Multi-stage magnetic separation equipment and magnetic separation method for high-purity iron powder

Abstract

The present invention provides a multi-stage magnetic separation device for high-purity iron powder and a magnetic separation method thereof, which relates to the field of magnetic separation technology. The device includes: a component characteristic calling module for calling multimodal iron powder component characteristics; a control parameter mapping module for performing multi-stage magnetic separation control parameter mapping to generate N-stage benchmark magnetic separation control parameters; a magnetic separation processing module for executing N-stage magnetic separation processing of the iron powder to be purified. The magnetic separation processing module includes: a closed-loop control unit for executing single-level magnetic separation parameter dynamic closed-loop control according to the results of synchronous laser-induced breakdown spectroscopy analysis; a joint adjustment optimization unit for executing joint adjustment optimization of cross-level Gaussian setting benchmarks; and an iterative execution unit. The present invention solves the technical problem that the magnetic separation process in the prior art is usually only applicable to a single magnetic separation stage and cannot be effectively optimized in multiple magnetic separation stages, resulting in the magnetic separation process being unable to fully optimize the processing conditions of each stage, thereby affecting the final iron powder purity.

Description

Technical Field

[0001] The present invention relates to the technical field of magnetic separation, and in particular to a multi-stage magnetic separation device for high-purity iron powder and a magnetic separation method thereof. Background Art

[0002] High-purity iron powder is widely used in fields such as powder metallurgy, magnetic materials, and high-end electronic devices. With the development of industrial technology, the demand for high-purity iron powder continues to increase, especially in high-end manufacturing and high-tech industries, where the purity requirements for iron powder are becoming increasingly stringent. Therefore, how to effectively remove impurities in iron powder and improve its purity has become a key technical challenge in the current iron powder production process.

[0003] However, most existing technologies rely on fixed parameters such as preset magnetic field strength and magnetic separation time to process iron powder. Such fixed control parameters make the magnetic separation process unable to flexibly cope with iron powder particles of different particle sizes and magnetic strengths, which in turn leads to unsatisfactory magnetic separation effects, which may cause incomplete impurity removal or excessive separation, affecting the purity of the final product; moreover, the magnetic separation process of the existing technology is usually only applicable to a single magnetic separation stage and cannot be effectively optimized in multiple magnetic separation stages, resulting in the magnetic separation process being unable to fully optimize the processing conditions of each stage, unable to maximize the removal of impurities, and affecting the purity of the final iron powder. Summary of the Invention

[0004] The present application provides a multi-stage magnetic separation equipment and a magnetic separation method for high-purity iron powder, aiming to solve the technical problem that the magnetic separation process in the prior art is usually only applicable to a single magnetic separation stage and cannot be effectively optimized in multiple magnetic separation stages, resulting in the magnetic separation process being unable to fully optimize the processing conditions of each stage, thereby affecting the final iron powder purity.

[0005] The first aspect disclosed in the present application provides a multi-stage magnetic separation device for high-purity iron powder, the device comprising: a component characteristic calling module for calling multimodal iron powder component characteristics across devices; a control parameter mapping module for performing multi-stage magnetic separation control parameter mapping based on the multimodal iron powder component characteristics to generate N-level benchmark magnetic separation control parameters, wherein each level of benchmark magnetic separation control parameters consists of a Gaussian setting benchmark, a Gaussian adjustable interval, and an execution time benchmark; a magnetic separation processing module for the magnetic separation hardware system to use N execution time benchmarks as magnetic separation time limits and perform N-level magnetic separation processing processes on the iron powder to be purified based on the N Gaussian setting benchmarks, the magnetic separation processing module comprising: a closed-loop control unit for performing dynamic closed-loop control of single-level magnetic separation parameters within N Gaussian adjustable intervals based on the synchronous laser induced breakdown spectroscopy analysis results of the LIBS probe; a joint optimization unit for performing joint optimization of cross-level Gaussian setting benchmarks based on the impurity distribution heat map at the end of the N execution times; and an iterative execution unit for iteratively executing until the N-level iron powder magnetic separation is completed to produce iron powder of target purity.

[0006] The second aspect disclosed in the present application provides a multi-stage magnetic separation method for high-purity iron powder, which is implemented by the above-mentioned multi-stage magnetic separation equipment for high-purity iron powder, and the method includes: calling multi-modal iron powder component characteristics across devices; performing multi-stage magnetic separation control parameter mapping based on the multi-modal iron powder component characteristics to generate N-level benchmark magnetic separation control parameters, wherein each level of benchmark magnetic separation control parameters is composed of a Gaussian setting benchmark, a Gaussian adjustable interval and an execution time benchmark; the magnetic separation hardware system uses N execution time benchmarks as the magnetic separation time limit, and executes N-level magnetic separation processing of the iron powder to be purified based on N Gaussian setting benchmarks: step a: according to the synchronous laser induced breakdown spectroscopy analysis results of the LIBS probe, perform single-level magnetic separation parameter dynamic closed-loop control within the N Gaussian adjustable intervals; step b: according to the impurity distribution heat map at the end of the N execution times, perform cross-level Gaussian setting benchmark joint optimization; iterative execution until the N-level iron powder magnetic separation is completed to produce target purity iron powder.

[0007] One or more technical solutions provided in this application have at least the following beneficial effects:

[0008] By calling the multimodal iron powder component characteristics across devices, the physical and chemical properties of the iron powder can be fully understood, which provides accurate data support for the subsequent magnetic separation process and ensures that the magnetic separation parameters can be adjusted in a targeted manner during the magnetic separation process; according to the multimodal iron powder component characteristics of the iron powder, N-level benchmark magnetic separation control parameters are generated through multi-level magnetic separation control parameter mapping. The control parameters of each level include Gaussian setting benchmark, Gaussian adjustable interval and execution time benchmark, ensuring that the magnetic separation process can accurately process iron powders of different particle sizes and magnetic strengths, so that the iron powder can be effectively graded to improve the purity; by using N execution time benchmarks as the magnetic separation time limit, the processing time of each level of magnetic separation is ensured to be accurately controlled to avoid over-processing or under-processing. Within the time limit of each magnetic separation level, the magnetic field strength and processing time are adjusted according to the results of the previous stage, so as to gradually remove the iron powder. Impurities; During the magnetic separation process, the concentration of impurities on the surface of the iron powder is detected in real time through the LIBS probe synchronous laser induced breakdown spectroscopy analysis, and the magnetic separation parameters can be dynamically adjusted according to the real-time data to form a closed-loop control. This dynamic closed-loop control can respond to changes in impurity concentration in real time to ensure that each magnetic separation level can achieve the best removal effect; At the end of each magnetic separation stage, an impurity distribution heat map is generated. Through the impurity distribution heat map, the control parameters can be optimized in the subsequent magnetic separation stage. Through the joint optimization of the Gaussian setting benchmark, the operation of each magnetic separation level is ensured to be coordinated and consistent, further improving the efficiency of impurity removal; Through the iterative execution of the magnetic separation process, the control strategy is continuously optimized until the required target purity is achieved. The results of each stage will affect the control strategy of the subsequent stage, thereby ensuring that the entire magnetic separation process can efficiently remove impurities and ultimately produce high-purity iron powder.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic structural diagram of a multi-stage magnetic separation device for high-purity iron powder provided in an embodiment of the present application.

[0011] Figure 2 A schematic flow chart of a multi-stage magnetic separation method for high-purity iron powder provided in an embodiment of the present application.

[0012] Explanation of the reference numerals: component characteristic calling module 10 , control parameter mapping module 20 , magnetic separation processing module 30 , closed-loop control unit 31 , joint debugging optimization unit 32 , iterative execution unit 33 . DETAILED DESCRIPTION

[0013] The embodiments of the present application provide a multi-stage magnetic separation equipment and a magnetic separation method for high-purity iron powder, which solves the technical problem that the magnetic separation process in the prior art is usually only applicable to a single magnetic separation stage and cannot be effectively optimized in multiple magnetic separation stages, resulting in the magnetic separation process being unable to fully optimize the processing conditions of each stage, thereby affecting the final iron powder purity.

