Graded magnetic separation purification method for zinc powder impurity separation

Abstract

The invention discloses a graded magnetic separation purification method for zinc powder impurity separation, and relates to the technical field of zinc powder impurity separation, and the graded magnetic separation purification method comprises the following steps: constructing an electromagnetic inversion scene of a three-dimensional particle cloud, and deducing to obtain a complex impedance spectrum in a micro-scale region; based on the obtained complex impedance spectrum, an inductive reactance component and a capacitive reactance component are extracted, and an inductance-capacitance parameter matrix reflecting the electromagnetic characteristics of local zinc powder particles is constructed; and taking the inductance-capacitance parameter matrix as input, executing electromagnetic excitation scanning in a continuous wide frequency range, applying an alternating magnetic field of a plurality of frequency points, and synchronously recording the unit energy absorption power density under the corresponding frequency. According to the method, the resonance risk of the zinc powder particle swarm is quantified by constructing a three-dimensional electromagnetic inversion and resonance recognition mechanism, pulse intensity regulation and control are implemented based on the resonance matching intensity coefficient, agglomeration is effectively inhibited, the grading and separation efficiency is improved, the stability and control precision of the magnetic separation process are remarkably enhanced, and continuous and stable preparation of the high-purity zinc powder is achieved.

Description

Technical Field

[0001] This invention relates to the field of zinc powder impurity separation technology, and more specifically to a graded magnetic separation purification method for separating zinc powder impurities. Background Technology

[0002] Zinc powder impurity separation, grading, and magnetic separation purification is a multi-step process for efficiently purifying non-zinc components (such as iron, manganese, oxides, and carbon particles) mixed in zinc powder, aiming to improve the purity and particle size uniformity of zinc powder. First, a separation process initially removes impurities with significant density differences (such as heavy minerals or light organic impurities). Then, grading is performed to finely separate the zinc powder according to particle size, removing excessively coarse or fine particles and improving particle distribution consistency. Next, magnetic separation is conducted, utilizing the characteristic that impurities often contain magnetic substances (such as ferromagnetic particles) to precisely remove them from the non-magnetic zinc powder under the influence of a magnetic field. Finally, a purification stage, including chemical cleaning, airflow dust removal, or high-temperature reduction, further removes residual trace impurities and oxide films, resulting in a high-quality zinc powder product with uniform particle size, high purity, and stable physical properties. The entire process combines physical and chemical separation mechanisms and is widely used in industrial fields with stringent zinc powder quality requirements, such as high-end alloys, battery materials, and chemical catalysts.

[0003] The existing technology has the following shortcomings: During the separation of impurities in zinc powder, zinc powder particles are simultaneously affected by induced current excitation and electrostatic coupling effects under the action of a high-frequency alternating magnetic field. If the magnetic field frequency resonates with the inductance-capacitance parameters (LC structure) of the local particle group, electromagnetic resonance can easily be induced in the local particle group. This resonance effect significantly enhances the instantaneous electromagnetic attraction between particles, causing nonlinear agglomeration behavior of zinc powder particles, which in turn forms abnormally large particle clusters. This not only seriously deviates from the predetermined particle size standard and destroys the uniformity of particle size distribution, but also easily encapsulates or embeds some magnetic or non-magnetic impurity particles inside the clusters, significantly increasing the difficulty of subsequent impurity removal. This reduces the overall impurity separation efficiency and the final zinc powder purity, resulting in impurity enrichment effect and separation performance fluctuations during the process, seriously affecting the stable preparation process of high-purity zinc powder.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a graded magnetic separation purification method for zinc powder impurity separation. By constructing a three-dimensional particle cloud electromagnetic inversion scenario, combined with complex impedance spectroscopy extraction and resonance fingerprint modeling, the electromagnetic resonance characteristics of zinc powder particles in a high-frequency alternating magnetic field are accurately identified, and the agglomeration risk is quantified using the resonance matching intensity coefficient. Based on this, an amplitude gradient control function is established, and a frequency-band differentiated pulse intensity modulation strategy is introduced to achieve dynamic control of the magnetic drive excitation parameters. This method can effectively suppress high-risk frequency bands and enhance the magnetic separation efficiency in low-risk frequency bands throughout the process, avoiding particle agglomeration problems caused by resonance in traditional magnetic separation. It significantly improves the consistency of zinc powder particle size classification, impurity removal efficiency, and process stability, ensuring the continuous and stable preparation of high-purity zinc powder, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a graded magnetic separation purification method for separating zinc powder impurities, comprising the following steps: In the process of separating zinc powder impurities, an electromagnetic inversion scenario of a three-dimensional particle cloud is constructed. By setting alternating magnetic field parameters and spatial detection arrangements, the phase distribution of the magnetic field vector and the change of instantaneous potential difference of zinc powder particles in the applied magnetic field are collected simultaneously. Then, the complex impedance spectrum representing the electromagnetic properties of the zinc powder particle group in the microscale region is derived. Based on the obtained complex impedance spectrum, a multi-scale adaptive recursive algorithm is used to dynamically decompose the impedance data and extract the inductive reactance and capacitive reactance components of the corresponding zinc powder particle swarm. Based on this, an inductor-capacitor parameter matrix containing the electromagnetic properties of the local particle swarm is constructed to describe the equivalent electromagnetic response structure of the local zinc powder particles under the action of a magnetic field. Using the constructed inductor-capacitor parameter matrix as the input basis, an electromagnetic excitation scanning operation is performed over a continuous wide frequency range. Alternating magnetic fields at multiple frequency points are applied in real time, and the unit energy absorption power density absorbed by the zinc powder particle swarm at the corresponding frequency points is recorded simultaneously to obtain the initial resonant response profile characterizing the electromagnetic response behavior of the particle swarm. Based on the obtained initial resonance response profile, an improved Lorentz function fitting process is performed to extract the full width at half maximum (FWHM) parameter and peak energy absorption power of the main resonance peak of the particle swarm. At the same time, environmental parameters are introduced for attenuation factor correction, and finally a resonance response fingerprint model that can quantitatively describe the resonance characteristics of zinc powder particle swarm is formed. Based on the resonance response fingerprint model and the set target magnetic field driving frequency, the peak energy absorption power ratio and phase synchronization rate corresponding to the frequency are calculated, and the product of the two is used as the quantitative definition of the resonance matching strength coefficient to characterize the possibility and intensity of electromagnetic resonance and induced agglomeration behavior of zinc powder particles at a specific frequency. Based on the coupling relationship between the constructed resonance matching strength coefficient and the spatial distribution characteristics of the particle swarm, an amplitude gradient control function is designed, and a frequency band-differentiated pulse intensity modulation strategy is introduced. For high-risk frequency bands where the resonance matching strength coefficient is higher than a set threshold, adaptive pulse intensity suppression control is implemented to weaken the magnetic drive intensity in that frequency band and avoid particle agglomeration. For frequency bands where the resonance matching strength coefficient is lower than a safety threshold, adaptive pulse intensity enhancement intervention is implemented to improve magnetic separation efficiency and particle swarm response control effect, thereby achieving differentiated magnetic drive regulation and stable graded purification control of zinc powder particle swarm throughout the process.

