Machine learning based shaker parameter adaptive control method and system

Abstract

The application relates to the technical field of parameter control, and discloses a vibration screen parameter self-adaptive control method and system based on machine learning. The method comprises the following steps: collecting coal characteristic parameters through an image device and a sensor; calculating a load coefficient and a difficulty index according to the parameters, adjusting double-motor exciter parameter control vibration; adjusting spring pre-tightening force and screen surface inclination according to vibration feedback, and optimizing screening track; adjusting feeding speed and baffle height based on screening efficiency data, and realizing self-adaptive control. The application solves the technical problems that the traditional vibration screen fixed parameter control cannot adapt to the change of multiple coal types, the vibration screen control system lacks real-time sensing and analysis ability of coal material characteristics, and there is a lack of accurate mapping and dynamic optimization mechanism between vibration parameters and screening efficiency.

Description

Technical Field

[0001] The present application relates to the field of parameter control technology, and in particular to a vibrating screen parameter adaptive control method and system based on machine learning. Background Art

[0002] Vibrating screens are key equipment for material sorting and grading in coal processing. Traditional vibrating screen control systems use a preset parameter control method. During the equipment commissioning phase, parameters such as vibration frequency, amplitude, and screen surface inclination are determined based on the specific coal type and maintained constant during production. This control method calculates key parameters through empirical formulas. When processing a single coal type and operating under stable conditions, it can meet basic production requirements and achieve coal material screening.

[0003] However, with the diversification of coal production processes, the properties of the coal materials that vibrating screens need to process are becoming increasingly complex and variable, with the fluctuation range of characteristic parameters such as hardness, moisture content, and particle size distribution continuing to expand. Fixed parameter control methods are clearly insufficient in the face of such variable coal characteristics, mainly manifested in three aspects: first, the control parameters cannot be automatically adjusted according to the changes in coal characteristics, resulting in a significant decrease in screening efficiency during the coal type conversion process; second, energy utilization efficiency is low, and the same vibration parameters are used for coal types with different characteristics, resulting in energy waste; third, there is a lack of real-time monitoring and feedback mechanism for material characteristics and equipment status, which cannot promptly respond to screening anomalies caused by changes in operating conditions.

[0004] The deeper technical problem lies in how to effectively identify the characteristics of variable coal types and establish an accurate dynamic control model. Although the existing semi-automatic control system has introduced some sensor monitoring, it lacks the ability to fuse multi-source data and cannot fully characterize the characteristics of coal materials. At the same time, the vibrating screen system itself is highly nonlinear and coupled, and traditional control strategies have difficulty dealing with the collaborative optimization of multiple objectives such as vibration direction, intensity, and uniformity. In addition, the screening efficiency evaluation and feedback adjustment mechanism is not perfect, especially when dealing with coal types with high moisture content or wide particle size distribution. It is impossible to achieve accurate matching and dynamic adjustment of screening parameters, resulting in large fluctuations in screening process efficiency, high energy consumption, and frequent screen blockage. Summary of the Invention

[0005] The present application provides a method and system for adaptive control of vibrating screen parameters based on machine learning, which is used to solve the technical problems that the fixed parameter control of traditional vibrating screens cannot adapt to the changes of multiple coal types, the vibrating screen control system lacks the ability to perceive and analyze coal material characteristics in real time, and there is a lack of accurate mapping and dynamic optimization mechanism between vibration parameters and screening efficiency.

[0006] In the first aspect, the present application provides a vibrating screen parameter adaptive control method based on machine learning, which includes: collecting data on the coal material at the feed end of the vibrating screen through an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen surface to obtain a characteristic parameter set; calculating the current coal load coefficient and screening difficulty index based on the characteristic parameter set, and adjusting the phase difference and current input of the dual-motor vibrator through the control system to obtain the vibration direction and intensity; based on the feedback data of the vibration direction and intensity, adjusting the preload force of the screen box support spring and the extension and contraction amount of the screen surface inclination hydraulic cylinder to obtain the screening trajectory and material movement speed; controlling the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers.

[0007] Optionally, the coal material at the feed end of the vibrating screen is collected through an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen to obtain a characteristic parameter set, including:

[0008] Acquire a multi-angle continuous image sequence of the coal surface by the image acquisition device, perform coal block boundary recognition on the image sequence, and obtain a coal particle size distribution feature matrix;

[0009] The near-infrared spectral sensor array installed at the front end of the screen is used to perform multi-point scanning of the coal material. After eliminating background noise using wavelet transform, characteristic spectral peaks are extracted to obtain the coal moisture content distribution map and mineral composition fingerprint.

[0010] The signals collected by the strain gauge weight sensor installed at the bottom of the screen box and the triaxial acceleration sensor on the side wall of the screen box are separated by an adaptive filter to obtain the dynamic load characteristic curve of the coal material.

[0011] The coal particle size distribution characteristic matrix, the coal moisture content distribution map and the mineral composition fingerprint, and the coal material dynamic load characteristic curve are subjected to dimensionality reduction and feature fusion through nonlinear embedding to generate a characteristic parameter set.

[0012] Optionally, the current coal load factor and screening difficulty index are calculated according to the characteristic parameter set, and the phase difference and current input of the dual-motor vibrator are adjusted by the control system to obtain the vibration direction and intensity, including:

[0013] Multiplying the particle size distribution characteristic matrix in the characteristic parameter set by a preset weight coefficient, summing the weighted values ​​of all particle size intervals, and multiplying the sum by the spatial average value of the moisture content distribution map to obtain a first coal load coefficient;

[0014] Based on the mineral composition fingerprint and dynamic load characteristic curve in the characteristic parameter set, the ratio of clay mineral content to hard mineral content is calculated, and combined with the fluctuation coefficient to generate a first screening difficulty index;

[0015] Applying a nonlinear mapping function to the first coal load coefficient and performing weighted fusion with the first screening difficulty index to obtain a second screening difficulty index and a second coal load coefficient;

[0016] Constructing a parameter mapping table according to the second coal load factor and the second screening difficulty index, and calculating the phase difference and current input value of the dual-motor vibrator by an interpolation algorithm;

[0017] Convert the phase difference into a phase control signal, convert the current input value into a frequency control signal, and send the signal to the motor drive unit through a controller to generate a vibration direction and vibration intensity;

[0018] The actual vibration parameters are measured by the vibration sensor installed on the screen box, compared with the theoretical calculated values, and the compensation amount is calculated through closed-loop control to adjust the vibration direction and vibration intensity.

[0019] Optionally, the preload force of the screen box support spring and the expansion and contraction amount of the screen surface inclination hydraulic cylinder are adjusted based on the feedback data of the vibration direction and intensity to obtain the screening trajectory and material movement speed, including:

[0020] Decomposing the three-axis acceleration signal in the feedback data of the vibration direction and intensity into horizontal and vertical components, calculating the ratio and phase difference of the two components, judging the elliptical shape of the current vibration trajectory, and triggering the adjustment of the screen surface inclination angle when the horizontal-vertical ratio deviates from the target value;

[0021] The vibration signals of the four supporting points of the screen box are compared in the time domain, and the time difference of the vibration peak of each supporting point is measured to judge the uniformity of the screen box vibration. When the vibration time difference of the left and right supporting points exceeds the preset threshold, an adjustment instruction is sent to the left or right spring preload device;

[0022] According to the difference between the current vibration trajectory and the target trajectory, the hydraulic system is controlled to extend or shorten the screen surface inclination hydraulic cylinder, changing the angle between the screen surface and the horizontal plane until the screen surface inclination reaches the calculated optimal angle;

[0023] The servo motor drives the spring preload screw to rotate and adjust the spring support stiffness so that the vibration intensity of the screen box reaches the optimal value matching the current coal type until the movement speed of the material on the screen surface meets the preset requirements.