[0014] After introducing the basic principles of this application, various non-limiting embodiments of this application will be specifically described below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only used to explain this application and are not used to limit this application.

[0015] Example 1, as Figure 1 As shown, an embodiment of the present application provides a multi-stage magnetic separation device for high-purity iron powder, the device comprising:

[0016] The component characteristic calling module 10 is used to call multimodal iron powder component characteristics across devices.

[0017] The control parameter mapping module 20 is used to perform multi-level magnetic separation control parameter mapping based on the multi-modal iron powder component characteristics to generate N-level reference magnetic separation control parameters, wherein each level of reference magnetic separation control parameters consists of a Gaussian setting reference, a Gaussian adjustable interval and an execution time reference.

[0018] The magnetic separation processing module 30 is used for the magnetic separation hardware system to perform N-level magnetic separation processing of the iron powder to be purified based on N execution time bases as the magnetic separation time limit and N Gauss setting bases. The magnetic separation processing module includes:

[0019] The closed-loop control unit 31 is used to perform dynamic closed-loop control of single-layer magnetic separation parameters within N Gaussian adjustable intervals according to the synchronous laser-induced breakdown spectroscopy analysis results of the LIBS probe.

[0020] The joint debugging optimization unit 32 is used to perform joint debugging optimization with a cross-level Gaussian setting benchmark based on the impurity distribution heat map at the end of N execution times.

[0021] The iterative execution unit 33 is used for iterative execution until N levels of iron powder magnetic separation are completed to produce iron powder with target purity.

[0022] Furthermore, the joint debugging and optimization unit also includes:

[0023] The pre-selection magnetic separation processing channel is used to operate the magnetic separation hardware system with the first Gauss setting benchmark of the first-level benchmark magnetic separation control parameter, and then start the feeder to transport the iron powder to be purified into the magnetic separation hardware system to perform pre-selection magnetic separation processing.

[0024] The spectral analysis channel is used to synchronously run the LIBS probe to perform laser-induced breakdown spectroscopy analysis on the conveyed iron powder during the pre-selection magnetic separation process, and output a real-time surface impurity concentration sequence.

[0025] A dynamic parameter update channel is used to perform dynamic parameter update of the first Gaussian setting benchmark based on the multivariate impurity fluctuation characteristics of the real-time surface impurity concentration sequence, with the first Gaussian adjustable interval as a constraint, until the first execution time benchmark is reached, and generate a first impurity distribution heat map.

[0026] Furthermore, the iterative execution unit further includes:

[0027] The cross-level pre-correction channel is used to perform cross-level pre-correction on the second Gaussian setting benchmark according to the first impurity distribution heat map to obtain the second Gaussian joint tuning parameters.

[0028] A parameter update channel is used to execute parameter update of the second Gaussian joint tuning parameters according to the surface impurity concentration data synchronously detected by the LIBS probe during the secondary magnetic separation process of the magnetic separation hardware system using the second Gaussian joint tuning parameters, until the magnetic separation time reaches the second execution time benchmark, and output a second impurity distribution heat map.

[0029] The Gaussian benchmark pre-tuning channel is used to iteratively perform dynamic adjustment of magnetic separation parameters in a single-layer magnetic separation process based on the surface impurity concentration, and iteratively perform cross-layer Gaussian benchmark pre-tuning based on the layer impurity thermal distribution until the N-level benchmark magnetic separation control parameters are executed and the target purity iron powder is produced.

[0030] Furthermore, the dynamic parameter update channel includes:

[0031] The Gaussian adjustment action library construction node is used to interactively obtain multiple sets of Gaussian adjustment actions for various iron powder impurities under multiple sets of impurity fluctuation scenarios to build a Gaussian adjustment action library.

[0032] The parallel matching node is used to input the multivariate impurity fluctuation characteristics as multivariate search conditions into the Gaussian adjustment action library to parallel match the multivariate Gaussian adjustment actions of the multivariate impurity fluctuation scenario.

[0033] The union solving node is used to perform a union solving on the multivariate Gaussian adjustment action to locate the first real-time adjustment action.

[0034] A parameter update node is used to use the first real-time adjustment action to perform parameter update of the first Gaussian setting benchmark.

[0035] The data acquisition node is used to drive the LIBS probe to collect instantaneous impurity concentration data and synthesize the first impurity distribution thermodynamic map when the first Gaussian setting benchmark is dynamically adjusted and updated and the first execution time benchmark is reached.

[0036] Furthermore, the cross-stage pre-correction channel includes:

[0037] A thermal feature detection network construction node is used to pre-construct a thermal feature detection network, wherein the thermal feature detection network includes M thermal map feature recognition channels.

[0038] The feature parallel recognition node is used to load the first impurity distribution thermodynamic map into the thermal feature detection network, perform feature parallel recognition through the M thermodynamic map feature recognition channels, and output M real-time thermodynamic map quantitative features.

[0039] The cross-level pre-correction node is used to perform cross-level pre-correction on the second Gaussian setting benchmark after pre-correction action matching according to the M real-time heat map quantitative features to obtain the second Gaussian joint adjustment parameters.

[0040] Furthermore, the cross-level pre-correction node includes:

[0041] The standard pre-correction action setting sub-node is used to locally set M groups of standard pre-correction actions for M groups of heat map quantitative feature thresholds.

[0042] The action extraction sub-node is used to extract M standard pre-correction actions from the M groups of standard pre-correction actions by using the intersection relationship of the M real-time heat map quantitative features in the M groups of heat map quantitative feature thresholds.

[0043] The conflict compensation sub-node is configured to perform conflict compensation on the M standard pre-correction actions to obtain a first pre-correction action.

[0044] The cross-level pre-correction sub-node is used to perform cross-level pre-correction on the second Gaussian setting benchmark by using the first pre-correction action to obtain the second Gaussian joint tuning parameter.

[0045] Furthermore, the dynamic parameter update channel is used to: input the spatial mapping result of the spatial coordinate mapping of the instantaneous impurity concentration data into a predefined impurity concentration-color level mapping table for gradient coloring synthesis to obtain the first impurity distribution heat map.

[0046] Furthermore, the multimodal iron powder composition characteristics include a particle size distribution curve, a saturation magnetization distribution, a coercive force distribution, and an initial impurity spectrum.

[0047] Furthermore, the control parameter mapping module includes:

[0048] The first matching unit is used to analyze the particle size distribution curve, locate the median particle size position and the median particle size distribution width ratio, and then obtain the first multi-stage initialization magnetic separation control parameter by matching the residence time spectrum mapping table according to the median particle size position and the median particle size distribution width ratio.

[0049] The second matching unit is used to obtain the second multi-level initialization magnetic separation control parameter by matching the field intensity response spectrum mapping table according to the calculated central tendency value of the saturation magnetization intensity distribution and the high value area ratio of the coercive force distribution.

[0050] The third matching unit is configured to extract an impurity concentration group corresponding to a core impurity group from the initial impurity spectrum, and then obtain a third multi-stage initialization magnetic separation control parameter by matching the element removal response spectrum mapping table.

[0051] The conflict parameter arbitration unit is used to perform conflict parameter arbitration on the first multi-level initialization magnetic separation control parameter, the third multi-level initialization magnetic separation control parameter and the third multi-level initialization magnetic separation control parameter, and output the N-level reference magnetic separation control parameter.

[0052] Through the subsequent detailed description of a multi-stage magnetic separation method for high-purity iron powder in this specification, those skilled in the art can clearly understand the multi-stage magnetic separation equipment for high-purity iron powder in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple. For relevant details, please refer to the method part.