[0007] Preferably, step S101 includes: A controllable three-dimensional magnetic field excitation region was constructed, the frequency and amplitude range of the alternating magnetic field were set, and a magnetic field vector acquisition device and a potential measurement electrode were deployed in the zinc powder particle distribution area. The zinc powder sample was uniformly dispersed in the area, and the magnetic field vector phase distribution and the instantaneous potential difference of the particles were collected simultaneously under magnetic field excitation. Based on the collected data, a spatiotemporal joint inversion algorithm was used for analysis to obtain the complex impedance spectrum representing the electromagnetic behavior of zinc powder particle groups in the microscale region.

[0008] Preferably, step S102 includes: The complex impedance spectrum is reconstructed in the frequency domain to construct a complex function array containing changes in the real and imaginary parts at multiple frequency points, and then subjected to multi-scale decomposition and sliding window processing. For the impedance data of each frequency sub-interval, an adaptive recursive algorithm is used for dynamic decomposition to extract the inductive reactance component and the capacitive reactance component. Based on the extraction results, an inductance-capacitance parameter matrix for each spatial region is constructed to reflect the equivalent electromagnetic response characteristics of different particle swarms.

[0009] Preferably, step S103 includes: Using the inductor-capacitor parameter matrix as input, a frequency response prediction model is established to determine the electromagnetic excitation frequency range and step accuracy. An alternating magnetic field of constant amplitude is applied at the set frequency point, and the unit energy absorption power density absorbed by the zinc powder particle group at each frequency is recorded simultaneously. The duration of excitation at each frequency is greater than twice the corresponding particle swarm response delay. The power is normalized using an integral energy measurement method. After removing background electromagnetic interference, the power density data are arranged in ascending order of frequency to form an initial resonance response profile curve, which is used for subsequent resonance identification and control strategy formulation.

[0010] Preferably, step S104 includes: The initial resonance response profile is smoothed and denoised to extract the target frequency band of the main resonance peak. For the main resonant frequency band, an improved Lorentz function containing an offset term and a stretching factor is used for nonlinear fitting to extract the center frequency, full width at half maximum (FWHM), and peak power. An environmental attenuation correction factor is introduced based on the airfield baseline data to correct the error of the fitting parameters; A resonance response fingerprint model was constructed based on the modified parameters to reflect the resonance characteristics and energy response trend of zinc powder particle groups at a specific frequency.

[0011] Preferably, step S105 includes: The main resonance peak parameters are extracted from the resonance response fingerprint model, and the target magnetic field driving frequency is set for matching analysis. Calculate the percentage of energy absorbed per unit frequency to characterize the intensity of energy resonance excitation; Extract the potential response signals at multiple locations at the target frequency and calculate the phase synchronization rate to reflect the response consistency of the particle swarm. Multiplying the energy absorption power ratio by the phase synchronization rate yields the resonance matching strength coefficient.

[0012] Preferably, step S106 includes: A model for the coupling relationship between the resonance matching strength coefficient and the spatial distribution of zinc powder particle swarm was established, and a resonance risk heatmap was constructed. Design an amplitude gradient control function with the resonant matching strength coefficient as input and the magnetic field amplitude as output to achieve continuous dynamic adjustment of the excitation amplitude at the frequency point; The frequency band risk level is divided according to the resonance matching strength coefficient, and pulse intensity suppression or enhancement strategies are implemented respectively to adjust the pulse amplitude, frequency and interval. By combining amplitude gradient control and pulse modulation strategies throughout the entire process, spatial frequency coordinated control of the excitation parameters of zinc powder magnetic drive is achieved.

[0013] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention constructs an electromagnetic inversion scenario of a three-dimensional particle cloud, combining complex impedance spectroscopy extraction, frequency response scanning, and resonance fingerprint modeling to comprehensively identify the electromagnetic resonance characteristics of zinc powder particle groups under high-frequency alternating magnetic fields. Furthermore, by quantitatively defining the resonance matching intensity coefficient, it achieves accurate assessment of particle agglomeration risk. Based on this, according to the coupling relationship between the resonance matching intensity coefficient and the spatial distribution of the particle group, an amplitude gradient control function is designed, and a frequency-band differentiated pulse intensity modulation strategy is introduced. This enables dynamic control of the magnetic drive excitation parameters, effectively suppressing high-risk frequency bands and enhancing the processing efficiency of low-risk frequency bands throughout the process. This not only successfully avoids the nonlinear agglomeration of zinc powder particles induced by electromagnetic resonance in traditional magnetic separation processes, significantly improving the consistency of zinc powder particle size classification and the thoroughness of impurity separation, but also greatly enhances the response stability and fine control capability of the entire process, thereby achieving continuous and stable preparation of high-purity zinc powder. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0015] Figure 1 This is a flowchart of the graded magnetic separation purification method for separating zinc powder impurities according to the present invention. Detailed Implementation

[0016] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0017] This invention provides, for example Figure 1 The graded magnetic separation purification method for separating zinc powder impurities, as shown, includes the following steps: S101. During the separation of zinc powder impurities, an electromagnetic inversion scenario of a three-dimensional particle cloud is constructed. By setting alternating magnetic field parameters and spatial detection arrangements, the phase distribution of the magnetic field vector and the change of instantaneous potential difference of zinc powder particles in the applied magnetic field are collected simultaneously. Then, the complex impedance spectrum representing the electromagnetic properties of the zinc powder particle group in the microscale region is derived. To address the issue of zinc powder particle agglomeration caused by electromagnetic resonance induced by high-frequency alternating magnetic fields, thus affecting the fractional purification effect, a controllable three-dimensional magnetic field distribution region is first constructed using a spatially adjustable magnetic field excitation structure. Within this region, a continuously variable alternating magnetic field is applied within a set frequency and amplitude range, forming a spatial magnetic field excitation scene with directional and phase characteristics. Simultaneously, by deploying multi-point magnetic field vector acquisition devices and high-sensitivity potential measurement electrodes within the zinc powder particle distribution region, a particle observation area with three-dimensional response capability is constructed to capture the instantaneous response behavior of zinc powder particles in the spatial magnetic field.