[0024] Optionally, the vibration signals of the four fulcrums of the screen box are compared in the time domain, the time difference of the vibration peaks of each fulcrum is measured, and the uniformity of the vibration of the screen box is judged. When the vibration time difference of the left and right fulcrums exceeds a preset threshold, an adjustment instruction is sent to the left or right spring preload device, including:

[0025] Collect data from acceleration sensors installed at the four support points of the screen box: the left front, right front, left rear, and right rear. Perform low-pass filtering on the raw data and extract the peak time stamp within each signal cycle.

[0026] Calculate the difference between the peak timestamps of the diagonal support points to obtain the torsional vibration parameter of the screen box, and when the torsional vibration parameter is greater than a first preset threshold, increase the preload force of the spring on the side with lower torsional stiffness;

[0027] Calculate the difference in the peak timestamps of the support points on the same side to obtain the swing vibration parameter of the screen box. When the swing vibration parameter is greater than a second preset threshold, increase the preload force of the spring on the side with greater swing.

[0028] The vibration amplitude data of the four supporting points of the screen box are input into the imbalance calculation unit to generate the screen box vibration uniformity index. When the screen box vibration uniformity index exceeds the third preset threshold, the adjustment amount of the spring preload of each supporting point is calculated;

[0029] The calculated spring preload adjustment amount is sent to the servo motor controller of each fulcrum, which drives the servo motor to rotate forward or reverse to adjust the spring compression amount until the vibration time difference of the four fulcrums is less than a fourth preset threshold.

[0030] Optionally, the step of inputting the vibration amplitude data of the four supporting points of the screen box into the imbalance calculation unit to generate the screen box vibration uniformity index includes:

[0031] Normalize the vibration amplitude data of the four supporting points of the screen box, namely the left front, right front, left rear and right rear, so that the value range of each supporting point is unified and a normalized vibration amplitude set is obtained;

[0032] Calculating the difference between the maximum value and the minimum value in the normalized vibration amplitude set and dividing the difference by the average value of the vibration amplitudes of the four supports to obtain the vibration imbalance of the first screen box;

[0033] Performing variance analysis on the vibration amplitudes of the four supporting points, calculating the ratio of the standard deviation to the mean, obtaining the dispersion coefficient of the vibration amplitude, and using the dispersion coefficient as the vibration imbalance degree of the second screen box;

[0034] The first screen box vibration imbalance degree and the second screen box vibration imbalance degree are synthesized by weighted average to obtain a screen box vibration uniformity index;

[0035] According to the numerical range of the screen box vibration uniformity index, the vibration state is divided into a uniform zone, a slightly uneven zone, a moderately uneven zone and a severely uneven zone, and corresponding spring preload adjustment strategies are adopted for different zones.

[0036] Optionally, controlling the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers includes:

[0037] The material samples are collected through the oversize sampler installed at the end of the screen surface and the undersize sampler at the screen outlet, and the screening particle size analysis and quality measurement are performed on the samples to calculate the screening efficiency value of the current screening process;

[0038] Comparing the screening efficiency value with a preset target efficiency threshold, and generating a screening state evaluation result when the screening efficiency value deviates from the preset target efficiency threshold by more than a first deviation range;

[0039] Based on the screening status evaluation results, the frequency converter of the feeder is controlled to adjust the feeding speed, thereby changing the residence time of the material on the screen surface, and the hydraulic cylinder is controlled to adjust the height position of the longitudinal baffle of the screen box to adjust the thickness of the material layer;

[0040] The screening efficiency of the adjusted working condition is monitored for multiple consecutive cycles, and the slope of the screening efficiency change is calculated. When the absolute value of the slope is less than the stability judgment threshold and the efficiency value is within the second deviation range of the preset target efficiency threshold, the current feeding speed and baffle height position are locked;

[0041] When it is detected that the screening efficiency value exceeds the third deviation range of the preset target efficiency threshold or the particle size distribution of the undersize material is abnormal, the start signal of the screen cleaning mechanism is triggered and the vibration intensity parameter is temporarily adjusted to the maximum limit value.

[0042] In a second aspect, the present application provides a vibrating screen parameter adaptive control system based on machine learning, the vibrating screen parameter adaptive control system based on machine learning includes:

[0043] The acquisition module is used to collect data of the coal material at the feed end of the vibrating screen through an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen surface to obtain a characteristic parameter set;

[0044] A control module is used to calculate the current coal load factor and screening difficulty index based on the characteristic parameter set, and adjust the phase difference and current input of the dual-motor vibrator through the control system to obtain the vibration direction and intensity;

[0045] An adjustment module is used to adjust the preload of the screen box support spring and the extension and contraction amount of the screen surface inclination hydraulic cylinder based on the feedback data of the vibration direction and intensity, so as to obtain the screening trajectory and material movement speed;

[0046] The feeding module is used to control the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers.

[0047] In a third aspect, a vibrating screen parameter adaptive control device based on machine learning is provided, comprising: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the vibrating screen parameter adaptive control device based on machine learning executes the above-mentioned vibrating screen parameter adaptive control method based on machine learning.

[0048] In a fourth aspect, a computer-readable storage medium is provided, wherein instructions are stored in the computer-readable storage medium, which, when executed on a computer, enables the computer to execute the above-mentioned vibrating screen parameter adaptive control method based on machine learning.

[0049] In the technical solution provided by this application, by installing an image acquisition device above the feed hopper and deploying a material sensor at the front end of the screen surface, multi-dimensional real-time perception of coal material characteristics is achieved, solving the technical problem of the traditional vibrating screen control system's lack of accurate identification of material characteristics; the construction of the characteristic parameter set adopts deep learning image processing and multi-source heterogeneous data fusion technology to accurately extract and characterize key characteristics such as coal particle size distribution, moisture content, and mineral composition, significantly improving the accuracy and stability of coal type identification; based on the coal load factor and screening difficulty index calculated based on the characteristic parameter set, an accurate mapping relationship between coal physical properties and vibration control parameters is established. The phase difference and current input of the dual-motor vibrator can be adjusted differently according to the characteristics of different coal types, effectively solving the core technical problem that the fixed parameter control method cannot adapt to the changes of multiple coal types; the closed-loop feedback mechanism of vibration direction and intensity realizes precise control of the screening trajectory through the coordinated adjustment of spring preload and screen surface inclination, ensuring that the movement speed of the material on the screen surface always remains in the optimal range, overcoming the problem of screening efficiency fluctuations caused by changes in coal types; finally, the screening efficiency monitoring and feeding speed and baffle height adjustment links build a complete quality closed-loop control system, so that the entire screening system can dynamically respond to changes in coal types and maintain efficient and stable operation.