[0053] Example 2, based on the same inventive concept as the multi-stage magnetic separation device for high-purity iron powder in the above embodiment, Figure 2 As shown, an embodiment of the present application provides a multi-stage magnetic separation method for high-purity iron powder, the method comprising:

[0054] Call multimodal iron powder component characteristics across devices;

[0055] Multimodal iron powder component characteristics refer to the different physical and chemical properties of iron powder samples obtained through multiple testing devices, including particle size distribution curve, saturation magnetization intensity distribution, coercive force distribution and initial impurity spectrum. For example, the particle size distribution curve of iron powder is obtained by a particle size analyzer, the saturation magnetization intensity distribution of iron powder is obtained by a magnetization intensity testing device, and the coercive force distribution of iron powder is obtained by a coercive force testing device. Cross-device calling means that data between different testing devices can be shared and combined with each other, thereby generating a complete, multi-angle multimodal iron powder component characteristic.

[0056] Multi-level magnetic separation control parameter mapping is performed according to the multi-modal iron powder component characteristics to generate N-level reference magnetic separation control parameters, wherein each level of reference magnetic separation control parameters consists of a Gaussian setting reference, a Gaussian adjustable interval and an execution time reference.

[0057] Multi-stage magnetic separation control parameter mapping is performed based on the multimodal iron powder component characteristics, aiming to generate precise control parameters for each magnetic separation level to optimize the magnetic separation process. For example, the magnetic separation parameters for different particle size segments are determined based on the particle size distribution curve of the iron powder. For example, for finer particles, a higher magnetic field strength and a shorter execution time are required to avoid over-purification of particles due to excessive magnetic field strength. For saturation magnetization intensity distribution, iron powder particles with high magnetization intensity will be affected by a stronger magnetic force, so the magnetic separation parameters for these particles will be different, and the corresponding magnetic field strength and execution time are calculated based on the saturation magnetization intensity distribution. For coercive force distribution, particles with larger coercive force are more difficult to remove, so longer magnetic separation time or a stronger magnetic field is required, and the magnetic separation parameters are adjusted according to the coercive force distribution. For the initial impurity spectrum, the initial impurity spectrum provides the concentration distribution of impurities. By matching it with the core impurity group, the impurity removal target is obtained, and the control parameters of each level of magnetic separation are adjusted accordingly.

[0058] Based on the above mapping, N-level benchmark magnetic separation control parameters are generated. The divided magnetic separation levels can include pre-selection level, fine selection level, and reverse selection level. Among them, the Gaussian setting benchmark is the set value of the initial condition, which serves as the starting standard for the magnetic separation operation; the Gaussian adjustable range refers to the magnetic field intensity range that can be adjusted in each magnetic separation level, representing the adjustment space in the magnetic separation process; the execution time benchmark is the time requirement for each level of magnetic separation operation to ensure that the magnetic separation process is completed within the specified time.

[0059] This parameter mapping approach ensures that the magnetic separation process can be tailored to each stage, resulting in a higher purity of the final iron powder.

[0060] The magnetic separation hardware system uses N execution time benchmarks as the magnetic separation time limit. When executing N-level magnetic separation processes of the iron powder to be purified according to N Gauss setting benchmarks:

[0061] Step a: Based on the results of the synchronous laser-induced breakdown spectroscopy analysis of the LIBS probe, dynamic closed-loop control of the single-layer magnetic separation parameters is performed within N Gaussian adjustable intervals.

[0062] Step b: Based on the impurity distribution heat map at the end of N execution times, perform cross-level Gaussian setting benchmark joint optimization.

[0063] The process is iterated until N levels of iron powder magnetic separation are completed, producing iron powder of the target purity.

[0064] The execution time benchmark is the time requirement for each level of magnetic separation operation. Each level of magnetic separation operation has a predetermined duration, which controls the duration of the magnetic separation process. The N execution time benchmarks of the N-level magnetic separation process are used as the magnetic separation time limit to ensure that each level of magnetic separation processing is completed within the specified time. The Gaussian setting benchmark is the set value of the initial condition, which is usually based on the Gaussian distribution. During the magnetic separation process, the N Gaussian setting benchmarks of the N-level magnetic separation process are used to determine the control strategy of each stage. The magnetic separation control of each level will be optimized according to the Gaussian setting benchmark, so that each level of magnetic separation operation can be maximized according to the current iron powder characteristics. The N-level magnetic separation process of the iron powder to be purified includes:

[0065] LIBS (Laser-Induced Breakdown Spectroscopy) is a technique commonly used to analyze the composition of materials. It uses lasers to excite the surface of a material, generating plasma and obtaining elemental composition information through spectral analysis. This process is particularly useful for analyzing the concentration of surface impurities in real time. The LIBS probe measures the concentration of impurities in the iron powder to be purified in real time, generating a laser-induced spectrum to determine the distribution of different impurity elements in the iron powder.

[0066] The Gaussian adjustable range refers to the range of magnetic field strength that can be adjusted in each magnetic separation level, representing the adjustment space in the magnetic separation process. Within these ranges, the magnetic separation hardware system is dynamically adjusted according to the LIBS analysis results. Among them, the dynamic closed-loop control of single-level magnetic separation parameters refers to the real-time adjustment of the parameters of the current magnetic separation level based on the impurity concentration data obtained by the LIBS probe. By real-time monitoring of the impurity concentration and feeding back to the control system, the magnetic field strength, execution time and other magnetic separation parameters can be dynamically adjusted to optimize the magnetic separation effect. Through dynamic closed-loop control, this process can respond to changes in impurity concentration in iron powder in real time, accurately adjust the magnetic separation parameters, ensure that each level of magnetic separation achieves the optimal effect, avoid unnecessary excessive or insufficient magnetic separation, and thus improve the purity of the iron powder.

[0067] At the end of each magnetic separation level, impurity distribution heat maps are generated. These impurity distribution heat maps show the distribution of impurities at different execution time points, reflecting the impurity removal effect after each level of magnetic separation operation, and helping to identify which magnetic separation stages still need to be optimized.

[0068] Cross-level cascade optimization refers to the optimization and adjustment of the Gaussian set benchmark at the end of the execution time of all magnetic separation levels, combined with the impurity distribution heat map of each level. Specifically, during the processing of each magnetic separation level, the magnetic separation parameters of the next level are adjusted according to the information feedback of the impurity distribution heat map. The optimization process involves adjusting the Gaussian set benchmark according to the information of the impurity distribution heat map to ensure that each level of magnetic separation can effectively remove impurities, and that the processing parameters of subsequent levels can be coordinated with the operations of the previous levels, thereby further improving the purity. Cross-level optimization ensures the coordination between the magnetic separation operations at each level, enabling the system to accurately adjust the parameters at each stage to avoid parameter conflicts or mismatches between different levels. By real-time feedback of impurity distribution and adjustment of the set benchmark, the magnetic separation process can be gradually optimized to ensure that the purity of the final iron powder meets the target requirements.

[0069] The magnetic separation process is continuously executed, adjusting the parameters of subsequent stages based on the results of the previous stage. This iterative process ensures that each level of magnetic separation is optimized based on real-time feedback data. The output data of each stage is fed back into the system to further optimize the magnetic separation parameters. By continuously adjusting the magnetic separation parameters, impurities can be removed to the maximum extent during the multi-stage magnetic separation process, resulting in increasingly higher iron powder purity. After N stages of magnetic separation, the final iron powder output will meet the required purity standard and obtain the target purity iron powder.

[0070] Furthermore, the method further comprises:

[0071] After operating the magnetic separation hardware system using the first Gaussian setting benchmark of the first-level benchmark magnetic separation control parameters, the feeder is started to convey the iron powder to be purified into the magnetic separation hardware system for pre-selection magnetic separation processing; during the pre-selection magnetic separation processing, the LIBS probe is synchronously operated to perform laser induced breakdown spectroscopy analysis on the conveyed iron powder, and a real-time surface impurity concentration sequence is output; based on the multivariate impurity fluctuation characteristics of the real-time surface impurity concentration sequence, the dynamic parameter adjustment update of the first Gaussian setting benchmark is executed with the first Gaussian adjustable interval as a constraint until the first execution time benchmark is reached and the first impurity distribution heat map is generated.

