Method for monitoring concentrate grade in jigging enrichment process

Abstract

The invention relates to the field of concentrate grade monitoring, in particular to a method for monitoring the concentrate grade in the jigging enrichment process. The method comprises the following steps: synchronously acquiring multi-source original data in the same time window, and carrying out primary processing to obtain monitoring basic state parameters; basic reference data is obtained by executing a calibration process before monitoring. And based on the monitored basic state parameters and basic reference data, introducing a pulsation constraint four-domain coupling hierarchical inversion algorithm, constructing an effective enrichment potential index and a jump suppression factor in combination with a hierarchical inversion mode, and obtaining the concentrate grade. The problems that a traditional concentrate grade monitoring method is long in detection period, obvious in hysteresis quality and incapable of meeting the requirement for real-time data under the continuous production condition, meanwhile, sampling points are limited and random, detection results are insufficient in representativeness, the real state of the whole discharge flow is difficult to comprehensively reflect, and the detection efficiency is poor are solved. Therefore, continuous, accurate and stable online monitoring of the concentrate grade in the jigging enrichment process is difficult to realize.

Description

A method for monitoring concentrate grade in a jigging enrichment process Technical Field

[0001] This invention relates to the field of concentrate grade monitoring, and more particularly to a method for monitoring concentrate grade during a jigging enrichment process. Background Technology

[0002] Jigging enrichment, a typical gravity separation method, is widely used in the separation of coal, iron ore, and non-ferrous metal minerals. Its basic principle is to utilize the density differences between mineral particles to form a stratified structure under the action of periodic pulsating water flow, thereby achieving the separation of valuable minerals from gangue. In actual production, the concentrate grade is one of the core indicators for evaluating the jigging separation effect, directly affecting the efficiency of subsequent processing and the stability of product quality. Therefore, accurate monitoring of the concentrate grade is of great significance.

[0003] Currently, the acquisition of concentrate grade mainly relies on manual sampling and laboratory analysis, indirectly reflecting the sorting effect by detecting ash content, grade, or composition of samples. However, this method has a long detection cycle and significant lag, failing to meet the demand for real-time data under continuous production conditions. Furthermore, the sampling points are limited and random, resulting in insufficient representativeness of the test results and an inability to comprehensively reflect the true state of the entire discharge flow. This leads to the problem of continuous, accurate, and stable online monitoring of concentrate grade during the jigging enrichment process. Summary of the Invention

[0004] This invention provides a method for monitoring the grade of concentrates during the jigging enrichment process. This method addresses the problems of traditional concentrate grade monitoring methods, which have long detection cycles, significant lag, and cannot meet the demand for real-time data under continuous production conditions. Furthermore, the limited and random sampling points result in insufficient representativeness of the detection results, making it difficult to fully reflect the true state of the entire discharge flow. Consequently, it is difficult to achieve continuous, accurate, and stable online monitoring of concentrate grade during the jigging enrichment process.

[0005] The present invention provides a method for monitoring the grade of concentrate in a jigging enrichment process, comprising the following steps:

[0006] S1. Simultaneously collect raw data from multiple sources within the same time window, perform preliminary processing, and obtain basic monitoring status parameters; obtain basic reference data by performing a calibration process before monitoring;

[0007] S2. Based on the monitoring of basic state parameters and basic reference data, a pulse-constrained four-domain coupled hierarchical inversion algorithm is introduced. Combined with the hierarchical inversion method, an effective enrichment potential index and a jump suppression factor are constructed, and the fine material grade is obtained.

[0008] Preferably, the multi-source raw data and the basic monitoring status parameters respectively include:

[0009] The multi-source raw data includes incident and transmitted ray intensity, electromagnetic signal emission intensity, electromagnetic signal reception intensity, particle image data, and bed vibration data; the basic state parameters monitored include mineral flow apparent density, average particle size, particle size dispersion, median particle size, electromagnetic response characteristic value of mineral flow, and stable density.