[0050] The present invention adopts a multi-level machine learning algorithm in the field of coal processing, and the algorithm has made significant contributions to various links: the deep convolutional neural network in image processing can accurately identify the boundaries of coal blocks from complex backgrounds, solving the problem that traditional image recognition algorithms are easily interfered with in harsh industrial environments; the nonlinear embedding algorithm of multi-source data fusion retains the topological relationship between features of different dimensions, effectively solving the difficulty of characterizing the characteristics of multiple coal types; the parameter optimization mechanism based on the improved gray wolf optimization algorithm introduces material adaptability weight adjustment, greatly improving the global search capability in complex nonlinear systems, and solving the defect that traditional optimization methods are prone to falling into local optimality; the multi-objective evaluation system in spring preload adjustment can balance the contradiction between vibration uniformity and energy consumption, meeting the dual requirements of equipment reliability and economy in the coal screening process. In summary, the present invention realizes the adaptive control of the vibrating screen for multiple coal types through the system integration of material characteristic perception, parameter optimization control, vibration state monitoring and quality closed-loop adjustment, solving the key problem that fixed parameters in traditional technology cannot cope with changes in coal types. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0052] Figure 1 Schematic diagram of an embodiment of a vibrating screen parameter adaptive control method based on machine learning in an embodiment of the present application;

[0053] Figure 2 This is a schematic diagram of an embodiment of a vibrating screen parameter adaptive control system based on machine learning in an embodiment of the present application;

[0054] Figure 3 It is a schematic block diagram of the structure of a vibrating screen parameter adaptive control device based on machine learning in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The embodiments of the present application provide a method and system for adaptive control of vibrating screen parameters based on machine learning. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0056] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the vibrating screen parameter adaptive control method based on machine learning includes:

[0057] Step S101: Data of the coal material at the feed end of the vibrating screen is collected by an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen surface to obtain a characteristic parameter set;

[0058] Step S102: Calculate the current coal load factor and screening difficulty index based on the characteristic parameter set, and adjust the phase difference and current input of the dual-motor vibrator through the control system to obtain the vibration direction and intensity;

[0059] Step S103: Based on the feedback data of vibration direction and intensity, the preload of the screen box support spring and the expansion and contraction amount of the screen surface inclination hydraulic cylinder are adjusted to obtain the screening trajectory and material movement speed;

[0060] Step S104: Control the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers.

[0061] It is understandable that the execution subject of this application can be a vibrating screen parameter adaptive control system based on machine learning, or a terminal or a server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0062] Specifically, coal material characteristic data is collected using multi-source sensors. A high-definition camera mounted above the feed hopper captures coal surface images at a rate of 30 frames per second. An improved Canny edge detection algorithm is used to identify coal block boundaries. Regional segmentation is used to calculate the distribution of particle size distributions within different ranges, generating a particle size distribution matrix. A near-infrared spectral sensor array located at the front of the screen detects the spectral reflectance characteristics of the coal. Wavelet transforms are used to eliminate background noise and extract peaks in characteristic bands. These peaks are directly correlated with the coal's moisture content and mineral composition, generating a moisture distribution map and mineral composition fingerprint. Simultaneously, a strain gauge weight sensor at the bottom of the screen box measures material load changes, while triaxial accelerometers on the side walls record vibration. An adaptive filter separates the effective signal from the noise, generating a dynamic load characteristic curve for the coal material. This multi-source heterogeneous data is fused through dimensionality reduction and nonlinear embedding to generate a comprehensive characteristic parameter set. To calculate control indicators based on this characteristic parameter set, the particle size distribution matrix is ​​multiplied by a preset weight coefficient. The sum of these factors is then multiplied by the average moisture content to obtain the first coal load factor. The ratio of clay minerals to hard minerals is extracted from the mineral composition fingerprint and combined with the dynamic load fluctuation coefficient to generate a first screening difficulty index. These two first-level indicators are combined through a nonlinear mapping function and weighted fusion to form a second coal load factor and a second screening difficulty index, the latter of which more accurately reflects the material's sieving ability. A parameter mapping table is constructed based on these indicators to calculate the optimal dual-motor parameters and convert them into control signals. For example, when processing high-hardness, low-moisture coal, the calculated load factor is high, so the system automatically increases the vibration frequency to 45 Hz and reduces the phase difference to 15 degrees to provide stronger vibration energy. When processing low-hardness, high-moisture coal, the system reduces the vibration frequency to 35 Hz and increases the phase difference to 30 degrees to enhance material dispersion. After the vibration control parameters are implemented, the system continuously monitors the vibration effect. The triaxial accelerometer data is decomposed into horizontal and vertical components, and the elliptical shape of the vibration trajectory is determined by calculating the ratio and phase difference. The vibration signals of the four support points of the screen box are compared in the time domain, measuring the time difference between the vibration peaks at each support point to assess vibration uniformity. If the horizontal-to-vertical ratio deviates from the target or the fulcrum vibrates unevenly, the system automatically adjusts the screen box support spring preload and the screen surface inclination angle. The spring preload is adjusted by a servo motor-driven preload screw, while the screen surface inclination angle is controlled by a hydraulic cylinder's extension and retraction. This dual-parameter linkage adjustment ensures the material's optimal trajectory and speed on the screen surface, avoiding the unstable material movement associated with traditional fixed-parameter control. The system evaluates the screening effect in real time using an oversize sampler installed at the end of the screen surface and an undersize sampler at the screen outlet. The sampler collects material samples, performs particle size analysis, and calculates the current screening efficiency value. This value is compared with a preset target threshold to generate a screening status assessment result. Based on this assessment result, the control system adjusts the operating frequency of the feeder's frequency converter and the height position of the screen box's longitudinal baffles to change the material feed rate and the thickness of the material on the screen surface.For example, if the system detects that the screening efficiency is below the target value, it reduces the feed rate and raises the baffle height to extend the material's residence time on the screen surface. When the screening efficiency returns to within the target range, the current parameter settings are locked. This closed-loop control mechanism can cope with the fluctuations in screening difficulty caused by changes in coal type, maintaining stable production quality and efficiency, and solving the problem of decreased efficiency of traditional vibrating screens during coal type changes.

[0063] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0064] The image acquisition device is used to obtain a multi-angle continuous image sequence of the coal surface, and the coal block boundary is identified in the image sequence to obtain a coal particle size distribution characteristic matrix;

[0065] The near-infrared spectral sensor array installed at the front end of the screen is used to perform multi-point scanning of the coal material. After eliminating background noise using wavelet transform, characteristic spectral peaks are extracted to obtain the coal moisture content distribution map and mineral composition fingerprint.

[0066] The signals collected by the strain gauge weight sensor installed at the bottom of the screen box and the triaxial acceleration sensor on the side wall of the screen box are separated by an adaptive filter to obtain the dynamic load characteristic curve of the coal material.

[0067] The coal particle size distribution characteristic matrix, coal moisture content distribution map and mineral composition fingerprint as well as the dynamic load characteristic curve of coal material are reduced in dimension and fused through nonlinear embedding to generate a characteristic parameter set.