[0072] Pre-selection magnetic separation is the first step in the magnetic separation process. Its purpose is to remove relatively simple and easy-to-separate impurities in the iron powder. This stage usually deals with coarse-grained impurities in the iron powder and those components that can be effectively removed by simple magnetic separation. The first-level benchmark magnetic separation control parameters provide preliminary magnetic separation settings for the pre-selection magnetic separation process. The first Gauss setting benchmark of the first-level benchmark magnetic separation control parameters is used to run the magnetic separation hardware system. In the magnetic separation hardware system, the feeder is responsible for conveying the iron powder to be purified into the magnetic separation hardware system. The feeder's conveying volume and speed need to be adjusted according to the processing capacity of the magnetic separation equipment and the characteristics of the iron powder to ensure that the iron powder can enter the magnetic separation system evenly.

[0073] During the pre-selection magnetic separation process, the LIBS probe detects the concentration of impurities in the iron powder to be purified in real time, generates a laser-induced spectrum, and thus obtains the distribution of different impurity elements in the iron powder. Synchronous operation means that while the pre-selection magnetic separation process is being carried out, the LIBS probe will continue to analyze and obtain real-time impurity concentration data on the surface of the iron powder. This process is carried out in parallel with the magnetic separation process to ensure that the changes in impurities at each stage can be accurately captured. The real-time surface impurity concentration series is the data collected by the LIBS probe at each moment, indicating the concentration changes of impurities on the iron powder surface. This data series provides the concentration changes of different impurity elements in the iron powder surface, helping to analyze which impurities have been removed during the magnetic separation process and which still exist.

[0074] The multivariate impurity fluctuation characteristics refer to the time-varying patterns of the concentrations of different types of impurities in the surface layer of iron powder. Since iron powder contains a variety of impurity elements, the fluctuation characteristics of the impurity concentration are often multivariate. Different impurities exhibit different fluctuation characteristics in different time periods. For example, some impurities may be removed quickly in a short period of time, while other impurities may be removed more slowly.

[0075] The Gaussian adjustable range refers to the range of magnetic field strength that can be adjusted during the magnetic separation process. In this step, the first Gaussian adjustable range is used as an adjustment range to limit the dynamic adjustment of the magnetic separation parameters. This constraint ensures that the magnetic separation parameters will not deviate from the set range during the dynamic adjustment process, thereby avoiding over-adjustment of the system. With the first Gaussian adjustable range as a constraint, the first Gaussian setting benchmark is dynamically updated. The dynamic parameter update is based on the impurity concentration data to ensure that the magnetic separation process can effectively remove different types of impurities. The goal is to continuously optimize the magnetic separation process based on real-time data and improve the impurity removal efficiency. At each time point, the magnetic separation parameters are adjusted according to the current impurity concentration changes to ensure that each level of the magnetic separation process can adapt to the different characteristics of the impurities.

[0076] The first execution time benchmark is the maximum duration of the pre-selection magnetic separation. When the time in the pre-selection magnetic separation process reaches the first execution time benchmark, the adjustment process stops and prepares to enter the next stage of processing. At the end of the first execution time benchmark, the first impurity distribution heat map is generated according to the impurity removal in the magnetic separation process. The heat map is a visualization tool that shows the changes in impurity distribution at various stages of the magnetic separation process. The first impurity distribution heat map presents the distribution of impurity concentration in the form of color levels, where high-concentration areas are displayed in different colors to facilitate analysis and optimization of magnetic separation processing. The first impurity distribution heat map is used to further analyze which areas have poor impurity removal effects, so as to make corresponding adjustments in subsequent magnetic separation processing.

[0077] Furthermore, the method further comprises:

[0078] According to the first impurity distribution thermodynamic map, the second Gaussian setting benchmark is pre-corrected across levels to obtain the second Gaussian joint tuning parameters; when the magnetic separation hardware system is operated using the second Gaussian joint tuning parameters to perform secondary magnetic separation processing, the second Gaussian joint tuning parameters are updated according to the surface impurity concentration data synchronously detected by the LIBS probe until the magnetic separation time reaches the second execution time benchmark, and the second impurity distribution thermodynamic map is output; based on the surface impurity concentration, the magnetic separation parameters of the single-layer magnetic separation process are dynamically adjusted, and based on the layer impurity thermal distribution, cross-layer Gaussian benchmark pre-tuning is iteratively performed until the N-level benchmark magnetic separation control parameters are executed and the target purity iron powder is produced.

[0079] Cross-stage pre-correction adjusts the control parameters of the secondary magnetic separation process based on the results of the first-stage magnetic separation (i.e., the first impurity distribution heat map). By analyzing the impurity distribution characteristics in the first heat map, the needs of the second-stage magnetic separation process can be inferred and related control parameters such as magnetic field strength and execution time can be adjusted. The second Gaussian setting benchmark is the control benchmark for the secondary magnetic separation process (such as the fine selection stage). Based on the data in the first heat map, the second Gaussian setting benchmark is optimized to ensure that residual impurities can be better removed during the secondary magnetic separation process. The second Gaussian joint adjustment parameters are the magnetic separation control parameters obtained through cross-stage correction, ensuring that the second-stage magnetic separation process can adapt to the results of the first-stage magnetic separation.

[0080] After obtaining the second Gaussian tuning parameters, these optimized parameters are used to run the magnetic separation hardware system for secondary magnetic separation. This process, performed at the secondary magnetic separation level, further removes remaining impurities. A LIBS probe operates synchronously during the secondary magnetic separation process, providing real-time monitoring of surface impurity concentrations on the iron powder. This data provides insights into the efficiency of impurity removal at this stage and any impurity components that remain problematic.

[0081] According to the surface impurity concentration data collected in real time, the parameter update of the second Gaussian joint tuning is executed. This means that according to the current impurity removal situation, the magnetic separation parameters, such as magnetic field strength, execution time, etc., are adjusted in real time to improve the magnetic separation effect. The second execution time benchmark is the set secondary magnetic separation processing time. The duration of the secondary magnetic separation process is controlled according to this benchmark. When the secondary magnetic separation processing time reaches the second execution time benchmark, the secondary magnetic separation process automatically stops and prepares to enter the next stage. At the end of the secondary magnetic separation process, a second impurity distribution heat map is generated. The heat map shows the distribution of impurities in each area during the secondary magnetic separation process, which provides intuitive visual data for subsequent analysis.

[0082] The surface impurity concentration is the concentration data of impurities on the surface of iron powder obtained by real-time monitoring of the LIBS probe. During the single-level magnetic separation process, the magnetic separation parameters, including magnetic field strength and execution time, are dynamically adjusted according to the surface impurity concentration. This dynamic adjustment is achieved through real-time feedback, ensuring that each level of magnetic separation process can maximize impurity removal and maintain process efficiency.

[0083] The hierarchical impurity thermal distribution is used to represent the distribution of impurities in different magnetic separation levels. The heat map can intuitively show the effect of impurity removal in each magnetic separation stage and help determine which areas have poor impurity removal effects. The cross-level Gaussian benchmark pre-tuning is based on the hierarchical impurity thermal distribution generated after each level of magnetic separation processing, and iteratively adjusts the Gaussian benchmark and other magnetic separation control parameters. This means that not only the magnetic separation parameters of the current level are adjusted, but also the magnetic separation parameters of the subsequent stages are adjusted according to the impurity distribution of the previous stage to ensure the coordination between the magnetic separation operations at all levels. This process optimizes the various parameters in the magnetic separation process step by step through iterative optimization to ensure the maximum impurity removal effect.