[0010] Preferably, the preliminary processing specifically includes:

[0011] The incident and transmitted ray intensities are processed separately to obtain stable incident and transmitted ray intensities, and the apparent density of the mineral flow is calculated. The particle image data is processed to obtain the particle diameter and the particle size weight is calculated. Based on the particle diameter and particle size weight, the average particle size, particle size dispersion, and median particle size are obtained. The electromagnetic signal emission intensity and electromagnetic signal reception intensity are processed separately to obtain stable emission intensity and reception intensity, and the electromagnetic response characteristic values ​​of the mineral flow are obtained.

[0012] Preferably, the preliminary processing further includes:

[0013] A bed vibration data sequence was constructed based on bed vibration data, and a spectrum analysis was performed to extract the dominant frequency. Based on the dominant frequency, the apparent density of the mineral flow was stabilized under operating conditions to obtain a stable density.

[0014] Preferably, the calibration process before monitoring specifically includes:

[0015] By introducing low-grade and high-grade concentrate samples with known grades, and applying the same methods as those used in the above calculations of mineral flow apparent density, electromagnetic response characteristic values, and dominant frequency extraction, reference apparent density, reference electromagnetic response characteristic values, and reference bed pulsation frequency are obtained. Furthermore, by performing chemical analysis on the low-grade and high-grade concentrate samples, the upper and lower grade boundaries are obtained.

[0016] Preferably, the implementation process of the pulsation-constrained four-domain coupled hierarchical inversion algorithm specifically includes:

[0017] Based on the monitoring of basic state parameters and basic reference data, a stable density gain term, an electromagnetic gain term, a granular discrete penalty term, and a pulsation confidence term are constructed, and an effective enrichment potential index is generated.

[0018] Preferably, the implementation process of the pulsation-constrained four-domain coupled hierarchical inversion algorithm further includes:

[0019] By introducing a bounded mapping, the effective enrichment potential index is mapped to a bounded grade ratio, thus obtaining the static inversion result.

[0020] Preferably, the implementation process of the pulsation-constrained four-domain coupled hierarchical inversion algorithm further includes:

[0021] Based on the effective enrichment potential index of two adjacent time windows, the effective enrichment potential index and jump suppression factor after time consistency correction are obtained; based on the effective enrichment potential index and jump suppression factor after time consistency correction, the time consistency constraint term is generated.

[0022] Preferably, the implementation process of the pulsation-constrained four-domain coupled hierarchical inversion algorithm further includes:

[0023] Based on the dominant oscillation frequency, a dynamic fusion weight is constructed. Combined with the static inversion results and time consistency constraints, the final grade ratio is obtained. Combining the upper and lower grade boundaries, the final grade ratio is mapped to the actual fine feed grade.

[0024] The beneficial effects of the technical solution of the present invention are:

[0025] 1. By constructing a pulse-constrained four-domain coupled hierarchical inversion algorithm, four types of physical information—stability density, electromagnetic response, particle size structure, and vibration state—are coupled and uniformly characterized using an effective enrichment potential index, achieving a high-dimensional mapping from multi-source physical characteristics to a single grade index. The pulse-constrained four-domain coupled hierarchical inversion algorithm maintains reasonable sensitivity for each physical quantity across different numerical ranges through the construction of a nonlinear gain and constraint mechanism, thus avoiding excessive dominance of any single parameter in the results. In particular, by introducing a particle size discrete penalty term and a pulse confidence term, the algorithm can automatically identify abnormal conditions such as uneven particle distribution and enhanced bed disturbance, and suppress the detection results, thereby significantly improving the stability of grade estimation.