[0018] After establishing the three-dimensional magnetic field excitation and particle observation area, the zinc powder sample is uniformly dispersed within this space, controlling its particle size range and concentration within a predetermined range, and data is synchronously acquired under the action of the excitation magnetic field. Specifically, multiple spatially distributed magnetic field vector probes are used to measure the vector magnetic field distribution of zinc powder particles in the alternating magnetic field at each moment, obtaining complete magnetic field vector change information including direction, amplitude, and phase; simultaneously, micro-potential difference acquisition electrodes deployed at different locations in the particle area are used to record the instantaneous potential difference changes between particles driven by the alternating magnetic field, thereby achieving synchronous observation of magnetic response behavior and potential behavior.

[0019] Based on the collected magnetic field vector phase distribution data and particle instantaneous potential difference change data, combined with the spatial distribution of particles and the response sequence in the time dimension, a spatiotemporal joint analytical method is used to inversely analyze the electromagnetic response behavior of zinc powder particle groups. Through multidimensional Fourier transform and generalized impedance transform techniques, the dynamic magnetic field response information and potential change data are transformed into complex impedance functions representing the electromagnetic behavior of the zinc powder particle group in a microscale region, thereby obtaining an important parameter reflecting the electromagnetic properties of this region—the complex impedance spectrum. This spectrum information simultaneously reflects the frequency response characteristics, energy absorption capacity, and phase coupling degree of the particle group, and can reflect whether zinc powder particles may form an LC equivalent resonance structure under the current magnetic field excitation conditions.

[0020] To ensure the reliability and representativeness of the complex impedance spectrum, noise filtering and background signal removal are necessary for the acquired data. Specifically, wavelet denoising and phase drift correction are applied to the magnetic field acquisition signal to ensure the temporal consistency of the magnetic field vector phase data; baseline correction and time window normalization are performed on the potential difference data to improve the accuracy and stability of the inversion results. The complex impedance spectrum constructed based on this can serve as crucial foundational data for subsequent extraction of inductive-capacitive parameters, construction of the inductance-capacitance matrix, and prediction of resonance behavior. This ensures proactive identification and precise control of the risk of electromagnetically induced agglomeration during zinc powder impurity separation, enhancing the stability and purity assurance capabilities of the graded magnetic separation purification process.

[0021] S102. Based on the obtained complex impedance spectrum, the impedance data is dynamically decomposed using a multi-scale adaptive recursive algorithm to extract the inductive reactance component and capacitive reactance component of the corresponding zinc powder particle group. Based on this, an inductor-capacitor parameter matrix containing the electromagnetic properties of the local particle group is constructed to describe the equivalent electromagnetic response structure of the local zinc powder particles under the action of a magnetic field. To effectively identify and quantify the electromagnetic characteristic parameters of zinc powder particle groups that may induce resonance behavior under an alternating magnetic field, high-precision modeling and analysis of the previously acquired complex impedance spectrum data are first required. To this end, the complex impedance spectrum is first reconstructed in the frequency domain to construct a complex function array containing the variation characteristics of the real and imaginary parts at multiple frequency points. Then, a multi-scale decomposition strategy is introduced to hierarchically divide the complex impedance spectrum according to the frequency dimension. A sliding window processing method is applied within each frequency sub-interval to extract local extrema, phase inflection points, and impedance transition points for subsequent refined feature extraction. The purpose of this step is to establish a high-resolution impedance structure that preserves the global frequency response trend while highlighting local anomalous response characteristics, providing a valid data foundation for the subsequent recursive algorithm.

[0022] After completing the multi-scale frequency domain partitioning, an improved adaptive recursive algorithm is used for dynamic decomposition of the impedance data within each frequency interval. This algorithm comprehensively considers the frequency response stability, phase change rate, and energy density distribution of the impedance spectrum. By constructing a dynamic state transition function centered on impedance admittance transformation, it achieves frequency-wise extraction of the inductive reactance component (corresponding to inductive response) and capacitive reactance component (corresponding to capacitive response) in the complex impedance spectrum. In each recursive operation, the algorithm automatically determines the dominant response characteristics of the current frequency band based on the local phase gradient and amplitude jump trend, and adjusts the decomposition parameters in real time to adapt to changes in frequency scale and particle population behavior, thereby ensuring the accuracy and stability of the extraction results. This method avoids the problem of aliasing of inductive and capacitive reactance components in multi-particle coupling states using traditional Fourier inversion or Bode plot analysis, effectively identifying the true physical response modes hidden behind the complex impedance spectrum. The improved adaptive recursive algorithm employed can specifically utilize a complex impedance component estimation algorithm based on the extended Kalman filter framework as its implementation path. This algorithm introduces nonlinear state-observation relationship modeling capabilities on top of the state estimation of traditional Kalman filtering, making it suitable for processing dynamic response signals such as complex impedance spectra that exhibit amplitude-phase coupling and are sensitive to frequency changes. In its implementation, the trends of the real and imaginary parts of the impedance spectrum with frequency variation can be used as the system state vector, the observed values ​​of the complex impedance as the measurement vector, and a time-varying model of inductive and capacitive reactance (such as a first-order Markov process) can be introduced as a priori structure to construct nonlinear state transition equations and observation equations. By recursively updating the state estimate and error covariance matrix within each frequency step, the algorithm can extract the dominant components of the impedance response during frequency changes in real time, thereby achieving high-precision dynamic separation of inductive and capacitive reactance components. Furthermore, to adapt to multi-scale feature analysis, this complex impedance component estimation algorithm based on the extended Kalman filter framework can also be combined with wavelet transform preprocessing or variational mode decomposition (VMD) to perform scale separation and noise suppression of the signal, improving decomposition accuracy and stability. Therefore, this improved adaptive recursive algorithm combines nonlinear modeling, dynamic estimation, and adaptive noise filtering capabilities, making it highly suitable for the dynamic multi-scale analytical task of complex impedance spectra in this invention.