[0084] The N-level benchmark magnetic separation control parameters are continuously iterated until all levels of magnetic separation processing are completed. Each stage of magnetic separation is adjusted according to the results of the previous stage to ensure that the effect and purity of each stage of magnetic separation are gradually improved. By iteratively executing all levels of magnetic separation processing and dynamically adjusting the magnetic separation parameters according to the results of each level, the target purity iron powder is finally obtained, thereby improving the overall effect of magnetic separation.

[0085] Furthermore, based on the multivariate impurity fluctuation characteristics of the real-time surface impurity concentration sequence, and with the first Gaussian adjustable interval as a constraint, dynamic parameter adjustment and updating of the first Gaussian setting benchmark are executed until a first execution time benchmark is reached, thereby generating a first impurity distribution heat map. The method includes:

[0086] Interactively obtain multiple groups of Gaussian adjustment actions for multiple iron powder impurities under multiple groups of impurity fluctuation scenarios to construct a Gaussian adjustment action library; input the multivariate impurity fluctuation characteristics as multivariate search conditions into the Gaussian adjustment action library and parallelly match the multivariate Gaussian adjustment actions of the multivariate impurity fluctuation scenarios; perform a union solution on the multivariate Gaussian adjustment actions to locate the first real-time adjustment action; use the first real-time adjustment action to execute the parameter update of the first Gaussian setting benchmark; when the first Gaussian setting benchmark is dynamically updated and the end of the first execution time benchmark is reached, drive the LIBS probe to collect instantaneous impurity concentration data and synthesize the first impurity distribution heat map.

[0087] Multiple iron powder impurities refer to different types of impurity components in iron powder. These impurities have different physical and chemical properties. For example, impurities include elements such as silicon, phosphorus, and aluminum. Their removal effect will be affected by parameter changes during the magnetic separation process. Multiple groups of impurity fluctuation scenarios refer to different patterns of concentration changes of impurities during the magnetic separation process. These fluctuation scenarios reflect the removal characteristics of impurities. For example, some impurities may be removed quickly, while others may be removed slowly. The removal performance of each impurity is different under different conditions, so different adjustment strategies are required. The Gaussian adjustment action library generates a set of optional magnetic separation adjustment actions by collecting fluctuation scenarios of multiple impurities. These adjustment actions are based on Gaussian distribution. Each Gaussian adjustment action represents a specific control action, such as increasing the magnetic field strength, extending the magnetic separation time, etc. By analyzing different impurity fluctuation scenarios, selective adjustments can be made under different fluctuation scenarios.

[0088] Multivariate impurity fluctuation characteristics refer to the concentration fluctuation characteristics of multiple impurity elements during the magnetic separation process. The multivariate impurity fluctuation characteristics are input into the Gaussian adjustment action library as retrieval conditions. Through the parallel matching mechanism, the fluctuation characteristics of multiple impurities are considered at the same time to query the corresponding Gaussian adjustment actions. The multivariate Gaussian adjustment action is a set of Gaussian distribution control parameter adjustment strategies generated for multiple impurity fluctuation scenarios. These actions include not only increasing or decreasing the magnetic field intensity, but also adjusting the time, changing the sorting intensity, etc., to ensure that each impurity can achieve the optimal removal effect according to its fluctuation characteristics.

[0089] After matching the multivariate Gaussian adjustment actions, the resulting adjustment actions are then unioned. This involves combining multiple Gaussian adjustment actions to find the most suitable adjustment strategy for the current magnetic separation process. During the union, the maximum value of the magnetization action is used as the baseline offset, the minimum value of the magnetization action is used as the protection threshold, and the median of the adjustable range is used for the conflicting action. Specifically, the magnetization action increases the magnetic field strength. When the magnetic field strength needs to be increased, the maximum value is selected as the offset to ensure that the magnetic force can be effectively increased to remove difficult-to-separate impurities. The baseline offset is the maximum adjustment value within the control range of the Gaussian distribution to enhance the system's adjustment capability. The magnetization action decreases the magnetic field strength. When it is detected that certain impurities have a high removal efficiency or a strong magnetic field may cause over-separation, the minimum value is selected as the protection threshold to avoid excessive magnetic separation caused by an overly strong magnetic field. When a conflicting action occurs in the same scenario and requires mutual adjustment (such as magnetization and demagnetization), the median of the adjustable range is selected as a compromise value to ensure that the system does not over- or under-adjust during the adjustment process. By solving the union of multiple adjustment actions, the first real-time adjustment action is finally determined.

[0090] The first Gaussian setting benchmark is the preliminary control benchmark for the first-level magnetic separation process. The first real-time adjustment action is used to update the parameters of the first Gaussian setting benchmark. Parameter update means adjusting the initial first Gaussian setting benchmark to adapt to the current impurity concentration, iron powder characteristics and magnetic separation effect. Through this adjustment, the magnetic separation process will become more flexible and efficient, and can make precise control according to the characteristics of different impurities to ensure better removal effect.

[0091] During the magnetic separation process, dynamic parameter adjustments and updates continue until the preset first execution time benchmark is reached. When the magnetic separation process reaches the end of the first execution time benchmark, the LIBS probe is started and driven to collect instantaneous impurity concentration data. Based on the instantaneous impurity concentration data, a first impurity distribution heat map is synthesized to display the impurity distribution at various locations in the iron powder during the entire magnetic separation process.

[0092] Furthermore, a cross-level pre-correction is performed on a second Gaussian setting reference according to the first impurity distribution heat map to obtain a second Gaussian joint tuning parameter. The method includes:

[0093] A thermal feature detection network is pre-constructed, wherein the thermal feature detection network includes M thermal map feature recognition channels; after the first impurity distribution thermal map is loaded into the thermal feature detection network, parallel feature recognition is performed through the M thermal map feature recognition channels to output M real-time thermal map quantitative features; after pre-correction action matching is performed based on the M real-time thermal map quantitative features, the second Gaussian setting benchmark is cross-level pre-corrected to obtain the second Gaussian joint tuning parameters.

[0094] Construct a thermal feature detection network. The thermal feature detection network consists of M thermal map feature recognition channels, which are specifically used to analyze and process features in the thermal map. The M thermal map feature recognition channels are multiple channels in the network that process the thermal map in parallel. Each channel is responsible for extracting a certain type of feature in the thermal map, such as the concentration change trend, impurity distribution in the hot spot area, etc. These thermal map feature recognition channels process the input thermal map in different ways to identify different features. Through parallel processing of multiple channels, the network can extract information in the thermal map more comprehensively and accurately.

[0095] The first impurity distribution heat map is input into the pre-built thermal feature detection network, and different features are identified in parallel through M thermal map feature recognition channels. For example, concentration hotspot identification is performed to identify areas with high impurity concentrations to help determine which areas have not effectively removed impurities; impurity removal trend analysis is performed to analyze the removal trends of different areas during the magnetic separation process to help determine which impurities have been removed and which still remain; removal efficiency analysis is performed to analyze the efficiency of impurity removal in each area to help determine whether the magnetic separation operation needs further optimization. Parallel recognition means that these recognition tasks are performed simultaneously rather than sequentially. Through parallel processing, various features in the heat map can be extracted in a shorter time.

[0096] The result of each channel identification is quantified into a characteristic value to form M real-time thermal map quantitative features. These features include impurity concentration, removal efficiency, size of hot spot area, etc. The effect of the magnetic separation process can be judged based on these features, and it can be decided whether the magnetic separation parameters need to be adjusted or further optimized.

[0097] Pre-correction action refers to the preliminary adjustment of the current magnetic separation parameters according to the feature data analyzed in real time during the magnetic separation process. By analyzing the quantitative characteristics of M kinds of thermal maps, it can be determined which magnetic separation parameters need to be adjusted in advance to optimize the effect of the subsequent magnetic separation stage. Pre-correction action matching refers to searching in the preset correction action library according to the quantitative characteristics of M kinds of real-time thermal maps, and selecting the correction action that is most suitable for the current magnetic separation stage.