[0026] 2. By constructing a jump suppression factor based on the geometric mean enrichment potential and logarithmic rate of change of adjacent time windows, a continuous constraint on the grade change trend is achieved. This prevents instantaneous anomalies from being directly mapped to grade mutations when facing short-term detection fluctuations, effectively avoiding the "peak misjudgment" problem common in traditional online monitoring. Simultaneously, by coupling the time consistency constraint strength coefficient with the vibration state, the time smoothing strength can automatically adjust with changes in operating conditions, maintaining rapid response capability under stable operating conditions and enhancing anti-interference capability under disturbed operating conditions, achieving a balance between response speed and stability. Attached Figure Description

[0027] Figure 1 is a flowchart of a method for monitoring the grade of concentrate in a jigging enrichment process according to the present invention. Detailed Implementation

[0028] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0030] The following describes in detail, with reference to the accompanying drawings, a specific scheme for monitoring the concentrate grade in a jigging enrichment process provided by the present invention.

[0031] Referring to Figure 1, a flowchart of a method for monitoring the grade of concentrate in a jigging enrichment process according to an embodiment of the present invention is shown. The method includes the following steps:

[0032] S1. Simultaneously collect raw data from multiple sources within the same time window and perform preliminary processing to obtain basic monitoring status parameters; obtain basic reference data by performing a calibration process before monitoring.

[0033] By arranging gamma-ray emitters and receivers on both sides of the feed discharge channel, the gamma-ray beam traverses the entire cross-section of the feed flow. Electromagnetic excitation coils and electromagnetic receiving coils are then positioned downstream of the gamma-ray detection zone (i.e., the entire cross-section of the feed flow traversed by the gamma-ray beam), allowing the same batch of feed particles to continue entering the electromagnetic response detection zone after density testing. Simultaneously, an industrial camera and surface light source are installed outside the feed discharge channel to continuously acquire particle images. The lens axis of the industrial camera forms a fixed angle with the discharge surface, determined based on the ore particle size. For example, 50°–55° is used when the ore particle size is greater than 15mm, and 55°–65° is used when the ore particle size is less than 5mm. Vibration sensors are fixed on the jig bed housing to collect the bed pulsation frequency.

[0034] Through the above arrangement, multiple sources of raw data are collected simultaneously within the same time window, including incident ray intensity and transmitted ray intensity, electromagnetic signal emission intensity and electromagnetic signal reception intensity, particle image data and bed vibration data.

[0035] At the start of monitoring, a fixed time window is used. For example, for 4 seconds, multiple sources of raw data are continuously acquired. Within each time window, the incident ray intensity sequence and transmitted ray intensity sequence obtained by the gamma-ray receiver are first averaged and then denoised by pulse, respectively, to obtain the result at the 4th time window. Stable incident ray intensity within a time window and intensity of transmitted rays Then, based on the existing feasible ray attenuation relationships, calculate the apparent density of mineral flow for this time window:

[0036] in, It is the first Apparent density of mineral flow within a time window; This represents the linear attenuation coefficient, which reflects the linear attenuation coefficient of the mineral at the energy level of the radiation used. It is obtained by consulting a standard physical property database. The effective transmission thickness of the fine material discharge channel in the direction of the X-ray path is obtained directly from the equipment design drawings; this step is a mature and feasible existing detection method.

[0037] Within the same time window, firstly, the multi-frame particle image data captured continuously by an industrial camera is converted to grayscale using the averaging method to obtain grayscale particle image data. Then, the grayscale particle image data is used for background separation using an adaptive thresholding method such as mean-adaptive thresholding to obtain separated particle image data. Next, the separated particle image data is enhanced using the Canny edge detection algorithm to obtain enhanced particle image data. Finally, based on the enhanced particle image data, connected component analysis is performed on the binarized foreground region to determine the contacting particle regions. The watershed segmentation algorithm is then used to further divide the contacting particle regions, thus obtaining independent particle regions. Next, a boundary tracking algorithm is used to extract the contour of each independent particle region, generating closed particle boundaries. After obtaining the complete contour, the number of pixels within each independent particle region is counted and multiplied by the square of the actual pixel size obtained from camera calibration to obtain the pixel area, i.e., the projected area of ​​the particle. Finally, this projected area is converted into an equivalent circle diameter to characterize the size features of particles of different shapes. The diameter of each particle is obtained through the above process, and the proportion of the projected area of ​​each particle in the total projected area of ​​all particles is used as the particle size weight. Furthermore, based on the particle size weight, the particle image data after contour extraction is calculated at the [missing information - likely a specific value or parameter]. Average particle size within a time window With particle size dispersion (i.e., standard deviation) Then, after sorting the particle size weights from smallest to largest, the median particle size is calculated based on the particle size corresponding to the cumulative particle size weight reaching 50%. .

