal in the crystallizer according to claim 6, characterized in that, The evaluation module includes a search process, as detailed below: Initialization phase: Determine the required initial conditions and parameters, including the search range, the frequency and phase of the initial reference signal. Coarse search phase: Within one complete cycle of the reference signal, the signal conditioning module gradually adjusts the phase of the reference signal starting from 0° with a step size of 22.5°; After each step, the reference signal and the liquid level signal are multiplied, and then sent to the ADC sampling module after passing through the low-pass filter; the evaluation module determines the alignment degree of the current phase based on the obtained amplitude, and determines an approximate optimal phase range accordingly. Fine search phase: Taking the phase obtained in the coarse search phase as the center, expand to the left and right by 45° to form a search range. Within this range, make further adjustments in steps of 2.8125°. Precise search phase: Within ±5.625° of the phase center value obtained in the fine search phase, a higher resolution search is continued with a step size of 0.3515625°. After completing the precise search phase and determining the optimal phase, the phase parameter is recorded, thus completing the entire automatic alignment process. The phase-aligned output signal can be expressed by the following formula:

Citation Information

Patent Citations

  • Weak magnetic field detection device and method and non-contact current measurement method

    CN118884311A