[0068] Specifically, an industrial camera array mounted above the feed hopper captures images of the coal surface from different angles at a frequency of 30 frames per second and a resolution of 1920 × 1080 pixels. After grayscale conversion, the acquired raw images are then applied to a modified Canny edge detection algorithm to identify coal lump boundaries. This algorithm uses a dual-threshold method to detect strong and weak edges. The threshold is dynamically determined based on the image histogram, avoiding the edge loss problem caused by a fixed threshold. After edge connection, a morphological closing operation is performed to fill in the edge gaps and obtain a complete coal lump outline. Based on this outline information, the equivalent diameter of each coal lump is calculated, and the number distribution of particles in different size ranges is statistically analyzed to form an n × m dimensional particle size distribution feature matrix, where n represents the number of particle size classes (typically 5–8 classes) and m represents the time series length. Simultaneously, a near-infrared spectral sensor array mounted at the front of the screen scans the coal material at a frequency of 100 Hz, collecting reflectance spectral data within the wavelength range of 900–1700 nm. The raw spectral data contains ambient light interference and random noise, so wavelet transform is used to reduce noise. Specifically, the db4 wavelet basis is selected and the spectral data is decomposed into six layers to obtain wavelet coefficients of different frequencies. The high-frequency coefficients are processed using a soft thresholding method, retaining low-frequency information reflecting material properties and removing high-frequency noise. Wavelet reconstruction yields purified spectral data, and the peak and valley positions of the characteristic bands are extracted. Characteristic peaks related to moisture content are primarily located near 1450nm and 1940nm, while characteristic peaks related to mineral composition are distributed in the 1000-1300nm region. The coal moisture content distribution map and mineral composition fingerprint are calculated using the characteristic peak intensity ratio and displacement.

[0069] To capture the dynamic load characteristics of the coal material, a strain gauge weight sensor at the bottom of the screen box and a triaxial accelerometer on the side wall simultaneously collect data at a sampling frequency of 1000 Hz. These two signals contain a mixture of material load information, equipment vibration information, and environmental interference. Adaptive filtering technology is used to separate the effective signals. Specifically, a minimum mean square error (LMS) adaptive filtering algorithm is used. The acceleration signal is used as the reference input and the weight signal as the primary input. By iteratively adjusting the filter coefficients, the effective components of the material load's impact on the equipment are extracted. The filtered signal is then analyzed through envelope detection to produce a dynamic load curve for the material, which reflects the material's fluidity and resistance characteristics during the screening process.

[0070] The acquired multi-source heterogeneous data (particle size distribution matrix, moisture content distribution map, mineral composition fingerprint, and dynamic load characteristic curve) is fused and processed. Because the various data types have different dimensions and scales, each type of data is first standardized so that all data is mapped to the interval [0,1]. Principal component analysis (PCA) is then used to reduce the dimensionality of the high-dimensional data, selecting principal components with a cumulative contribution exceeding 95%. Finally, the t-SNE nonlinear embedding algorithm is used to project the data into a low-dimensional feature space, preserving the topological relationships between the data and generating a compact set of characteristic parameters. This processing method solves the problem of insufficient material characteristic perception in traditional vibrating screen control, enabling the control system to obtain rich material information and provide a data foundation for subsequent intelligent decision-making. For example, when the type of coal being processed changes from anthracite to bituminous coal, the hardness index in the characteristic parameter set decreases significantly, while the moisture content index increases. The system automatically adjusts the vibration parameters accordingly to maintain stable screening efficiency.

[0071] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0072] The particle size distribution characteristic matrix in the characteristic parameter set is multiplied by a preset weight coefficient, the weighted values ​​of all particle size intervals are summed, and the sum is multiplied by the spatial average value of the moisture content distribution map to obtain a first coal load factor;

[0073] Based on the mineral composition fingerprint and dynamic load characteristic curve in the characteristic parameter set, the ratio of clay mineral content to hard mineral content is calculated, and combined with the fluctuation coefficient to generate the first screening difficulty index;

[0074] Applying a nonlinear mapping function to the first coal load coefficient and performing weighted fusion with the first screening difficulty index to obtain a second screening difficulty index and a second coal load coefficient;

[0075] A parameter mapping table is constructed based on the second coal load factor and the second screening difficulty index, and the phase difference and current input value of the dual-motor vibrator are calculated through an interpolation algorithm;

[0076] The phase difference is converted into a phase control signal, and the current input value is converted into a frequency control signal, which is sent to the motor drive unit through the controller to generate the vibration direction and vibration intensity;

[0077] The actual vibration parameters are measured by the vibration sensor installed on the screen box, compared with the theoretical calculated values, and the compensation amount is calculated through closed-loop control to adjust the vibration direction and vibration intensity.

[0078] Specifically, the particle size distribution characteristic matrix is ​​multiplied by the preset weight coefficients. These weight coefficients reflect the degree of influence of different particle sizes on the difficulty of screening. The coarse particle size (such as >10mm) has a lower weight, and the fine particle size (such as <3mm) has a higher weight because fine-grained materials are more likely to clog the screen. The weighted data of each particle size are summed to obtain a comprehensive particle size index, which is then multiplied by the spatial average value of the moisture content distribution map to form the first coal load coefficient. The higher the moisture content, the worse the material fluidity and the higher the load coefficient. At the same time, the clay mineral content and hard mineral content data are extracted from the mineral composition fingerprint. Clay minerals (such as kaolinite and montmorillonite) make coal easier to stick together, and hard minerals (such as quartz and pyrite) increase the screening resistance. After calculating the ratio of the two, it is combined with the fluctuation coefficient of the dynamic load characteristic curve. The fluctuation coefficient indicates the severity of the load change and reflects the stability of the material movement on the screen surface. These two indicators are weighted and summed to generate the first screening difficulty index.

[0079] The first coal load factor is nonlinearly mapped through a Sigmoid function, which converts the original value into a value in the range of 0-1, making the control smoother while highlighting the differences in the medium load area. The mapped load factor is weighted and fused with the first screening difficulty index. The weight is dynamically adjusted according to the focus of the current screening task to form a more accurate second screening difficulty index and second coal load factor. Based on these two secondary indicators, a two-dimensional parameter mapping table is constructed, with the horizontal axis being the load factor and the vertical axis being the difficulty index. The table stores the optimal vibration parameters corresponding to each combination. The bilinear interpolation algorithm is used to obtain the optimal dual-motor vibrator phase difference and current input value under the current working conditions from the parameter mapping table. The phase difference directly determines the ellipticity of the vibration direction, while the current input value affects the vibration intensity.

[0080] The control system converts the phase difference into a digital instruction for the phase controller, controlling the time difference between the timing triggers of the dual motors. It also converts the current input into a frequency control signal for the inverter, regulating the motor speed and output torque. These control signals are transmitted to the motor drive unit via the fieldbus, generating the desired vibration direction and intensity. A three-axis vibration sensor installed on the screen box collects vibration data in real time, calculating the actual vibration elliptical trajectory parameters and energy distribution. These are compared with the theoretically calculated values ​​to determine the deviation. The control system calculates the compensation amount using a proportional-integral-differential (PID) algorithm, dynamically adjusting the control output to eliminate the effects of external interference and model errors, stabilizing the vibration parameters within the optimal range. For example, when the system detects a decrease in vibration intensity due to material accumulation, it automatically increases the current input value. If the vibration direction deviates from the optimal trajectory, it fine-tunes the phase difference to ensure the highest screening efficiency at all times.

[0081] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0082] Decompose the three-axis acceleration signal in the feedback data of vibration direction and intensity into horizontal and vertical components, calculate the ratio and phase difference of the two components, and judge the elliptical shape of the current vibration trajectory. When the horizontal-vertical ratio deviates from the target value, the screen surface inclination angle adjustment is triggered;

[0083] The vibration signals of the four supporting points of the screen box are compared in the time domain, and the time difference of the vibration peak of each supporting point is measured to judge the uniformity of the screen box vibration. When the vibration time difference of the left and right supporting points exceeds the preset threshold, an adjustment instruction is sent to the left or right spring preload device;

[0084] According to the difference between the current vibration trajectory and the target trajectory, the hydraulic system is controlled to extend or shorten the screen surface inclination hydraulic cylinder, changing the angle between the screen surface and the horizontal plane until the screen surface inclination reaches the calculated optimal angle;

[0085] The servo motor drives the spring preload screw to rotate and adjust the spring support stiffness so that the vibration intensity of the screen box reaches the optimal value matching the current coal type until the movement speed of the material on the screen surface meets the preset requirements.