[0038] Apply a fixed frequency to the concentrate stream within each time window An alternating excitation magnetic field (set by technicians, e.g., 20kHz) is used to measure the electromagnetic signal transmission and reception intensities. Electromagnetic signal transmission and reception intensity sequences are then constructed. The transmission and reception intensities are then averaged and nonlinearly filtered to obtain the result at the [missing value]. Stable emission intensity within a time window and reception strength Then, regarding the launch intensity and reception strength After taking the logarithm of the ratio, it is divided by a fixed frequency. The square root of the value yields the characteristic value of the electromagnetic response of the mineral flow. .

[0039] Bed vibration data were acquired within each time window, and a bed vibration data sequence was constructed. Spectral analysis was performed using Fast Fourier Transform to extract the dominant frequency. Simultaneously calculate the past The average master frequency of operation within each time window As a specific embodiment, here The possible value is 5. To avoid bed pulsation directly altering the instantaneous particle accumulation state in the detection zone, the apparent density of the mineral flow is first stabilized using existing compensation techniques to obtain a stable density:

[0040] in, It is in the Stable density within a time window; It is the pulsation frequency influencing factor, which reflects the impact of bed pulsation frequency on detection stability. It is used to mitigate and compensate for the intensity of vibration, so that the impact of vibration is not excessively amplified; This represents the density value after correction for particle size, reflecting the compensation for the deviation in density measurement caused by particle structure. It is a typical S-shaped function that can map vibration frequency to disturbance intensity factor.

[0041] After the above preliminary processing of the multi-source raw data, the basic state parameters for monitoring are obtained, including the apparent density of mineral flow, average particle size, particle size dispersion, median particle size, electromagnetic response characteristic value of mineral flow, and stable density.

[0042] Based on the above calculation method, a calibration process is performed before monitoring begins to obtain basic reference data for dimensionless conversion: one set of low-grade concentrate sample and one set of high-grade concentrate sample of known grade are taken, and under stable operating conditions, the reference apparent density of the low-grade sample is calculated in the same manner as described above. Reference electromagnetic response characteristic value and reference bed pulsation frequency The reference bed pulsation frequency is the dominant frequency extracted by fast Fourier transform under stable operating conditions. Standard chemical analysis is performed on low-grade and high-grade concentrate samples respectively. That is, based on the mineral type, standard laboratory quantitative analysis methods are used to accurately determine the content of target components in the calibrated samples, thus determining the lower grade boundary. and the boundaries of taste The stable operating condition refers to the operating state in which the equipment operating parameters and material state remain relatively constant within a certain time interval during the jigging enrichment process.

[0043] S2. Based on the monitoring of basic state parameters and basic reference data, a pulse-constrained four-domain coupled hierarchical inversion algorithm is introduced. Combined with the hierarchical inversion method, an effective enrichment potential index and a jump suppression factor are constructed, and the fine material grade is obtained.