[0098] Cross-level pre-correction refers to the adjustment of the control parameters of the next magnetic separation level (i.e., the second-level magnetic separation). Although the adjustment in the current stage is based on the data of the current level, the cross-level correction will affect the subsequent magnetic separation levels to ensure that the control strategies between different levels are coordinated with each other. According to the M real-time thermal map quantitative characteristics and matching pre-correction actions, the second Gaussian setting benchmark is cross-level pre-corrected. This correction is to ensure that the second-level magnetic separation can fully utilize the effect of the first-level magnetic separation and further optimize it on this basis. The second Gaussian joint adjustment parameters are the adjusted parameters obtained based on the cross-level pre-correction. These parameters will serve as the control benchmark for the second-level magnetic separation process.

[0099] Furthermore, after performing pre-correction action matching based on the M real-time heat map quantitative features, cross-level pre-correction is performed on the second Gaussian setting benchmark to obtain the second Gaussian joint tuning parameter. The method includes:

[0100] Locally set M groups of standard pre-correction actions for M groups of thermal map quantitative feature thresholds; use the intersection relationship of the M real-time thermal map quantitative features in the M groups of thermal map quantitative feature thresholds to extract M standard pre-correction actions from the M groups of standard pre-correction actions; perform conflict compensation on the M standard pre-correction actions to obtain a first pre-correction action; use the first pre-correction action to perform cross-level pre-correction on the second Gaussian setting benchmark to obtain the second Gaussian joint tuning parameter.

[0101] The M-group thermal map quantitative feature threshold refers to the specific threshold set by the system for classifying and judging the distribution of impurities in the thermal map. These thresholds determine which areas or features need to be adjusted or optimized. These thresholds are usually set based on empirical data, experimental results or optimization goals. For example, a specific impurity concentration value may be considered a sign of insufficient removal, or when the impurity concentration in a certain area is higher than the set threshold, it is determined that the area needs further optimization. The M-group standard pre-correction action is a series of preset correction actions based on the set M-group thermal map quantitative feature threshold. These correction actions are classified based on different impurity concentrations and thermal map features, and provide corresponding magnetic separation adjustment strategies. For example, if the impurity concentration in a certain area exceeds the set threshold, the standard pre-correction action includes increasing the magnetic field strength, extending the magnetic separation time, or adjusting other control parameters.

[0102] Intersection relationships refer to the overlap of different real-time thermographic features within their respective thermographic thresholds. For example, if certain thermographic features indicate excessively high impurity concentrations in a specific area, and they intersect with the concentration threshold, that area is deemed to require special treatment. M standard pre-correction actions are selected from a pre-set set of M standard pre-correction actions based on these intersection relationships. By analyzing the intersection of different thermographic features, the most appropriate corrective action is extracted from a library of pre-set actions, ensuring precise adjustments for different impurity characteristics.

[0103] During the magnetic separation process, conflicts may arise between the M standard pre-correction actions. For example, the removal of some impurities may require an increase in magnetic field strength, while other impurities may overreact due to an overly strong magnetic field, resulting in reduced removal efficiency. In this case, there is a conflict between increasing the magnetic field strength and decreasing it. The purpose of conflict compensation is to resolve these conflicts and ensure that the magnetic separation process does not suffer a decrease in efficiency due to inconsistencies between different corrective actions by merging or adjusting the operation strategy.

[0104] By analyzing the characteristics and effects of M standard pre-correction actions, it is determined which correction actions conflict with each other. For example, if one action requires magnetic raising and another requires magnetic lowering, the optimal adjustment method is selected based on factors such as specific impurity concentration and removal efficiency. In this process, conflicting actions are merged or adjusted to ensure that the final adjustment action can remove impurities without causing loss of other treatment effects. The first pre-correction action finally generated will be an integrated, conflict-free optimization strategy.

[0105] The first pre-correction action is applied to perform cross-stage pre-correction on the second Gaussian setting benchmark. In this way, it is ensured that the second-stage magnetic separation can be fully optimized on the basis of the first-stage magnetic separation, and the magnetic separation treatment of each stage can cooperate with each other to achieve the final optimization effect. The second Gaussian joint adjustment parameter is the adjustment parameter obtained by applying the first pre-correction action. The second Gaussian joint adjustment parameter will serve as the control benchmark for the second-stage magnetic separation to ensure that the magnetic separation process can be further optimized and impurities can be removed on the basis of the first-stage magnetic separation.

[0106] Furthermore, a spatial mapping result of performing spatial coordinate mapping on the instantaneous impurity concentration data is input into a predefined impurity concentration-color scale mapping table for gradient coloring synthesis to obtain the first impurity distribution heat map.

[0107] Instantaneous impurity concentration data is the concentration data of impurities on the surface of iron powder collected in real time by the LIBS probe, reflecting the impurity distribution on the iron powder surface at the current moment. Spatial coordinate mapping is to map the instantaneous impurity concentration data to a spatial coordinate. This process associates the concentration value of the impurity with its specific location in the magnetic separation equipment. Through this mapping, the impurity concentration information will be presented as a spatial distribution, helping the system to identify changes in impurity concentration in different areas.

[0108] The impurity concentration-color scale mapping table is a predefined table that maps impurity concentrations to color scales (i.e., colors). Different impurity concentration values are mapped to different colors within the color scale to more intuitively represent the distribution of impurities. For example, higher impurity concentrations are mapped to red, while lower concentrations are mapped to green or blue. Gradient coloring synthesis combines the spatial coordinate mapping results with the impurity concentration-color scale mapping table to generate a heat map with a gradient color. This color gradient reflects the changes in impurity concentration, making it easy to quickly identify areas on the iron powder surface with high impurity concentrations and areas with better impurity removal. For example, high-concentration areas (i.e., areas with high impurity removal) on the heat map appear dark red, while low-concentration areas (i.e., areas with better impurity removal) appear light green or blue. This gradient coloring method intuitively presents the spatial distribution of impurity distribution. The above steps ultimately result in the first impurity distribution heat map, a visual representation of the spatial distribution of impurity concentrations on the iron powder surface, providing a basis for subsequent magnetic separation adjustments.

[0109] Furthermore, the multimodal iron powder composition characteristics include a particle size distribution curve, a saturation magnetization distribution, a coercive force distribution, and an initial impurity spectrum.

[0110] The multimodal iron powder component characteristics include particle size distribution curve, saturation magnetization intensity distribution, coercive force distribution and initial impurity spectrum. Among them, the particle size distribution curve is a graphical representation of the particle size distribution in the iron powder, reflecting the proportion of particles of different sizes in the iron powder. The particle size distribution is usually measured by screening or laser diffraction. The particle size distribution is very important for the magnetic separation process because particles of different sizes respond differently to the magnetic field. Larger particles are easier to be magnetically separated, while smaller particles require a stronger magnetic field or a longer processing time. The saturation magnetization intensity refers to the maximum magnetization intensity that the iron powder can reach under the action of an external magnetic field. This characteristic indicates the ability of the iron powder to be magnetized in a magnetic field. Usually It is related to factors such as the chemical composition and crystal structure of the iron powder. Coercivity refers to the magnetic strength that a magnetic material retains after the external magnetic field is removed. In the case of iron powder, the coercivity distribution shows how easily the iron powder particles regain their magnetism during the demagnetization process. Particles with high coercivity in iron powder are often difficult to remove by magnetic separation. Therefore, during the magnetic separation process, the magnetic field strength or processing time needs to be adjusted to remove these high-coercivity particles. The initial impurity spectrum refers to the initial concentration distribution of all impurity elements in the iron powder. It includes information on the initial concentrations of all impurity elements (such as silicon, aluminum, and calcium) in the iron powder and is used to determine which impurities are targeted for removal during magnetic separation and which require special treatment. By using multimodal iron powder component characteristics, the behavior and removal difficulty of different iron powder particles can be more effectively identified, and parameters such as magnetic field strength and execution time can be adjusted accordingly, thereby improving impurity removal efficiency and the purity of the final iron powder output.