[0023] Based on the extracted inductive and capacitive reactance data, an inductance-capacitance parameter structure for the local particle swarm is constructed. Specifically, multiple zinc powder particles within each spatial sub-region are considered as a sub-group with electromagnetic coupling. The average inductive and capacitive reactance values ​​for each region are calculated and converted into equivalent inductance (in Henry) and equivalent capacitance (in Farad) values. A two-dimensional parameter matrix is ​​then constructed, where each element represents the equivalent inductance-capacitance combination characteristics of the zinc powder particle swarm at a specific frequency point. This matrix representation not only reflects the electromagnetic response characteristics of the entire particle swarm under alternating magnetic fields but also characterizes the response differences of particle swarms in different regions under magnetic field influence, providing a theoretical basis for subsequent spatially selective control.

[0024] To further enhance the representativeness and engineering adaptability of the constructed inductor-capacitor parameter matrix, the initial matrix needs to be normalized and its stability verified. Normalization involves scaling each inductor and capacitor parameter in the matrix proportionally to its maximum value to facilitate comparison of diverse characteristics. Stability verification involves constructing a frequency response simulation model to verify whether the local particle swarm represented by the parameter matrix may exhibit resonant behavior under different excitation frequencies. If sharp jumps occur in the matrix eigenvalues ​​at certain frequency points, it indicates a high risk of resonance in that region, requiring close monitoring and intervention in subsequent electromagnetic control stages. Through these methods, the constructed inductor-capacitor parameter matrix not only achieves digital modeling of the response characteristics of microscopic zinc powder particle swarms but also provides an accurate foundation for resonance prediction throughout the entire graded magnetic separation purification process, effectively supporting subsequent resonance matching evaluation and differentiated magnetic drive control operations.

[0025] S103. Using the constructed inductor-capacitor parameter matrix as the input basis, perform electromagnetic excitation scanning operation in a continuous wide frequency range, apply alternating magnetic fields at multiple frequency points in real time, and simultaneously record the unit energy absorption power density absorbed by the zinc powder particle swarm at the corresponding frequency points to obtain the initial resonant response profile characterizing the electromagnetic response behavior of the particle swarm. To accurately obtain the electromagnetic response characteristics of zinc powder particle swarms under alternating magnetic fields and identify potential resonant behavior features, the previously constructed inductance-capacitance parameter matrix is ​​used as the basic input. This matrix contains equivalent inductance and capacitance values ​​in multiple microscale spatial sub-regions and accurately reflects the electromagnetic response capabilities of local zinc powder particles under different frequency excitations. Using this parameter matrix as input, a set of theoretical prediction models describing the frequency response relationship of the zinc powder particle swarm can be established to determine the coverage and step accuracy of subsequent excitation frequencies, thereby ensuring the response integrity and data acquisition effectiveness of the entire excitation scanning process.

[0026] Based on the potential resonant frequency range indicated by the inductance-capacitance parameter matrix, an electromagnetic excitation sequence covering a wide frequency range (e.g., 10 kHz to 10 MHz) is constructed. At each set frequency point, an alternating magnetic field with a constant amplitude but varying frequency is applied. This alternating magnetic field can be achieved using a high-frequency electromagnetic exciter. Simultaneously, the excitation direction of the magnetic field is ensured to be as consistent as possible with the natural alignment of the zinc powder particles to enhance the consistency and measurability of the response. While the magnetic field is applied, multiple energy response detection devices positioned around the three-dimensional distribution space of the particle swarm record the unit absorbed power density of the zinc powder particles at each frequency point in real time. The recording unit is watts per cubic centimeter (W / cm³), and a one-to-one functional relationship is established between this density and the excitation frequency, thereby constructing the energy absorption response spectrum of the zinc powder particle swarm during the magnetic field scanning process.

[0027] To ensure the accuracy and physical meaning of the absorbed power density data, the excitation at each frequency point must meet the steady-state response condition. This means the excitation duration at each frequency should be more than twice the average response delay of the particle swarm at that frequency to avoid the influence of transient disturbances. Simultaneously, an integral energy measurement method is used within each excitation cycle to normalize the overall absorption efficiency of the sample using the average absorbed power density, making subsequent data analysis more comparable. Furthermore, to eliminate environmental electromagnetic interference, background noise power density is simultaneously acquired during the energy absorption data acquisition process and removed in later data processing to ensure that the obtained response curve truly reflects the resonant energy absorption behavior of the zinc powder particle swarm.

[0028] After scanning and recording data across the entire frequency range, the unit energy absorption power density at all frequency points was arranged in ascending order to form a complete initial resonance response profile curve. This curve reflects the overall energy absorption trend of the zinc powder particle group under different excitation frequencies and includes multiple local peak regions, preliminarily revealing the location distribution and response intensity of possible resonant coupling in the particle group. This response profile not only provides fundamental information for identifying the primary and secondary resonance frequencies but also provides highly reliable experimental evidence for subsequent fitting analysis, fingerprint model construction, and the formulation of magnetic field-driven frequency control strategies. The initial resonance response profile obtained in this step can significantly enhance the parameter control capability of the zinc powder classification magnetic separation process, effectively avoid the risk of particle agglomeration induced by electromagnetic resonance, and improve the stability and precision control level of the entire purification process.