[0086] Specifically, effective information is extracted from the vibration feedback data collected by the three-axis accelerometer. The three-dimensional acceleration signal (X, Y, and Z) is decomposed into a horizontal component (XY plane) and a vertical component (Z axis) through coordinate transformation. A bandpass filter is applied to the original signal to remove high-frequency noise and low-frequency drift, retaining the effective vibration signal within the range of 0.5 to 2 times the vibration frequency. The amplitude ratio of the horizontal composite acceleration to the vertical acceleration (H / V ratio) is calculated. This ratio directly reflects the elliptical shape of the vibration trajectory. The larger the ratio, the flatter the ellipse, and the more the material's motion trajectory on the screen surface tends to be horizontal. Simultaneously, the fundamental frequency phase of the two components is extracted through Fourier transform, and the phase difference is calculated. This phase difference determines the inclination angle of the elliptical trajectory. When the H / V ratio deviates by more than ±10% from the target value calculated for the current coal type (typically 1.2-1.8), the automatic adjustment mechanism for the screen surface inclination is triggered. For vibration uniformity monitoring, the system collects vibration signals from the four support points of the screen box: the left front, right front, left rear, and right rear. A peak detection algorithm is used to determine the peak moment of each fulcrum signal, and the time interval between adjacent cycles is calculated to ensure the periodic stability of the signal. Then, taking the left front fulcrum as the benchmark, the time difference between the peak occurrence time of the other three fulcrums and the benchmark point is calculated to form three time difference values. Ideally, the time difference between the diagonal fulcrums should be half a cycle (corresponding to a 180-degree phase difference), and the time difference between the fulcrums on the same side should be close to zero. When the absolute value of the vibration time difference between the left and right fulcrums exceeds the preset threshold (usually 5% of the vibration period), it indicates that there is torsional uneven vibration in the screen box, and the spring preload on the corresponding side needs to be adjusted. Specifically, if the left fulcrum is ahead of the right fulcrum, increase the spring preload on the right side; otherwise, increase the preload on the left side.

[0087] Based on the results of vibration trajectory analysis, when the screen surface inclination angle needs to be adjusted, the control system first calculates the optimal angle adjustment. This calculation is based on the deviation between the material properties and the current H / V ratio, and the optimal inclination angle value is predicted using a pre-trained neural network model. The control signal is converted into a standard 4-20mA current signal, which is then used to control the hydraulic system through a proportional valve to extend or shorten the screen surface inclination hydraulic cylinder. The displacement of the hydraulic cylinder is fed back in real time via a position sensor, forming a closed-loop control system that precisely adjusts the angle between the screen surface and the horizontal plane. The hydraulic system uses a slow start-stop strategy to avoid impact during angle adjustment until the screen surface inclination angle reaches the calculated optimal angle, with an average accuracy of ±0.5 degrees.

[0088] Based on the results of the fulcrum vibration uniformity analysis, the control system sends adjustment instructions to the spring preload device. The spring preload device consists of a servo motor, a reducer, and a screw mechanism. The servo motor receives a PWM control signal and drives the preload screw in forward or reverse rotation, changing the compression of the spring and thus adjusting the spring support stiffness. The preload adjustment amount is proportional to the degree of vibration unevenness. A staged adjustment strategy is adopted. After each adjustment, wait 5-10 vibration cycles to observe the effect before deciding whether to continue adjustment. When the vibration parameters are consistent with the preset targets and the material movement speed on the screen surface meets the requirements, the current settings are locked.

[0089] In a specific embodiment, the process of performing the time domain comparison of the vibration signals of the four supporting points of the screen box may specifically include the following steps:

[0090] Collect data from acceleration sensors installed at the four support points of the screen box: the left front, right front, left rear, and right rear. Perform low-pass filtering on the raw data and extract the peak time stamp within each signal cycle.

[0091] Calculate the difference between the peak timestamps of the diagonal support points to obtain the torsional vibration parameter of the screen box, and when the torsional vibration parameter is greater than a first preset threshold, increase the preload force of the spring on the side with lower torsional stiffness;

[0092] Calculate the difference in the peak timestamps of the support points on the same side to obtain the swing vibration parameter of the screen box. When the swing vibration parameter is greater than a second preset threshold, increase the preload force of the spring on the side with greater swing.

[0093] The vibration amplitude data of the four supporting points of the screen box are input into the imbalance calculation unit to generate the screen box vibration uniformity index. When the screen box vibration uniformity index exceeds the third preset threshold, the adjustment amount of the spring preload of each supporting point is calculated;

[0094] The calculated spring preload adjustment amount is sent to the servo motor controller of each fulcrum, which drives the servo motor to rotate forward or reverse to adjust the spring compression amount until the vibration time difference of the four fulcrums is less than a fourth preset threshold.

[0095] Specifically, high-precision piezoelectric accelerometers were installed at the left front, right front, left rear, and right rear pivots of the screen box, with a sampling frequency set to 1000 Hz to ensure the capture of complete vibration waveform characteristics. The collected raw data, which contains mechanical noise and high-frequency interference, was processed through a Butterworth low-pass filter with a cutoff frequency set to three times the fundamental vibration frequency, retaining the main vibration information while filtering out high-frequency noise. After filtering, the data was peak detected. A sliding window method and a dynamic threshold strategy were used to identify the peak points within each vibration cycle. The timestamps of the peak occurrences were recorded to form time series data for the four pivots. Using the acquired peak timestamp data, the difference in the peak occurrence time of the diagonal pivots (left front and right rear, and right front and left rear) was calculated. Ideally, the diagonal pivots should maintain a 180-degree phase difference, corresponding to half the vibration cycle. The deviation between the actual calculated time difference and the ideal value forms a torsional vibration parameter, which reflects the degree of torsion of the screen box around the vertical axis. When the torsional vibration parameter exceeds a first preset threshold (typically 5% of the vibration cycle), it indicates that the screen box has significant torsional deformation and requires adjustment of the torsional stiffness. By analyzing the torsion direction and determining the side with lower torsional stiffness, the control system automatically sends a command to the spring preload device on that side to increase the preload force to balance the torsional effect.

[0096] The screen box swing vibration parameter is calculated by calculating the difference in peak timestamps between the pivot points on the same side (left front and left rear, right front and right rear). Ideally, the pivot points on the same side should vibrate synchronously, with the difference close to zero. When the swing vibration parameter exceeds a second preset threshold (typically 3% of the vibration period), it indicates that the screen box is swaying back and forth. The control system identifies the side with greater swing and increases the spring preload on that side to suppress excessive swing. To assess vibration amplitude uniformity, the system extracts vibration amplitude data from the four pivot points and first normalizes it to eliminate the influence of sensor sensitivity differences. It then calculates the ratio of the difference between the maximum and minimum amplitudes, as well as the ratio of the standard deviation to the mean of the four amplitudes. These two indicators together constitute the screen box vibration uniformity index. When this index exceeds a third preset threshold, a comprehensive adjustment mechanism is triggered. The system calculates the adjustment amount based on the amplitude deviation of each pivot point. The larger the deviation, the larger the adjustment amount, forming a spring preload adjustment plan. The control system converts the calculated preload adjustment amount into a servo motor control signal, which is transmitted via the PROFIBUS fieldbus to the servo driver of each pivot point to control the servo motor's forward or reverse rotation. The motor drives the lead screw through a reduction mechanism, precisely adjusting the spring compression and changing the support stiffness. The system monitors the adjustment results in real time and completes the adjustment process when the vibration time difference between the four support points is less than a fourth preset threshold (less than 2% of the vibration period) and the vibration amplitude deviation rate is less than 5%. This method solves the technical problem of traditional vibrating screens' inability to precisely control vibration uniformity, significantly improving screening efficiency and equipment life.