[0111] Furthermore, multi-stage magnetic separation control parameter mapping is performed based on the multi-modal iron powder component characteristics to generate N-stage reference magnetic separation control parameters. The method includes:

[0112] After analyzing the particle size distribution curve and locating the median particle size position and the median particle size distribution width ratio, the first multi-level initialization magnetic separation control parameter is obtained by matching in the residence time spectrum mapping table according to the median particle size position and the median particle size distribution width ratio; the second multi-level initialization magnetic separation control parameter is obtained by matching in the field strength response spectrum mapping table according to the calculated central tendency value of the saturation magnetization intensity distribution and the high value area ratio of the coercive force distribution; after extracting the impurity concentration group corresponding to the core impurity group from the initial impurity spectrum, the third multi-level initialization magnetic separation control parameter is obtained by matching in the element removal response spectrum mapping table; conflicting parameters of the first multi-level initialization magnetic separation control parameter, the third multi-level initialization magnetic separation control parameter and the third multi-level initialization magnetic separation control parameter are arbitrated, and the N-level reference magnetic separation control parameter is output.

[0113] The particle size distribution curve is a statistical diagram of the particle size of iron powder, showing the distribution of particles of different particle sizes in the entire iron powder sample. Generally, the particle size distribution curve will show the different particle size segments of the particles and their corresponding number or mass distribution. Analyzing the particle size distribution curve can determine the particle size distribution of the particles in the iron powder, especially the median particle size, as well as the median particle size position and the median particle size distribution width percentage. Among them, the median particle size refers to the particle size value in the middle after the particle size distribution is sorted by size; the median particle size distribution width refers to the width of the particle size range, that is, the difference between the maximum and minimum particle sizes, which describes the breadth of the particle size distribution; the median particle size position is the midpoint in the particle size distribution, which divides the particle size distribution curve into two parts, making half of the particles smaller and the other half larger; the particle size distribution width percentage indicates the breadth of the particle size distribution. If the percentage is large, it means that the particle size is more varied, and if the percentage is small, it means that the particle size is more uniform.

[0114] The residence time spectrum mapping table is a predefined table used to calculate the residence time of particles of different sizes during the magnetic separation process based on the particle size distribution (specifically the median size position and the median size distribution width percentage). Residence time refers to the amount of time a particle remains in the magnetic separation system, with larger particles having shorter residence times and smaller particles having longer residence times. Based on the median size position and median size distribution width percentage, matching residence time parameters are found in the residence time spectrum mapping table to obtain the first multi-stage initialization magnetic separation control parameters, which are used in subsequent magnetic separation processes to appropriately process particles of different sizes.

[0115] The saturation magnetization distribution refers to the distribution of magnetization intensity within the iron powder particles under the influence of an external magnetic field. Central tendency values (e.g., average, peak, etc.) indicate the concentration of magnetization intensity within the saturation magnetization distribution. A higher central tendency value indicates that most particles have strong magnetism, while a lower central tendency value indicates weak magnetism. The coercivity distribution refers to the distribution of magnetic intensity retained by the iron powder particles after the external magnetic field is removed. The percentage of high-value areas indicates the proportion of particles with higher coercivity within the particles. Particles with higher coercivity are generally more difficult to remove through magnetic separation, requiring a stronger magnetic field or longer processing time.

[0116] The field intensity response spectrum mapping table is a predefined table used to adjust the magnetic field intensity during the magnetic separation process based on the magnetic properties of the particles (such as saturation magnetization and coercive force distribution). The choice of magnetic field intensity directly affects the magnetic separation efficiency. By analyzing the central tendency value of the saturation magnetization distribution and the proportion of high-value areas in the coercive force distribution, the most appropriate magnetic field intensity control parameters are found in the field intensity response spectrum mapping table. These second-level multi-stage initialization magnetic separation control parameters are then derived and used to optimize the magnetic separation process and ensure that the magnetic field intensity matches the magnetic properties of the iron powder particles.

[0117] The initial impurity spectrum refers to the concentration distribution of various impurity elements in the iron powder before magnetic separation. From the initial impurity spectrum, the impurity elements that are most challenging to the magnetic separation process are extracted, called the core impurity group. These impurity elements are the components that need to be removed during the magnetic separation process. The impurity concentration group refers to the concentration information of these core impurities, reflecting the distribution and concentration level of these impurities in the iron powder.

[0118] The Element Purge Response Spectrum Mapping Table is a predefined table that correlates impurity concentrations with magnetic separation control parameters (such as magnetic field strength and processing time). This table helps the system calculate the required magnetic separation parameters based on the concentration distribution of different impurities to effectively remove specific impurities. Based on the impurity concentration groups extracted from the core impurity group, matching magnetic separation parameters are found in the Element Purge Response Spectrum Mapping Table to obtain the third multi-stage initialization magnetic separation control parameters. This process provides targeted control parameters for subsequent magnetic separation steps to optimize the removal of these core impurities.

[0119] During the magnetic separation process, different initialization magnetic separation control parameters may conflict with each other. For example, some parameters recommend increasing the magnetic field strength, while others recommend reducing it. Conflicting parameter arbitration is to find an optimal adjustment strategy by resolving these conflicts. Arbitration methods include selecting optimal parameters, making compromises, or adjusting parameters according to actual conditions. Usually, these conflicts are resolved through weighted averaging, priority sorting, or other strategies. After the conflicts are resolved, N-level benchmark magnetic separation control parameters are generated based on the arbitration results. These parameters will serve as the control benchmark for the entire magnetic separation process.

[0120] In summary, the multi-stage magnetic separation method for high-purity iron powder provided in the embodiments of the present application has the following technical effects:

[0121] By calling the multimodal iron powder component characteristics across devices, the physical and chemical properties of the iron powder can be fully understood, which provides accurate data support for the subsequent magnetic separation process and ensures that the magnetic separation parameters can be adjusted in a targeted manner during the magnetic separation process; according to the multimodal iron powder component characteristics of the iron powder, N-level benchmark magnetic separation control parameters are generated through multi-level magnetic separation control parameter mapping. The control parameters of each level include Gaussian setting benchmark, Gaussian adjustable interval and execution time benchmark, ensuring that the magnetic separation process can accurately process iron powders of different particle sizes and magnetic strengths, so that the iron powder can be effectively graded to improve the purity; by using N execution time benchmarks as the magnetic separation time limit, the processing time of each level of magnetic separation is ensured to be accurately controlled to avoid over-processing or under-processing. Within the time limit of each magnetic separation level, the magnetic field strength and processing time are adjusted according to the results of the previous stage, so as to gradually remove the iron powder. Impurities; During the magnetic separation process, the concentration of impurities on the surface of the iron powder is detected in real time through the LIBS probe synchronous laser induced breakdown spectroscopy analysis, and the magnetic separation parameters can be dynamically adjusted according to the real-time data to form a closed-loop control. This dynamic closed-loop control can respond to changes in impurity concentration in real time to ensure that each magnetic separation level can achieve the best removal effect; At the end of each magnetic separation stage, an impurity distribution heat map is generated. Through the impurity distribution heat map, the control parameters can be optimized in the subsequent magnetic separation stage. Through the joint optimization of the Gaussian setting benchmark, the operation of each magnetic separation level is ensured to be coordinated and consistent, further improving the efficiency of impurity removal; Through the iterative execution of the magnetic separation process, the control strategy is continuously optimized until the required target purity is achieved. The results of each stage will affect the control strategy of the subsequent stage, thereby ensuring that the entire magnetic separation process can efficiently remove impurities and ultimately produce high-purity iron powder.