[0029] S104. Based on the obtained initial resonance response profile, an improved Lorentz function fitting process is performed to extract the full width at half maximum (FWHM) parameter and peak energy absorption power of the main resonance peak of the particle swarm. At the same time, environmental parameters are introduced to correct the attenuation factor, and finally a resonance response fingerprint model that can quantitatively describe the resonance characteristics of zinc powder particle swarm is formed. After completing the electromagnetic excitation scan of the zinc powder particle swarm within a continuous frequency range and obtaining the initial resonance response profile, in order to further identify the main resonance behavior characteristics of the particle swarm in a specific frequency range and establish a quantitative model, the obtained frequency response curve was first subjected to preliminary smoothing and denoising processing to eliminate random fluctuation signals that might be introduced into the test system. Specifically, a weighted moving average algorithm was used to perform interval filtering on the power density data to ensure that the overall curve trend was smooth without losing the true peak characteristics. Simultaneously, the first derivative of the curve was analyzed to accurately locate all possible local peak points, and the main resonance peak with the largest power amplitude and a clear half-width at half-maximum (HWHM) was selected to determine the target frequency band for fitting analysis. These operations provide an initial parameter estimation interval for the fitting process, thereby ensuring the accuracy and fitting efficiency of the subsequent model solution.

[0030] For the frequency band containing the main resonance peak in the initial resonance response profile, a modified Lorentz function is used for nonlinear fitting. The fitting function not only includes parameters such as the center frequency, peak power, and full width at half maximum (FWHM) of the standard Lorentz distribution, but also introduces a baseline offset term for absorbed power and an asymmetric stretching factor to correct for non-ideal response curve shapes caused by multi-peak coupling or uneven particle distribution. The fitting process is iteratively solved using a least-squares error optimization algorithm, gradually adjusting the function parameters until the error converges below a set threshold. During the fitting process, three key indicators of the main resonance peak are recorded and extracted: the center frequency, representing the frequency point where the particle swarm resonates; the FWHM, representing the frequency span of the resonance peak, reflecting the selectivity and resonance stability of the frequency response; and the peak absorbed power, representing the maximum energy response capability of the particle swarm at that frequency. These indicators together constitute the main resonance characteristics of the zinc powder particle swarm.

[0031] To enhance the adaptability and accuracy of the resonant response model in complex engineering environments, the fitted parameters need to be corrected using an environmental attenuation factor. In this step, we first analyze the non-ideal influencing factors in the test environment, including background magnetic field interference, temperature fluctuations, and electromagnetic interference signals. Based on control experiments, we measure the baseline energy absorption power per unit field under open-field conditions to quantify the background environmental noise. Then, we transform the aforementioned background interference into a mathematically defined exponential attenuation term and incorporate it into the Lorentz fitting function model to correct the errors in the peak power and full width at half maximum (FWHM) parameters. This correction mechanism effectively compensates for the resonance characteristic shift caused by the non-ideal nature of the test system, ensuring that the established model more closely approximates the true resonant response state of the zinc powder particle swarm under ideal, pure magnetic field conditions.

[0032] Based on the corrected resonance fitting parameters, a quantitative fingerprint model describing the resonance characteristics of zinc powder particle swarms is constructed. This model uses frequency as the horizontal axis and unit absorbed power density as the vertical axis, superimposed with a parameter set consisting of the center frequency of the main resonance peak, the full width at half maximum (FWHM), and the peak power. This parameter set characterizes the resonance behavior trend of zinc powder particle swarms under specific magnetic field excitation conditions. Furthermore, a fingerprint matching metric can be introduced to compare the resonance response fingerprint of the current sample with the fingerprint features in the historical sample database, helping to determine whether the batch of zinc powder exhibits abnormal aggregation tendencies or anomalous hierarchical response behavior. This resonance response fingerprint model not only possesses the advantages of visualization and quantification but also provides a basis for identifying resonance risks in the subsequent magnetic field control stage, significantly improving the dynamic control accuracy and safety margin in the magnetic drive purification process. The establishment of this model allows the zinc powder particle swarm to move from its original physical behavior to a measurable response description, laying a solid foundation for the intelligent control of the entire process.

[0033] S105. Based on the resonance response fingerprint model and the set target magnetic field driving frequency, calculate the peak energy absorption power ratio and phase synchronization rate corresponding to the frequency, and use the product of the two as the quantitative definition of the resonance matching strength coefficient to characterize the possibility and intensity of electromagnetic resonance and induced agglomeration behavior of zinc powder particles at a specific frequency. After extracting and fitting the resonance response profile of the zinc powder particle swarm, to further evaluate whether the target magnetic field driving frequency matches the main resonance characteristics of the zinc powder particle swarm and to quantitatively measure the risk intensity of its induced electromagnetic resonance and particle agglomeration behavior, the corresponding main resonance peak parameters, especially the peak energy absorption power and resonance response range, are first extracted from the resonance response fingerprint model, and the target magnetic field driving frequency used in the process is set. This target frequency should cover the frequency domain range formed by the previous continuous frequency scan and be within the adjustable range of the magnetic separator's operating frequency. By analyzing the intersection of this set frequency and the resonance response fingerprint model, it is preliminarily determined whether it falls near the resonance response peak, thereby determining whether further quantitative assessment of resonance intensity is needed.

[0034] For a given target magnetic field driving frequency, the energy absorption power density value at that frequency is extracted from the resonant response fingerprint model. This energy absorption power density value is then normalized to the peak power of the main peak across the entire resonant response frequency range, calculating the proportion of the target frequency energy absorption power in the main resonant response. This peak energy absorption power proportion characterizes the effectiveness of the current magnetic field excitation frequency in energy resonance absorption and is a core indicator describing the "energy matching degree." A higher proportion indicates that the frequency is closer to the main resonant frequency, resulting in a stronger electromagnetic absorption response from the zinc powder particle group under its influence, thus increasing the risk of resonant excitation.

[0035] After completing the energy level assessment, it is also necessary to consider the phase synchronization of zinc powder particles under the excitation of the target frequency, that is, to analyze whether the particle swarm exhibits highly coordinated response behavior due to electromagnetic coupling. To achieve this analysis, a phase synchronization rate calculation method is used to statistically analyze the phase distribution of the response waveform of the zinc powder particle swarm at the target frequency. The specific steps include: extracting the potential response signals collected at multiple locations in space, performing Hilbert transform on them to obtain the instantaneous phase; then calculating the phase concentration index (such as Rayleigh coherence or Kuramoto synchronization) of all sampling points at the target frequency to obtain a phase synchronization rate value between 0 and 1. The higher this value, the stronger the synchronous oscillation trend of the particle swarm at the target frequency, and the higher the probability of resonant cooperative aggregation behavior.