[0044] Based on the monitoring of fundamental state parameters and basic reference data, a pulsation-constrained four-domain coupled hierarchical inversion algorithm is introduced. Combining the hierarchical inversion method, factors beneficial to grade improvement and factors weakening detection reliability are respectively constructed as effective enrichment potential indices and jump suppression factors, and mapped to grade. The specific implementation process is as follows:

[0045] First, define the... Effective enrichment potential indicators for each time window Constructing an effective enrichment potential index The following physical logic applies: the higher the stable density, the higher the proportion of heavy useful minerals in the concentrate; the stronger the electromagnetic response, the higher the probability of the presence of a specific target mineral; the greater the particle size dispersion, the stronger the inhomogeneity of the current particle group, and the worse the representativeness of the detection; the more obvious the deviation of the current bed pulsation frequency, the lower the reliability of the real-time detection in the current time window. Based on this definition:

[0046] in, It is the first The effective enrichment potential index for each time window is used to describe the comprehensive results of the enrichment intensity of useful minerals in the concentrate stream, the stability of detection, and the reliability of data. The reason for using logarithmic form for the stable density gain term is that after the density increases to a certain level, the grade no longer increases infinitely in a linear proportion. Introducing logarithmic compression can suppress the excessive amplification of the high-density region. It is the electromagnetic gain term, and the square root form is used to reduce the dominance of the electromagnetic polar response on the overall result. It is a particle size dispersion penalty term, used to characterize the particle size dispersion of the particle group within the current time window; It is the pulsation reliability term, which is used to exponentially attenuate the reliability of the current detection data based on the degree of deviation of the bed pulsation frequency, thereby suppressing the non-true grade fluctuations caused by operating condition disturbances.

[0047] Furthermore, since the grade of concentrate is essentially a mass fraction with an upper limit, it cannot be determined by the effective enrichment potential index. As it increases indefinitely, a bounded mapping needs to be introduced for the first-level inversion to obtain the effective enrichment potential index. Mapped to bounded grade ratios:

[0048] in, , indicating the first The static grade ratio obtained by direct inversion based on the effective enrichment potential index within a time window is the static inversion result.

[0049] However, the jigging process has a clear temporal continuity, and the grade of the concentrate will not undergo a drastic change without a physical basis in the absence of material mutations. Therefore, only using This is still insufficient to suppress transient errors. Therefore, a second-level hierarchical inversion term is constructed. First, the geometric mean of the effective enrichment potential indexes of two adjacent time windows is taken to maintain unit consistency and highlight the material inertia of continuous production, thus obtaining the effective enrichment potential indexes corrected for time consistency. :

[0050] in, It is the first Effective enrichment potential indicators for a time window;

[0051] Then, based on the logarithmic rate of change of the effective enrichment potential index of adjacent time windows, a jump suppression factor is constructed. :

[0052] in, The larger the value, the greater the difference in enrichment potential between the current time window and the previous time window;

[0053] Furthermore, based on the effective enrichment potential index and jump suppression factor after time consistency correction, a time consistency constraint term is constructed:

[0054] in, It is in the Within a time window, the dimensionless estimate of the grade of the concentrate after time consistency constraint, i.e., the time consistency constraint term; This is the time consistency constraint strength coefficient, used to adjust the time consistency constraint strength, and is directly determined based on the vibration conditions. The derivation logic of the above formula is as follows: first use the effective enrichment potential index after time consistency correction. Preserve the shared enrichment trend of the current and previous time windows, and then add to the denominator. As a jump penalty term, it makes the time consistency constraint term when continuity difference occurs. Automatic reduction, thereby avoiding instantaneous spikes from being directly mapped to high-quality results;

[0055] Next, dynamic fusion weights are constructed. Since the reliability of the current time window is related to the degree of bed pulsation deviation, and the time consistency constraint term is suitable for bearing a larger weight under strong disturbances, a probabilistic mapping model based on the Sigmoid function is introduced to define the dynamic fusion weights. :

[0056] in, It is a dynamic fusion weight, representing the weight at the th... Within a time window, the proportion of the time consistency constraint in the final grade calculation; when the current bed pulsation frequency is close to the average operating dominant frequency. Smaller When the value approaches a smaller value, it indicates that the current detection is relatively stable, and static inversion results can be used more frequently; when the deviation increases, Increasing this value will automatically increase the proportion of the time consistency constraint, thus yielding the final grade ratio. :

[0057] Finally, the final grade ratio, combined with the upper and lower grade boundaries, is mapped to the actual refined feed grade:

[0058] in, For the first The quality of the refined materials output within each time window.