[0097] In a specific embodiment, the process of inputting the vibration amplitude data of the four supporting points of the screen box into the imbalance calculation unit may specifically include the following steps:

[0098] Normalize the vibration amplitude data of the four supporting points of the screen box, namely the left front, right front, left rear and right rear, so that the value range of each supporting point is unified and a normalized vibration amplitude set is obtained;

[0099] Calculate the difference between the maximum and minimum values ​​in the normalized vibration amplitude set and divide it by the average value of the vibration amplitudes of the four supports to obtain the vibration imbalance of the first screen box;

[0100] Perform variance analysis on the vibration amplitudes of the four supporting points, calculate the ratio of the standard deviation to the mean, and obtain the dispersion coefficient of the vibration amplitude. The dispersion coefficient is used as the vibration imbalance degree of the second screen box.

[0101] The vibration imbalance of the first screen box and the vibration imbalance of the second screen box are synthesized by weighted average to obtain the screen box vibration uniformity index;

[0102] According to the numerical range of the screen box vibration uniformity index, the vibration state is divided into uniform zone, slightly uneven zone, moderately uneven zone and severely uneven zone, and corresponding spring preload adjustment strategies are adopted for different zones.

[0103] Specifically, the vibration amplitude data collected from the four fulcrums of the screen box (left front, right front, left rear, and right rear) are normalized to eliminate the systematic errors caused by differences in sensor sensitivity and changes in installation position. Normalization adopts the maximum and minimum value standardization method, that is, the original vibration amplitude of each fulcrum is subtracted from the minimum value of the four points, and then divided by the four-point amplitude range (maximum value minus minimum value), so that the values ​​of each fulcrum are uniformly mapped to the 0-1 range. This processing makes the vibration data at different positions and times comparable, which facilitates subsequent uniformity analysis. Based on the normalized vibration amplitude set, the vibration imbalance of the first screen box is calculated. This indicator is obtained by dividing the difference between the maximum and minimum values ​​by the average value, and intuitively reflects the extreme value of the vibration amplitude distribution. A large imbalance indicates that the vibration of some fulcrums is too strong or too weak, which may lead to uneven stress distribution of the screen box structure and accelerate equipment wear. At the same time, in order to comprehensively evaluate the vibration uniformity, the vibration amplitudes of the four fulcrums are subjected to variance analysis, the standard deviation is calculated and then divided by the mean to obtain the dispersion coefficient of the vibration amplitude, that is, the vibration imbalance of the second screen box. This coefficient more sensitively reflects the degree of deviation of each support point from the average value and can detect relatively subtle uneven conditions.

[0104] These two complementary imbalance indices are combined into a comprehensive uniformity index through a weighted average. The weighting factors are dynamically adjusted based on the coal material characteristics. When processing high-hardness coal, the weight of the first imbalance is increased (focusing on extreme value control), and when processing high-humidity coal, the weight of the second imbalance is increased (focusing on overall uniformity). The combined uniformity index fully reflects the vibration status of the screen box and serves as the basis for subsequent control decisions.

[0105] Based on a large amount of experimental data and expert experience, the uniformity index is divided into four intervals: uniform zone (0-0.1), slightly uneven zone (0.1-0.2), moderately uneven zone (0.2-0.35) and severely uneven zone (>0.35). Differentiated spring preload adjustment strategies are adopted for different areas: no adjustment is required in the uniform zone; stability adjustment is adopted in the slightly uneven zone, and the preload of the fulcrum with the lowest vibration amplitude is adjusted by a small amplitude (3-5%); contrast adjustment is adopted in the moderate uneven zone, reducing the preload of the high-amplitude fulcrum (5-10%) while increasing the preload of the low-amplitude fulcrum (5-10%); reconstruction adjustment is adopted in the severely uneven zone, first resetting the preload of the four fulcrums to the standard value, and then redistributing the preload differences according to the current material characteristics to make the vibration energy distribution more reasonable. This vibration uniformity assessment and adjustment mechanism solves the technical problem that the traditional vibrating screen fixed parameter control method cannot accurately monitor and adjust the vibration state. Through data-driven adaptive control, it achieves accurate matching of vibration parameters and coal characteristics.

[0106] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0107] The material samples are collected through the oversize sampler installed at the end of the screen surface and the undersize sampler at the screen outlet, and the screening particle size analysis and quality measurement are performed on the samples to calculate the screening efficiency value of the current screening process;

[0108] Comparing the screening efficiency value with a preset target efficiency threshold, and generating a screening state evaluation result when the screening efficiency value deviates from the preset target efficiency threshold by more than a first deviation range;

[0109] Based on the screening status evaluation results, the feeder frequency converter is controlled to adjust the feeding speed, change the residence time of the material on the screen surface, and at the same time, the hydraulic cylinder is controlled to adjust the height position of the longitudinal baffle of the screen box to adjust the thickness of the material layer;

[0110] The screening efficiency of the adjusted working condition is monitored for multiple consecutive cycles, and the slope of the screening efficiency change is calculated. When the absolute value of the slope is less than the stability judgment threshold and the efficiency value is within the second deviation range of the preset target efficiency threshold, the current feeding speed and baffle height position are locked;

[0111] When it is detected that the screening efficiency value exceeds the third deviation range of the preset target efficiency threshold or the particle size distribution of the undersize material is abnormal, the start signal of the screen cleaning mechanism is triggered and the vibration intensity parameter is temporarily adjusted to the maximum limit value.

[0112] Specifically, a mechanical sampler installed at the end of the screen and an automatic sampling device at the screen outlet regularly collect samples of the oversize and undersize material. The sampling cycle is 5 minutes, and each sample volume is approximately 500 grams. The collected samples are then graded by an automatic screening system consisting of five layers of standard screens with corresponding apertures of 10mm, 5mm, 3mm, 1mm, and 0.5mm. The mass of each particle size is measured using vibratory screening and weight sensors, and the screening efficiency is calculated using the following formula: Screening efficiency = (mass of target particle size in the undersize material / total mass of target particle size in the feed) × 100%. The target particle size is determined based on production process requirements; in coal processing, it is typically fine particles less than 3mm. The calculated screening efficiency value is compared with a preset target efficiency threshold, which is pre-set based on the characteristics of the coal type and production requirements, generally between 85% and 95%. When the screening efficiency deviates from the target efficiency by more than a first deviation range (±5%), a status assessment mechanism is triggered. Evaluation results are categorized into three categories: under-efficiency (less than -5% of the target), moderate efficiency (within ±5% of the target), and over-efficiency (greater than +5% of the target). These evaluations serve as the basis for subsequent parameter adjustments, with under-efficiency requiring special attention as it directly impacts product quality and production efficiency.