[0036] The resonance matching strength coefficient is calculated by multiplying the obtained peak energy absorption power ratio with the phase synchronization rate. This coefficient serves as a quantitative indicator, describing whether the target magnetic field driving frequency can simultaneously excite the electromagnetic resonance behavior of zinc powder particles in both energy and phase dimensions. It is a key parameter characterizing potential agglomeration risk. To enhance its practicality, risk thresholds can be set in actual applications. For example, frequencies with a resonance matching strength coefficient higher than 0.7 can be considered "high-risk frequencies," requiring intervention measures such as pulse suppression or frequency shifting during subsequent magnetic field application. Conversely, frequencies with a resonance matching strength coefficient lower than 0.3 can be considered "low-risk frequencies," allowing for increased magnetic field amplitude to improve separation efficiency. Establishing a quantitative relationship between the target frequency and resonance risk through this method not only realizes the transformation from physical response to control strategy but also provides intelligent control based on measured response for the entire zinc powder grading and magnetic separation purification process, significantly improving process stability and finished product purity assurance.

[0037] S106. Based on the coupling relationship between the constructed resonance matching strength coefficient and the spatial distribution characteristics of the particle swarm, an amplitude gradient control function is designed, and a frequency band-differentiated pulse intensity modulation strategy is introduced. For high-risk frequency bands where the resonance matching strength coefficient is higher than a set threshold, adaptive pulse intensity suppression control is implemented to weaken the magnetic drive intensity in that frequency band and avoid particle aggregation. For frequency bands where the resonance matching strength coefficient is lower than a safety threshold, adaptive pulse intensity enhancement intervention is implemented to improve magnetic separation efficiency and particle swarm response control effect, thereby achieving differentiated magnetic drive regulation and stable graded purification control of zinc powder particle swarm throughout the process. After constructing the resonance matching strength coefficient based on the zinc powder particle swarm resonance response fingerprint model and the target magnetic field driving frequency, in order to effectively suppress the frequency band that may induce electromagnetic resonance agglomeration risk, while making full use of the safe frequency band to improve magnetic separation efficiency and particle size stability control, it is first necessary to establish a coupling relationship model between the resonance matching strength coefficient and the spatial distribution of the zinc powder particle swarm. Specifically, by combining the three-dimensional distribution data of the particle swarm and the local inductance-capacitance parameter matrix, the resonance matching strength coefficient of the zinc powder particles in each spatial sub-region is spatially mapped. The high-risk and low-risk regions of each frequency point in space are identified and classified separately, and a resonance risk heatmap is constructed using spatial visualization. This heatmap reflects the potential intensity of particle agglomeration in each region under specific frequency excitation, and is the basis for subsequent magnetic field amplitude control strategy design.

[0038] Based on the above coupling model, a continuously differentiable amplitude gradient control function is designed. This function takes the resonant matching strength coefficient as the input variable and the applied magnetic field amplitude as the output, constructing a set of frequency-amplitude response adjustment rules. The function design must satisfy two basic characteristics: first, a monotonically decreasing characteristic, meaning that as the resonant matching strength coefficient increases, the corresponding magnetic drive excitation amplitude automatically decreases; second, boundary stability characteristics, meaning that when the resonant matching strength coefficient is below a set safety threshold, the magnetic field amplitude is allowed to gradually increase within an adjustable range. To achieve continuous adjustment control, an exponential inverse function form is preferred for design; for example, the magnetic field amplitude can be set as: ,in: The magnetic field amplitude at frequency f. This is the resonance matching strength coefficient corresponding to this frequency. To adjust the sensitivity coefficient, This represents the upper limit of the amplitude. By implementing the above amplitude gradient function, real-time control of the excitation intensity can be achieved at each frequency point, ensuring that the magnetic drive energy input in high-risk frequency bands is effectively suppressed.

[0039] To further enhance the flexibility and dynamic response of the control, a frequency-band differentiated pulse intensity modulation strategy is introduced, building upon continuous amplitude control. Specifically, the entire frequency range is divided into several sub-bands, and each sub-band is further categorized into high-risk, intermediate transition, and low-risk bands based on the average resonant matching strength coefficient. For high-risk bands (i.e., those with an average resonant matching strength coefficient higher than 0.7), a pulse intensity suppression control strategy is employed. This strategy focuses on reducing the peak amplitude of the pulse magnetic field and increasing the pulse interval to weaken the energy input density, thereby preventing nonlinear agglomeration of particles under high resonance risk conditions. For low-risk bands (i.e., those with a resonant matching strength coefficient lower than 0.3), a pulse intensity enhancement strategy is used, including increasing the peak amplitude of the pulse magnetic field, increasing the pulse repetition frequency, or shortening the pulse interval. This enhances the response capability and impurity stripping efficiency of zinc powder particles in this band, thus improving the classification effect. In the intermediate-risk bands, the pulse amplitude and rhythm can be fine-tuned according to the resonant matching strength coefficient to maintain a smooth transition in the control.

[0040] This approach combines amplitude gradient control functions with frequency-band differentiated pulse modulation strategies throughout the zinc powder impurity separation and purification process, achieving differentiated magnetic drive control across the entire process. By using the resonant matching strength coefficient as a dynamic feedback input, the magnetic field application strategy is dynamically adjusted with each frequency switch and spatial position change, ensuring the magnetic drive process remains in a relatively safe and efficient operating state. Specifically, a sliding frequency window and spatial tracking algorithm are employed to continuously monitor the changing trends between excitation parameters and matching strength. This ensures that when a frequency is identified as high-risk in a certain region, excitation amplitude reduction and pulse energy suppression operations are immediately triggered; conversely, when identified as a low-risk region, excitation enhancement is implemented, thus forming a dynamically adaptive control grid throughout the magnetic separation space. This strategy effectively avoids the resonance excitation error and particle agglomeration problems caused by uniform excitation in traditional magnetic separation processes, improving the spatial resolution and local adjustment capability of the process.