[0059] In summary, a method for monitoring the grade of concentrates during the jigging enrichment process has been developed.

[0060] The order of the embodiments is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0061] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0062] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for monitoring the grade of concentrate in a jigging enrichment process, characterized in that, Includes the following steps: S1. Simultaneously collect raw data from multiple sources within the same time window, perform preliminary processing, and obtain basic monitoring status parameters; obtain basic reference data by performing a calibration process before monitoring; S2. Based on the monitoring of basic state parameters and basic reference data, a pulse-constrained four-domain coupled hierarchical inversion algorithm is introduced. Combined with the hierarchical inversion method, an effective enrichment potential index and a jump suppression factor are constructed, and the fine material grade is obtained.

2. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 1, characterized in that, The multi-source raw data and monitoring basic state parameters include: the multi-source raw data includes incident ray intensity and transmitted ray intensity, electromagnetic signal emission intensity, electromagnetic signal reception intensity, particle image data and bed vibration data; the monitoring basic state parameters include mineral flow apparent density, average particle size, particle size dispersion, median particle size, electromagnetic response characteristic value of mineral flow and stable density.

3. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 2, characterized in that, The preliminary processing specifically includes: processing the incident ray intensity and transmitted ray intensity separately to obtain stable incident ray intensity and transmitted ray intensity, and calculating the apparent density of the mineral flow; processing the particle image data to obtain the particle diameter and calculating the particle size weight, and obtaining the average particle size, particle size dispersion, and median particle size based on the particle diameter and particle size weight; processing the electromagnetic signal emission intensity and electromagnetic signal reception intensity separately to obtain stable emission intensity and reception intensity, and obtaining the electromagnetic response characteristic value of the mineral flow.

4. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 3, characterized in that, The preliminary processing also includes: constructing a bed vibration data sequence based on bed vibration data and performing spectral analysis to extract the dominant frequency; and stabilizing the apparent density of the mineral flow based on the dominant frequency to obtain a stable density.

5. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 4, characterized in that, The calibration process prior to monitoring specifically includes: introducing low-grade concentrate samples and high-grade concentrate samples of known grade; applying the same methods as the above-mentioned calculation of apparent density of mineral flow, calculation of electromagnetic response characteristic value of mineral flow, and extraction of dominant frequency to obtain reference apparent density, reference electromagnetic response characteristic value, and reference bed pulsation frequency; and obtaining the upper and lower grade boundaries by performing chemical analysis on the low-grade concentrate samples and high-grade concentrate samples.

6. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 5, characterized in that, The implementation process of the pulsation-constrained four-domain coupled hierarchical inversion algorithm specifically includes: based on the monitored basic state parameters and basic reference data, constructing a stable density gain term, an electromagnetic gain term, a granular discrete penalty term, and a pulsation confidence term, and generating an effective enrichment potential index.

7. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 6, characterized in that, The implementation process of the pulsed constraint four-domain coupled hierarchical inversion algorithm also includes: introducing a bounded mapping to map the effective enrichment potential index to a bounded grade ratio, thereby obtaining the static inversion result.

8. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 7, characterized in that, The implementation process of the pulsed constraint four-domain coupled hierarchical inversion algorithm also includes: obtaining the effective enrichment potential index and jump suppression factor after time consistency correction based on the effective enrichment potential index of two adjacent time windows; and generating time consistency constraint terms based on the effective enrichment potential index and jump suppression factor after time consistency correction.

9. The method for monitoring the grade of concentrate in a jigging enrichment process according to claim 8, characterized in that, The implementation process of the pulsed constraint four-domain coupled hierarchical inversion algorithm also includes: constructing dynamic fusion weights based on the dominant oscillation frequency, combining the static inversion results and time consistency constraints to obtain the final grade ratio; and mapping the final grade ratio to the actual fine feed grade by combining the upper and lower grade boundaries.

Images