[0113] Based on the evaluation results, the control system performs dual-parameter coordinated adjustment. When it detects that the efficiency is too low, it simultaneously slows down the operating frequency of the feeder frequency converter (reduced by 10-20%) and increases the height of the longitudinal baffle of the screen box (raised by 10-30mm). These two operations work together to extend the residence time of the material on the screen surface on the one hand, and increase the thickness of the material layer on the other hand, strengthening the mutual extrusion and friction between the materials, and promoting the passage of fine particles through the sieve holes. Conversely, when the efficiency is too high (which may be accompanied by a decrease in output), the feed speed is appropriately increased and the baffle height is lowered to increase the processing capacity.

[0114] After adjusting the parameters, the system enters the monitoring stabilization phase, continuously monitoring screening efficiency data for 5-8 cycles (approximately 25-40 minutes) and calculating the efficiency change slope. This slope, derived through linear regression, represents the trend of screening efficiency over time. When the absolute value of the slope falls below the stability threshold (typically 0.2% / minute) and the efficiency value is within the second deviation range (±3%) of the target threshold, the system is deemed to have reached stability and the current feed rate and baffle height settings are locked. The system also features abnormality detection and handling capabilities. If a sudden drop in screening efficiency exceeding the third deviation range (±10%) or a significant abnormal particle size distribution in the undersize material (such as a sudden increase in the proportion of large particles) is detected, it indicates a possible screen blockage. This automatically triggers the screen cleaning mechanism's activation signal and temporarily adjusts the vibration intensity to the maximum limit (increase by 15-25%) for 30-60 seconds before returning to normal settings. This short burst of intense vibration helps clear blocked screen holes and restore normal screening.

[0115] The above describes the vibrating screen parameter adaptive control method based on machine learning in the embodiment of the present application. The following describes the vibrating screen parameter adaptive control system based on machine learning in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the vibrating screen parameter adaptive control system based on machine learning includes:

[0116] The acquisition module 201 is used to acquire data of the coal material at the feed end of the vibrating screen through an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen surface to obtain a characteristic parameter set;

[0117] A control module 202 is configured to calculate the current coal load factor and screening difficulty index based on the characteristic parameter set, and adjust the phase difference and current input of the dual-motor vibrator through the control system to obtain the vibration direction and intensity;

[0118] An adjustment module 203 is configured to adjust the preload of the screen box support spring and the expansion and contraction of the screen surface inclination hydraulic cylinder based on the feedback data of the vibration direction and intensity, thereby obtaining a screening trajectory and a material movement speed;

[0119] The feeding module 204 is used to control the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers.

[0120] above Figure 2 The vibrating screen parameter adaptive control system based on machine learning in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The vibrating screen parameter adaptive control device based on machine learning in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0121] Reference Figure 3 In an embodiment of the present invention, a vibration screen parameter adaptive control device based on machine learning is also provided. The vibration screen parameter adaptive control device based on machine learning can be a server, and its internal structure can be as follows: Figure 3 As shown. The vibrating screen parameter adaptive control device based on machine learning includes a processor, a memory, a display screen, an input device, a network interface and a database connected via a system bus. The computer-designed processor is used to provide computing and control capabilities. The memory of the vibrating screen parameter adaptive control device based on machine learning includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the vibrating screen parameter adaptive control device based on machine learning is used to store the corresponding data in this embodiment. The network interface of the vibrating screen parameter adaptive control device based on machine learning is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above method is implemented.

[0122] Those skilled in the art will understand that Figure 3 The structure shown in is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the machine learning-based vibrating screen parameter adaptive control device to which the solution of the present invention is applied.

[0123] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of a vibrating screen parameter adaptive control method based on machine learning.

[0124] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0125] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a machine learning-based vibrating screen parameter adaptive control device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0126] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A vibrating screen parameter adaptive control method based on machine learning, characterized in that: The method comprises: The coal material at the feed end of the vibrating screen is collected for data by an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen surface to obtain a characteristic parameter set, including: obtaining a multi-angle continuous image sequence of the coal surface by the image acquisition device, identifying coal block boundaries on the image sequence, and obtaining a coal particle size distribution characteristic matrix; performing multi-point scanning on the coal material by a near-infrared spectral sensor array installed at the front end of the screen surface, extracting characteristic spectrum peaks after eliminating background noise by wavelet transform, and obtaining a coal moisture content distribution map and a mineral composition fingerprint; performing signal separation on the signals collected by the strain gauge weight sensor installed at the bottom of the screen box and the three-axis acceleration sensor on the side wall of the screen box through an adaptive filter to obtain a dynamic load characteristic curve of the coal material; performing dimensionality reduction on the coal particle size distribution characteristic matrix, the coal moisture content distribution map and the mineral composition fingerprint, and the coal material dynamic load characteristic curve, and performing feature fusion through nonlinear embedding to generate a characteristic parameter set; The current coal load factor and screening difficulty index are calculated according to the characteristic parameter set, and the phase difference and current input of the dual-motor vibrator are adjusted by the control system to obtain the vibration direction and intensity, including: multiplying the particle size distribution characteristic matrix in the characteristic parameter set by a preset weight coefficient, summing the weighted values ​​of all particle size intervals, and multiplying it with the spatial average value of the moisture content distribution map to obtain a first coal load factor; based on the mineral composition fingerprint and dynamic load characteristic curve in the characteristic parameter set, calculating the ratio of clay mineral content to hard mineral content, and generating a first screening difficulty index in combination with the fluctuation coefficient; applying a nonlinear mapping function to the first coal load coefficient number, and perform weighted fusion with the first screening difficulty index to obtain the second screening difficulty index and the second coal load factor; construct a parameter mapping table according to the second coal load factor and the second screening difficulty index, and calculate the phase difference and current input value of the dual-motor vibrator through the interpolation algorithm; convert the phase difference into a phase control signal, and convert the current input value into a frequency control signal, which is sent to the motor drive unit through the controller to generate the vibration direction and vibration intensity; use the vibration sensor installed on the screen box to measure the actual vibration parameters, compare them with the theoretical calculated values, calculate the compensation amount through closed-loop control, and adjust them to the vibration direction and vibration intensity; Based on the feedback data of the vibration direction and intensity, the preload of the screen box support spring and the expansion and contraction of the screen surface inclination hydraulic cylinder are adjusted to obtain the screening trajectory and material movement speed; According to the screening efficiency data detected by the undersize and oversize samplers, the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box are controlled.

2. The vibrating screen parameter adaptive control method based on machine learning according to claim 1 is characterized in that: Based on the feedback data of the vibration direction and intensity, the preload of the screen box support spring and the expansion and contraction amount of the screen surface inclination hydraulic cylinder are adjusted to obtain the screening trajectory and material movement speed, including: Decomposing the three-axis acceleration signal in the feedback data of the vibration direction and intensity into horizontal and vertical components, calculating the ratio and phase difference of the two components, judging the elliptical shape of the current vibration trajectory, and triggering the adjustment of the screen surface inclination angle when the horizontal-vertical ratio deviates from the target value; The vibration signals of the four supporting points of the screen box are compared in the time domain, and the time difference of the vibration peak of each supporting point is measured to judge the uniformity of the screen box vibration. When the vibration time difference of the left and right supporting points exceeds the preset threshold, an adjustment instruction is sent to the left or right spring preload device; According to the difference between the current vibration trajectory and the target trajectory, the hydraulic system is controlled to extend or shorten the screen surface inclination hydraulic cylinder, changing the angle between the screen surface and the horizontal plane until the screen surface inclination reaches the calculated optimal angle; The servo motor drives the spring preload screw to rotate and adjust the spring support stiffness so that the vibration intensity of the screen box reaches the optimal value matching the current coal type until the movement speed of the material on the screen surface meets the preset requirements.