[0041] This implementation method establishes a refined control strategy based on resonant matching strength as the core indicator, spatial distribution response as the foundation, and amplitude and pulse joint adjustment as the means. It realizes the optimization of magnetic drive response and stable control of particle behavior throughout the zinc powder impurity separation process, significantly improving the classification accuracy, impurity stripping efficiency and product purity consistency, and has good engineering adaptability and industrial application prospects.

[0042] The following is a further experimental verification and detailed analysis of the effectiveness of the technical solution of the present invention. Based on different working conditions and control conditions, a systematic comparison is made in terms of zinc powder purity, particle size distribution consistency, impurity removal efficiency, resonance agglomeration control capability, and multi-particle size adaptability. Combined with four embodiments, the technical advantages and significant effects of the present invention in practical engineering applications are fully demonstrated: [Example 1: Comparative Experiment of the Method of the Present Invention and Traditional Constant Frequency Magnetic Separation] In this embodiment, a batch of industrial zinc powder containing impurities was used as the subject. Its initial average particle size was approximately 60 μm, and the impurities mainly consisted of ferromagnetic particles, manganese oxides, and a small amount of carbon black particles. The raw materials were subjected to graded purification treatment using both conventional constant-frequency magnetic separation and the method described in this invention. The conventional method sets the magnetic separation frequency to a fixed 300 kHz and uses a constant amplitude magnetic field. It lacks a real-time monitoring and intervention mechanism for frequency response and fails to identify electromagnetic resonance behavior that may be excited at specific frequencies during the treatment process, leading to agglomeration of some particles and a significant shift in particle size distribution. The final results showed that the purity of the zinc powder increased to 98.3% after treatment, the particle size D90 shifted from the original 64 μm to 72 μm, and the standard deviation of the particle group increased to ±9.2 μm, indicating that agglomeration led to a deterioration in particle size uniformity. By adopting the method of this invention, a frequency identification mechanism and dynamic magnetic drive control logic based on complex impedance spectrum were established throughout the process, avoiding energy overload in the excitation frequency band. The final zinc powder purity reached 99.6%, the particle size D90 was 65μm, and the standard deviation was only ±3.7μm, indicating that the particle distribution was more concentrated and the purification effect was significantly better than that of traditional methods.

[0043] [Example 2: Comparison of the dynamic control strategy of the present invention with traditional linear frequency conversion magnetic separation] This experiment aimed to evaluate the response efficiency and aggregation control capability of the differentiated pulse modulation mechanism in this invention during dynamic processes. Two control groups were set up: one group used a traditional linear frequency conversion mode, with the frequency uniformly stepping from 100 kHz to 1 MHz, applying a constant amplitude excitation pulse at each step with a fixed period, without taking any adjustment measures for possible resonance frequency bands; the other group used the method of this invention, constructing a complex impedance spectrum and resonance fingerprint model to identify frequency bands with high resonance matching strength in real time, and adjusting the pulse amplitude and period rhythm according to the amplitude gradient function. The results showed that in the traditional mode, when the frequency reached close to 650 kHz and 840 kHz, the particle aggregation rate increased sharply, resulting in an impurity coating ratio as high as 0.87%. Some clustered particles failed to pass through sieving or magnetic separation to remove impurities, and the purity of the finished product fluctuated by more than 1%. In contrast, the method of this invention activated an energy suppression strategy within the above frequency bands, automatically reducing the pulse amplitude by 45% and extending the response period by 1.5 times, effectively avoiding particle aggregation. The final impurity coating ratio in the sample decreased to 0.21%, and the overall magnetic separation efficiency increased by 23.6%, demonstrating excellent resonance avoidance and separation accuracy.

[0044] [Example 3: Verifying the ability of the resonance recognition mechanism of the present invention to control aggregation behavior] To verify the inducing effect of electromagnetic resonance on the agglomeration behavior of zinc powder particles, and the effectiveness of this invention in identifying and suppressing this effect, two experimental environments were constructed: one environment eliminated the resonance identification and control strategy, applying magnetic field excitation without frequency identification; the other environment employed the method of this invention, performing complex impedance spectrum extraction and constructing a resonance matching intensity model before excitation to identify potentially high-risk frequency bands and adjust excitation parameters. In the reference group, the excitation process repeatedly triggered resonance conditions, generating a large number of particle clusters with diameters greater than 85 μm, accounting for 7.8% of the total number of particles. Many of these clusters contained embedded impurities such as carbon particles and iron oxide, leading to subsequent ineffective separation. In the experimental group applying this invention, the magnetic field excitation frequency was limited to a frequency band with a resonance matching intensity coefficient below 0.4. The results showed a significant decrease in the proportion of particle clusters, to only 1.2%, with an impurity embedding rate of less than 0.1%. This verifies that this invention has the ability to accurately identify resonance risks and effectively control agglomeration, significantly reducing the probability of irreversible failure during magnetic separation.

[0045] [Example 4: Evaluation of the adaptability and stability of the present invention's technology under different particle size raw material conditions] To further verify the versatility and industrial adaptability of this invention, three industrial-grade zinc powder raw materials with significantly different particle size distributions (D50=20μm, 40μm, and 60μm, respectively) were selected for processing, and the traditional magnetic separation method and the method of this invention were compared. When processing fine zinc powder with D50=20μm, the traditional method, due to its large surface area and concentrated resonant frequency, is prone to agglomeration in specific frequency bands, resulting in reduced purity and significantly decreased classification efficiency; the purity fluctuation range reached ±1.7%, and the separation stability was poor. In contrast, this invention, by pre-constructing complex impedance models and inductance-capacitance parameter matrices for each particle size grade, can accurately identify and avoid the resonant ranges of samples with different particle sizes in real time. In all three groups of samples, the particle size control deviation was maintained within ±0.3%, and the final sample purity remained above 99.5%. Experiments show that this invention not only performs excellently in single-particle-size samples but also possesses strong multi-particle-size adaptability and stability control capabilities under complex working conditions.