[0122] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-stage magnetic separation device for high-purity iron powder, characterized in that: The device comprises: Component feature calling module, used to call multi-modal iron powder component features across devices; a control parameter mapping module for performing multi-stage magnetic separation control parameter mapping based on the multi-modal iron powder component characteristics to generate N-stage reference magnetic separation control parameters, wherein each stage reference magnetic separation control parameter is composed of a Gaussian setting reference, a Gaussian adjustable interval, and an execution time reference; The magnetic separation processing module is used for the magnetic separation hardware system to perform N-level magnetic separation processing of the iron powder to be purified based on N execution time bases as the magnetic separation time limit and N Gauss setting bases. The magnetic separation processing module includes: A closed-loop control unit is used to perform dynamic closed-loop control of single-layer magnetic separation parameters within N Gaussian adjustable intervals based on the results of the synchronized laser-induced breakdown spectroscopy analysis of the LIBS probe; The joint debugging optimization unit is used to perform cross-level Gaussian benchmark joint debugging optimization based on the impurity distribution heat map of N execution time ends; The iterative execution unit is used for iterative execution until N-level iron powder magnetic separation is completed to produce iron powder with target purity.

2. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 1, characterized in that: The joint debugging and optimization unit further includes: a pre-selection magnetic separation processing channel for operating the magnetic separation hardware system using a first Gaussian setting reference of the first-level reference magnetic separation control parameter, and then starting a feeder to convey the iron powder to be purified into the magnetic separation hardware system for pre-selection magnetic separation processing; The spectral analysis channel is used to synchronously run the LIBS probe to perform laser-induced breakdown spectroscopy analysis on the conveyed iron powder during the pre-selection magnetic separation process, and output a real-time surface impurity concentration sequence; A dynamic parameter update channel is used to perform dynamic parameter update of the first Gaussian setting benchmark based on the multivariate impurity fluctuation characteristics of the real-time surface impurity concentration sequence, with the first Gaussian adjustable interval as a constraint, until the first execution time benchmark is reached, and generate a first impurity distribution heat map.

3. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 2, characterized in that: The iterative execution unit further includes: a cross-level pre-correction channel, configured to perform cross-level pre-correction on a second Gaussian setting reference according to the first impurity distribution thermodynamic map to obtain a second Gaussian joint tuning parameter; a parameter update channel for executing parameter update of the second Gaussian joint tuning parameters according to surface impurity concentration data synchronously detected by the LIBS probe during the secondary magnetic separation process of the magnetic separation hardware system using the second Gaussian joint tuning parameters, until the magnetic separation duration reaches a second execution time benchmark, and outputting a second impurity distribution heat map; The Gaussian benchmark pre-tuning channel is used to iteratively perform dynamic adjustment of magnetic separation parameters in a single-layer magnetic separation process based on the surface impurity concentration, and iteratively perform cross-layer Gaussian benchmark pre-tuning based on the layer impurity thermal distribution until the N-level benchmark magnetic separation control parameters are executed and the target purity iron powder is produced.

4. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 2, characterized in that: The dynamic parameter update channel includes: The Gaussian adjustment action library construction node is used to interactively obtain multiple sets of Gaussian adjustment actions for various iron powder impurities under multiple sets of impurity fluctuation scenarios to build a Gaussian adjustment action library; A parallel matching node, configured to input the multivariate impurity fluctuation characteristics as multivariate search conditions into the Gaussian adjustment action library to parallel match multivariate Gaussian adjustment actions for multivariate impurity fluctuation scenarios; a union solving node, configured to perform a union solving on the multivariate Gaussian adjustment actions to locate a first real-time adjustment action; a parameter update node, configured to perform parameter update of the first Gaussian setting benchmark using the first real-time adjustment action; The data acquisition node is used to drive the LIBS probe to collect instantaneous impurity concentration data and synthesize the first impurity distribution thermodynamic map when the first Gaussian setting benchmark is dynamically adjusted and updated and the first execution time benchmark is reached.

5. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 3, characterized in that: The cross-stage pre-correction channel includes: A thermal feature detection network construction node, used to pre-construct a thermal feature detection network, wherein the thermal feature detection network includes M thermal map feature recognition channels; a feature parallel recognition node, configured to load the first impurity distribution thermodynamic map into the thermodynamic feature detection network, perform feature parallel recognition via the M thermodynamic map feature recognition channels, and output M real-time thermodynamic map quantitative features; The cross-level pre-correction node is used to perform cross-level pre-correction on the second Gaussian setting benchmark after pre-correction action matching according to the M real-time heat map quantitative features to obtain the second Gaussian joint adjustment parameters.

6. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 5, characterized in that: The cross-level pre-correction node includes: The standard pre-correction action setting sub-node is used to locally set M groups of standard pre-correction actions for M groups of heat map quantitative feature thresholds; an action extraction subnode, configured to extract M standard pre-correction actions from the M groups of standard pre-correction actions by using the intersection relationship between the M real-time heat map quantitative features and the M groups of heat map quantitative feature thresholds; a conflict compensation subnode, configured to perform conflict compensation on the M standard pre-correction actions to obtain a first pre-correction action; The cross-level pre-correction sub-node is used to perform cross-level pre-correction on the second Gaussian setting benchmark by using the first pre-correction action to obtain the second Gaussian joint tuning parameter.

7. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 4, characterized in that: The dynamic parameter update channel is used to: input the spatial mapping result of spatial coordinate mapping of the instantaneous impurity concentration data into a predefined impurity concentration-color level mapping table for gradient coloring synthesis to obtain the first impurity distribution heat map.

8. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 1, characterized in that: The multimodal iron powder component characteristics include a particle size distribution curve, a saturation magnetization intensity distribution, a coercive force distribution, and an initial impurity spectrum.

9. The multi-stage magnetic separation equipment for high-purity iron powder according to claim 8, characterized in that: The control parameter mapping module includes: a first matching unit for, after analyzing the particle size distribution curve and locating the median particle size position and the median particle size distribution width ratio, matching the residence time spectrum mapping table to obtain first multi-stage initialization magnetic separation control parameters according to the median particle size position and the median particle size distribution width ratio; a second matching unit, configured to obtain a second multi-level initialization magnetic separation control parameter by matching the field intensity response spectrum mapping table according to the calculated central tendency value of the saturation magnetization distribution and the proportion of the high value region of the coercive force distribution; a third matching unit, configured to extract an impurity concentration group corresponding to a core impurity group from the initial impurity spectrum, and then obtain a third multi-stage initialization magnetic separation control parameter by matching the impurity concentration group in an element removal response spectrum mapping table; The conflict parameter arbitration unit is used to perform conflict parameter arbitration on the first multi-level initialization magnetic separation control parameter, the third multi-level initialization magnetic separation control parameter and the third multi-level initialization magnetic separation control parameter, and output the N-level reference magnetic separation control parameter.

10. A multi-stage magnetic separation method for high-purity iron powder, characterized in that: Based on the multi-stage magnetic separation equipment of high-purity iron powder according to any one of claims 1 to 9, the method comprises: Call multimodal iron powder component characteristics across devices; Performing multi-stage magnetic separation control parameter mapping according to the multi-modal iron powder component characteristics to generate N-stage reference magnetic separation control parameters, wherein each stage reference magnetic separation control parameter consists of a Gaussian setting reference, a Gaussian adjustable interval, and an execution time reference; The magnetic separation hardware system uses N execution time benchmarks as the magnetic separation time limit. When executing N-level magnetic separation processes of the iron powder to be purified according to N Gauss setting benchmarks: Step a: Based on the results of the synchronous laser-induced breakdown spectroscopy analysis of the LIBS probe, dynamic closed-loop control of the single-layer magnetic separation parameters is performed within N Gaussian adjustable intervals; Step b: Based on the impurity distribution heat map at the end of N execution times, perform cross-level Gaussian benchmark joint optimization; The process is iterated until N levels of iron powder magnetic separation are completed, producing iron powder of the target purity.

Images