3. The vibrating screen parameter adaptive control method based on machine learning according to claim 2 is characterized in that: The vibration signals of the four supports of the screen box are compared in the time domain, the time difference of the vibration peaks of each support is measured, and the uniformity of the screen box vibration is judged. When the vibration time difference of the left and right supports exceeds a preset threshold, an adjustment instruction is sent to the left or right spring preload device, including: Collect data from acceleration sensors installed at the four support points of the screen box: the left front, right front, left rear, and right rear. Perform low-pass filtering on the raw data and extract the peak time stamp within each signal cycle. Calculate the difference between the peak timestamps of the diagonal support points to obtain the torsional vibration parameter of the screen box, and when the torsional vibration parameter is greater than a first preset threshold, increase the preload force of the spring on the side with lower torsional stiffness; Calculate the difference in the peak timestamps of the support points on the same side to obtain the swing vibration parameter of the screen box. When the swing vibration parameter is greater than a second preset threshold, increase the preload force of the spring on the side with greater swing. The vibration amplitude data of the four supporting points of the screen box are input into the imbalance calculation unit to generate the screen box vibration uniformity index. When the screen box vibration uniformity index exceeds the third preset threshold, the adjustment amount of the spring preload of each supporting point is calculated; The calculated spring preload adjustment amount is sent to the servo motor controller of each fulcrum, which drives the servo motor to rotate forward or reverse to adjust the spring compression amount until the vibration time difference of the four fulcrums is less than a fourth preset threshold.

4. The vibrating screen parameter adaptive control method based on machine learning according to claim 3 is characterized in that: The vibration amplitude data of the four supporting points of the screen box are input into the imbalance calculation unit to generate the screen box vibration uniformity index, including: Normalize the vibration amplitude data of the four supporting points of the screen box, namely the left front, right front, left rear and right rear, so that the value range of each supporting point is unified and a normalized vibration amplitude set is obtained; Calculating the difference between the maximum value and the minimum value in the normalized vibration amplitude set and dividing the difference by the average value of the vibration amplitudes of the four supports to obtain the vibration imbalance of the first screen box; Performing variance analysis on the vibration amplitudes of the four supporting points, calculating the ratio of the standard deviation to the mean, obtaining the dispersion coefficient of the vibration amplitude, and using the dispersion coefficient as the vibration imbalance degree of the second screen box; The first screen box vibration imbalance degree and the second screen box vibration imbalance degree are synthesized by weighted average to obtain a screen box vibration uniformity index; According to the numerical range of the screen box vibration uniformity index, the vibration state is divided into a uniform zone, a slightly uneven zone, a moderately uneven zone and a severely uneven zone, and corresponding spring preload adjustment strategies are adopted for different zones.

5. The vibrating screen parameter adaptive control method based on machine learning according to claim 1, characterized in that: The method of controlling the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers includes: The material samples are collected through the oversize sampler installed at the end of the screen surface and the undersize sampler at the screen outlet, and the screening particle size analysis and quality measurement are performed on the samples to calculate the screening efficiency value of the current screening process; Comparing the screening efficiency value with a preset target efficiency threshold, and generating a screening state evaluation result when the screening efficiency value deviates from the preset target efficiency threshold by more than a first deviation range; Based on the screening status evaluation results, the frequency converter of the feeder is controlled to adjust the feeding speed, thereby changing the residence time of the material on the screen surface, and the hydraulic cylinder is controlled to adjust the height position of the longitudinal baffle of the screen box to adjust the thickness of the material layer; The screening efficiency of the adjusted working condition is monitored for multiple consecutive cycles, and the slope of the screening efficiency change is calculated. When the absolute value of the slope is less than the stability judgment threshold and the efficiency value is within the second deviation range of the preset target efficiency threshold, the current feeding speed and baffle height position are locked; When it is detected that the screening efficiency value exceeds the third deviation range of the preset target efficiency threshold or the particle size distribution of the undersize material is abnormal, the start signal of the screen cleaning mechanism is triggered and the vibration intensity parameter is temporarily adjusted to the maximum limit value.

6. A vibrating screen parameter adaptive control system based on machine learning, characterized in that: For implementing the vibrating screen parameter adaptive control method based on machine learning according to any one of claims 1 to 5, the vibrating screen parameter adaptive control system based on machine learning comprises: An acquisition module is used to acquire data from the coal material at the feed end of the vibrating screen through an image acquisition device installed above the feed hopper and a material sensor installed at the front end of the screen surface to obtain a characteristic parameter set, including: obtaining a multi-angle continuous image sequence of the coal surface through the image acquisition device, identifying coal block boundaries on the image sequence, and obtaining a coal particle size distribution characteristic matrix; performing multi-point scanning on the coal material using a near-infrared spectral sensor array installed at the front end of the screen surface, extracting characteristic spectral peaks after eliminating background noise using wavelet transform, and obtaining a coal moisture content distribution map and a mineral composition fingerprint; performing signal separation on the signals collected by the strain gauge weight sensor installed at the bottom of the screen box and the three-axis acceleration sensor on the side wall of the screen box through an adaptive filter to obtain a dynamic load characteristic curve of the coal material; performing dimensionality reduction on the coal particle size distribution characteristic matrix, the coal moisture content distribution map and the mineral composition fingerprint, and the coal material dynamic load characteristic curve, and performing feature fusion through nonlinear embedding to generate a characteristic parameter set; A control module is used to calculate the current coal load factor and screening difficulty index based on the characteristic parameter set, and adjust the phase difference and current input of the dual-motor vibrator through the control system to obtain the vibration direction and intensity, including: multiplying the particle size distribution characteristic matrix in the characteristic parameter set by a preset weight coefficient, summing the weighted values ​​of all particle size intervals, and multiplying it with the spatial average value of the moisture content distribution map to obtain a first coal load factor; based on the mineral composition fingerprint and dynamic load characteristic curve in the characteristic parameter set, calculating the ratio of clay mineral content to hard mineral content, and generating a first screening difficulty index in combination with the fluctuation coefficient; applying nonlinear dynamic load ... A mapping function is constructed and weightedly integrated with the first screening difficulty index to obtain a second screening difficulty index and a second coal load factor; a parameter mapping table is constructed according to the second coal load factor and the second screening difficulty index, and the phase difference and current input value of the dual-motor vibrator are calculated by an interpolation algorithm; the phase difference is converted into a phase control signal, and the current input value is converted into a frequency control signal, which is sent to the motor drive unit through a controller to generate a vibration direction and vibration intensity; the actual vibration parameters are measured using a vibration sensor installed on the screen box, compared with the theoretical calculated value, and the compensation amount is calculated through closed-loop control and adjusted to the vibration direction and vibration intensity; An adjustment module is used to adjust the preload of the screen box support spring and the extension and contraction amount of the screen surface inclination hydraulic cylinder based on the feedback data of the vibration direction and intensity, so as to obtain the screening trajectory and material movement speed; The feeding module is used to control the feeding speed of the feeder and the height position of the longitudinal baffle of the screen box according to the screening efficiency data detected by the undersize and oversize samplers.

7. A vibrating screen parameter adaptive control device based on machine learning, characterized in that: It includes a memory and a processor, the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the vibrating screen parameter adaptive control method based on machine learning as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the processor is caused to execute the vibrating screen parameter adaptive control method based on machine learning according to any one of claims 1 to 5.

Images