[0046] This invention constructs an electromagnetic inversion scenario of a three-dimensional particle cloud, combining complex impedance spectroscopy extraction, frequency response scanning, and resonance fingerprint modeling to comprehensively identify the electromagnetic resonance characteristics of zinc powder particle groups under high-frequency alternating magnetic fields. Furthermore, by quantitatively defining the resonance matching intensity coefficient, it achieves accurate assessment of particle agglomeration risk. Based on this, according to the coupling relationship between the resonance matching intensity coefficient and the spatial distribution of the particle group, an amplitude gradient control function is designed, and a frequency-band differentiated pulse intensity modulation strategy is introduced. This enables dynamic control of magnetic drive excitation parameters, effectively suppressing high-risk frequency bands and enhancing the processing efficiency of low-risk frequency bands throughout the process. Through the above scheme, this invention not only successfully avoids the nonlinear agglomeration of zinc powder particles induced by electromagnetic resonance in traditional magnetic separation processes, significantly improving the consistency of zinc powder particle size classification and the thoroughness of impurity separation, but also greatly enhances the response stability and fine control capability of the entire process. This achieves continuous and stable preparation of high-purity zinc powder, demonstrating promising industrial application prospects and promotional value.

[0047] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A graded magnetic separation purification method for separating impurities in zinc powder, characterized in that, Includes the following steps: S101. Construct an electromagnetic inversion scenario of a three-dimensional particle cloud and derive the complex impedance spectrum within the microscale region. S102. Based on the obtained complex impedance spectrum, extract the inductive reactance component and the capacitive reactance component, and construct an inductor-capacitor parameter matrix that reflects the electromagnetic properties of local zinc powder particles. S103. Using the inductor-capacitor parameter matrix as input, perform electromagnetic excitation scanning over a continuous wide frequency range, apply alternating magnetic fields at multiple frequency points, and simultaneously record the unit energy absorption power density at the corresponding frequency to form an initial resonant response profile. S104. The initial resonance response profile is fitted with an improved Lorentz function to extract the full width at half maximum (FWHM) and peak energy absorption power of the main resonance peak. Attenuation correction is performed in combination with environmental parameters to generate a resonance response fingerprint model. S105. Calculate the resonance matching strength coefficient based on the resonance response fingerprint model and the target magnetic field frequency to characterize the resonance and agglomeration risk of zinc powder particles at this frequency. S106. Based on the coupling relationship between the resonant matching strength coefficient and the spatial distribution of the particle swarm, a pulse intensity modulation strategy with frequency band differentiation is implemented, which performs pulse suppression control on the high matching strength frequency band and pulse enhancement control on the low matching strength frequency band.

2. The graded magnetic separation purification method for separating zinc powder impurities according to claim 1, characterized in that, Step S101 includes: A controllable three-dimensional magnetic field excitation region was constructed, the frequency and amplitude range of the alternating magnetic field were set, and a magnetic field vector acquisition device and a potential measurement electrode were deployed in the zinc powder particle distribution area. The zinc powder sample was uniformly dispersed in the area, and the magnetic field vector phase distribution and the instantaneous potential difference of the particles were collected simultaneously under magnetic field excitation. Based on the collected data, a spatiotemporal joint inversion algorithm was used for analysis to obtain the complex impedance spectrum representing the electromagnetic behavior of zinc powder particle groups in the microscale region.

3. The graded magnetic separation purification method for separating zinc powder impurities according to claim 1, characterized in that, Step S102 includes: The complex impedance spectrum is reconstructed in the frequency domain to construct a complex function array containing changes in the real and imaginary parts at multiple frequency points, and then subjected to multi-scale decomposition and sliding window processing. For the impedance data of each frequency sub-interval, an adaptive recursive algorithm is used for dynamic decomposition to extract the inductive reactance component and the capacitive reactance component. Based on the extraction results, an inductance-capacitance parameter matrix for each spatial region is constructed to reflect the equivalent electromagnetic response characteristics of different particle swarms.

4. The graded magnetic separation purification method for separating impurities in zinc powder according to claim 1, characterized in that, Step S103 includes: Using the inductor-capacitor parameter matrix as input, a frequency response prediction model is established to determine the electromagnetic excitation frequency range and step accuracy. An alternating magnetic field of constant amplitude is applied at the set frequency point, and the unit energy absorption power density absorbed by the zinc powder particle group at each frequency is recorded simultaneously. The duration of excitation at each frequency is greater than twice the corresponding particle swarm response delay. The power is normalized using an integral energy measurement method. After removing background electromagnetic interference, the power density data are arranged in ascending order of frequency to form the initial resonant response profile curve.

5. The graded magnetic separation purification method for separating zinc powder impurities according to claim 1, characterized in that, Step S104 includes: The initial resonance response profile is smoothed and denoised to extract the target frequency band of the main resonance peak. For the main resonant frequency band, an improved Lorentz function containing an offset term and a stretching factor is used for nonlinear fitting to extract the center frequency, full width at half maximum (FWHM), and peak power. An environmental attenuation correction factor is introduced based on the air field baseline data to correct the error of the fitting parameters. A resonance response fingerprint model is then constructed based on the corrected parameters to reflect the resonance characteristics and energy response trend of zinc powder particle groups at a specific frequency.

6. The graded magnetic separation purification method for separating zinc powder impurities according to claim 1, characterized in that, Step S105 includes: The main resonance peak parameters are extracted from the resonance response fingerprint model, and the target magnetic field driving frequency is set for matching analysis. Calculate the percentage of energy absorbed per unit frequency to characterize the intensity of energy resonance excitation; Extract the potential response signals at multiple locations at the target frequency and calculate the phase synchronization rate to reflect the response consistency of the particle swarm. Multiplying the energy absorption power ratio by the phase synchronization rate yields the resonance matching strength coefficient.

7. The graded magnetic separation purification method for separating impurities in zinc powder according to claim 1, characterized in that, Step S106 includes: A model for the coupling relationship between the resonance matching strength coefficient and the spatial distribution of zinc powder particle swarm was established, and a resonance risk heatmap was constructed. Design an amplitude gradient control function with the resonant matching strength coefficient as input and the magnetic field amplitude as output to achieve continuous dynamic adjustment of the excitation amplitude at the frequency point; The frequency band risk level is divided according to the resonance matching strength coefficient, and pulse intensity suppression or enhancement strategies are implemented respectively. The pulse amplitude, frequency and interval are adjusted, and the amplitude gradient control and pulse modulation strategy are combined and applied to the whole process to achieve spatial frequency coordinated control of the zinc powder magnetic drive excitation parameters.

Images